WEBVTT

00:00:03.680 --> 00:00:07.200
Hey there. Welcome to Data Democracy, a podcast

00:00:07.200 --> 00:00:10.300
where we explore ways to make data and AI more

00:00:10.300 --> 00:00:15.199
accessible to everybody. We do this by interviewing

00:00:15.199 --> 00:00:18.199
experts across industries and asking them how

00:00:18.199 --> 00:00:21.280
they think about data, what are some of the challenges

00:00:21.280 --> 00:00:25.000
they face when it comes to data, and also if

00:00:25.000 --> 00:00:27.960
they had a magic wand, to get whatever they needed.

00:00:28.320 --> 00:00:31.640
what kind of intelligence and models they would

00:00:31.640 --> 00:00:34.880
love to have. We've got a great guest today,

00:00:35.399 --> 00:00:39.140
Mary Beth Mashkovis. Mary Beth is an entrepreneur,

00:00:39.500 --> 00:00:43.380
a thought leader in data digital marketing. She

00:00:43.380 --> 00:00:46.719
is the founder of InsightLine Analytics, which

00:00:46.719 --> 00:00:51.679
is a digital consultancy agency focused on delivering

00:00:51.679 --> 00:00:56.500
customer impact. She also runs the digital team

00:00:56.500 --> 00:01:02.030
at Arch Digital, previously ran analytics marketing

00:01:02.030 --> 00:01:07.430
teams at eCapacity, Fantastic ES, Tyak, and various

00:01:07.430 --> 00:01:11.329
other companies in Europe and the U .S. Marybeth

00:01:11.329 --> 00:01:14.709
is a thought leader who talks about leadership,

00:01:14.989 --> 00:01:19.030
data coaching, customer -centric marketing, data

00:01:19.030 --> 00:01:23.409
strategies. I just love reading through Marybeth's

00:01:23.409 --> 00:01:27.140
blogs and articles because like me, she's focused

00:01:27.140 --> 00:01:32.120
on customers' needs first and then kind of building

00:01:32.120 --> 00:01:35.980
solutions to suit those needs versus the other

00:01:35.980 --> 00:01:39.299
way around, which I feel very frustrated about.

00:01:39.900 --> 00:01:42.480
But yeah, we have a ton of stuff to talk about,

00:01:43.319 --> 00:01:47.939
so without further ado, welcome, Mary Beth. Thanks

00:01:47.939 --> 00:01:50.319
so much for having me. Like I said before, it's

00:01:50.319 --> 00:01:52.519
so nice to catch up again. It's been a while

00:01:52.519 --> 00:01:55.140
since we really talked face -to -face. Awesome.

00:01:55.200 --> 00:01:57.359
Thank you so much again for being on our podcast.

00:01:58.340 --> 00:02:02.459
Yeah, absolutely. Awesome. Awesome. All right.

00:02:03.939 --> 00:02:07.260
Let's get started and tell us something about,

00:02:07.439 --> 00:02:09.659
tell us about your journey so far. How did you

00:02:09.659 --> 00:02:12.120
get started and how do you get where you are?

00:02:13.020 --> 00:02:15.479
Yeah, that's always a really important question

00:02:15.479 --> 00:02:18.060
to ask. And I tell this story a few different

00:02:18.060 --> 00:02:22.009
ways, but I think for this, How I ended up in

00:02:22.009 --> 00:02:25.289
digital analytics, which as a field, when I got

00:02:25.289 --> 00:02:28.189
started a little over 10 years ago, wasn't that

00:02:28.189 --> 00:02:31.090
established yet, was that I actually have an

00:02:31.090 --> 00:02:34.389
evolutionary biology background. And during that,

00:02:34.599 --> 00:02:37.479
I took classes in advanced statistics and I learned

00:02:37.479 --> 00:02:41.659
R. And just by some happy happenstance, I got

00:02:41.659 --> 00:02:44.759
recruited at a custom software company that also

00:02:44.759 --> 00:02:48.719
is a consulting firm as a marketing analyst because

00:02:48.719 --> 00:02:52.599
they had this idea that marketing was going to

00:02:52.599 --> 00:02:54.680
get a lot more data centric and that they were

00:02:54.680 --> 00:02:56.900
going to need people with like a scientific background

00:02:56.900 --> 00:02:59.439
to be able to evaluate results. They were right,

00:02:59.460 --> 00:03:01.879
obviously, because things have blown up since

00:03:01.879 --> 00:03:05.129
then. So that's really where I got started. really

00:03:05.129 --> 00:03:07.969
fell in love with digital as a whole but was

00:03:07.969 --> 00:03:10.990
really enticed by data and what it could do and

00:03:10.990 --> 00:03:14.030
how you could use it to make a tangible difference

00:03:14.030 --> 00:03:16.330
that you could see and measure in businesses.

00:03:17.169 --> 00:03:20.969
So my career took me abroad and I spent almost

00:03:20.969 --> 00:03:23.849
six years actually in Copenhagen working for

00:03:23.849 --> 00:03:26.310
several companies including Kayak like you mentioned

00:03:26.310 --> 00:03:29.430
and consulting which I've been back and forth

00:03:29.430 --> 00:03:32.860
in -house and as a consultant and so when I came

00:03:32.860 --> 00:03:35.620
back to the U .S. in 2020, trying to decide what

00:03:35.620 --> 00:03:37.960
to do. I felt like it was the right time for

00:03:37.960 --> 00:03:40.419
me to start my own consulting firm and consult

00:03:40.419 --> 00:03:42.280
in like the way I wanted and about the things

00:03:42.280 --> 00:03:45.080
that I wanted. That's awesome. That's awesome.

00:03:45.219 --> 00:03:49.360
Yeah, I mean, it definitely, when I talk to you

00:03:49.360 --> 00:03:52.020
and kind of read through your content, it definitely

00:03:52.020 --> 00:03:55.000
comes across that you're passionate about both

00:03:55.000 --> 00:03:58.680
data and marketing. So yeah, you've got the R

00:03:58.680 --> 00:04:01.259
background and data background that really shows.

00:04:01.780 --> 00:04:05.280
That's awesome. What does your day to day look

00:04:05.280 --> 00:04:10.280
like at the Insights line? Yeah, so I kind of

00:04:10.280 --> 00:04:12.840
wear a few different hats now because so the

00:04:12.840 --> 00:04:15.360
firm has grown and now we're about six people.

00:04:15.860 --> 00:04:18.540
So my day to day is kind of a combination of

00:04:18.540 --> 00:04:21.100
serving our clients and then also running a business

00:04:21.100 --> 00:04:25.769
now. So I often have one or two larger clients

00:04:25.769 --> 00:04:28.629
that I'm serving either an interim role with.

00:04:28.810 --> 00:04:31.050
So sometimes I'll sit as like a director of analytics

00:04:31.050 --> 00:04:33.829
or something like that, or running a data strategy

00:04:33.829 --> 00:04:36.250
project for a client. And then on the other side,

00:04:36.410 --> 00:04:39.670
I'm creating content for our business and then

00:04:39.670 --> 00:04:42.889
also coaching analysts and just running the team

00:04:42.889 --> 00:04:45.189
in general. So, you know, it's a pretty right

00:04:45.189 --> 00:04:47.290
now, a pretty balanced schedule between the two

00:04:47.290 --> 00:04:51.500
and business development, of course. but still

00:04:51.500 --> 00:04:54.360
get to do quite a lot of hands -on work as well.

00:04:54.959 --> 00:04:56.579
That's awesome. That's awesome. Yeah, I mean,

00:04:56.579 --> 00:04:59.519
that's a lot of different roles, different hats

00:04:59.519 --> 00:05:03.040
that you wear. And also, it's always good to

00:05:03.040 --> 00:05:08.399
kind of stay grounded with hands -on work. That

00:05:08.399 --> 00:05:13.649
keeps me sane a lot of times. Yeah, definitely.

00:05:14.329 --> 00:05:16.009
I have to say, you know, there's kind of two

00:05:16.009 --> 00:05:18.550
approaches to leadership. There's like some people

00:05:18.550 --> 00:05:20.769
who completely let go of whatever technical work

00:05:20.769 --> 00:05:23.230
and don't get into the weeds ever. And then there's

00:05:23.230 --> 00:05:25.050
others that try and keep that alive. And, you

00:05:25.050 --> 00:05:27.370
know, whether it will drive my team crazy or

00:05:27.370 --> 00:05:29.740
not, I still want to. stay at least a little

00:05:29.740 --> 00:05:31.519
bit up to date. I'm letting go of some of the

00:05:31.519 --> 00:05:34.399
details. I'm not going to be our, you know, in

00:05:34.399 --> 00:05:36.660
-house GA4 expert for every single tag or anything

00:05:36.660 --> 00:05:39.319
like that anymore. But, you know, I also make

00:05:39.319 --> 00:05:41.120
sure that I'm still involved in projects and

00:05:41.120 --> 00:05:42.699
understand, like, the trends of what's going

00:05:42.699 --> 00:05:47.019
on in the industry. Yeah, that's exactly where,

00:05:47.019 --> 00:05:51.019
you know, how I like to approach as well. I mean,

00:05:51.060 --> 00:05:53.339
it's a balance. You don't want to get too hands

00:05:53.339 --> 00:05:56.829
-on and, you know. frustrate the team. Right.

00:05:57.110 --> 00:05:59.290
Yeah. I don't know what their opinion of it is,

00:05:59.370 --> 00:06:00.990
but most of the time they tell me they're happy.

00:06:01.529 --> 00:06:03.410
So I think I'm doing a pretty good job. And I

00:06:03.410 --> 00:06:05.610
try and also like, if I feel like it's something

00:06:05.610 --> 00:06:07.449
that will step on their toes, I'll like run a

00:06:07.449 --> 00:06:09.269
project on my own or completely separate from

00:06:09.269 --> 00:06:11.329
the company, just to make sure that I'm still

00:06:11.329 --> 00:06:14.569
fresh with my knowledge. Okay. Yeah, that's a

00:06:14.569 --> 00:06:20.250
great approach. That's awesome. And any tips

00:06:20.250 --> 00:06:23.449
for aspiring entrepreneurs in the digital marketing

00:06:23.449 --> 00:06:26.829
world? What have you learned and what would you

00:06:26.829 --> 00:06:31.529
suggest? I think for someone who's starting out,

00:06:32.550 --> 00:06:37.529
it does really help to have a strong background

00:06:37.529 --> 00:06:40.930
in that field already. I know that there's a

00:06:40.930 --> 00:06:44.589
lot of people who do try and start. companies

00:06:44.589 --> 00:06:46.949
right out of school or whatnot with a lot of

00:06:46.949 --> 00:06:49.430
experience. But for me personally, having that

00:06:49.430 --> 00:06:51.209
really strong consulting background where I was

00:06:51.209 --> 00:06:55.269
able to get a lot of things right and get some

00:06:55.269 --> 00:06:58.350
things wrong on someone else's dime before I

00:06:58.350 --> 00:07:00.009
started my own company was really beneficial.

00:07:01.209 --> 00:07:04.490
Marketing and data are... you know, such broad

00:07:04.490 --> 00:07:06.889
fields that I'd say that you don't really know

00:07:06.889 --> 00:07:08.769
exactly where you want to niche down into. And

00:07:08.769 --> 00:07:12.170
you really do need to do that to be successful

00:07:12.170 --> 00:07:15.149
until you try on a few different things. And

00:07:15.149 --> 00:07:18.050
it's much easier to do that while you're employed

00:07:18.050 --> 00:07:21.649
and you have that safety net than it is to just

00:07:21.649 --> 00:07:23.970
kind of throw yourself into something and be

00:07:23.970 --> 00:07:25.610
not sure if that's like the direction that you

00:07:25.610 --> 00:07:28.550
really want to go. Yeah, I mean, I totally agree

00:07:28.550 --> 00:07:31.689
with that. I mean, I started my career as an

00:07:31.689 --> 00:07:35.089
entrepreneur. we started building websites for

00:07:35.089 --> 00:07:39.250
pocket money. But then we quickly realized that

00:07:39.250 --> 00:07:43.629
we make so many mistakes that it costs us a lot

00:07:43.629 --> 00:07:47.370
if we are running a company and also learning

00:07:47.370 --> 00:07:50.550
at the same time. Like you said, if you can,

00:07:50.670 --> 00:07:54.110
just go learn, make mistakes on somebody else's.

00:07:56.339 --> 00:07:59.100
And if any of my past employers are listening,

00:07:59.279 --> 00:08:02.319
I promise I didn't make that many mistakes. It

00:08:02.319 --> 00:08:05.720
just gives you some more flexibility and you're

00:08:05.720 --> 00:08:08.360
not worrying about, like you said, running a

00:08:08.360 --> 00:08:10.420
business is a whole skill set in its own. And

00:08:10.420 --> 00:08:13.079
I felt really happy that I really knew how to

00:08:13.079 --> 00:08:15.240
consult and what I want to consult on and how

00:08:15.240 --> 00:08:16.839
I was going to do my packages and everything.

00:08:17.139 --> 00:08:20.720
And I could just focus on how do I do taxes and

00:08:20.720 --> 00:08:22.139
how do I do all these other things that I've

00:08:22.139 --> 00:08:25.279
never had to worry about before. Right. And if

00:08:25.279 --> 00:08:28.240
any of my employers are listening, yes, I'm making

00:08:28.240 --> 00:08:30.300
a lot of mistakes. I will make a lot of mistakes,

00:08:31.100 --> 00:08:34.740
but don't make that rule. I mean, from my friends

00:08:34.740 --> 00:08:38.159
that don't make the same mistake twice as long

00:08:38.159 --> 00:08:42.279
as you keep that going, you know. Cool. That's

00:08:42.279 --> 00:08:46.600
awesome. You talked about R and kind of data

00:08:46.600 --> 00:08:51.779
analysis and everything. How is data after your

00:08:51.779 --> 00:08:54.620
career so far, how has this data played a role

00:08:54.620 --> 00:08:57.740
in your career? Ooh, yeah, that's a good one.

00:08:57.940 --> 00:09:01.940
I mean, I think that having that kind of hard

00:09:01.940 --> 00:09:06.919
quote unquote data skills to work with has really

00:09:06.919 --> 00:09:09.840
defined how I've gone about roles, even when

00:09:09.840 --> 00:09:12.779
I was in more marketing roles. So when I was

00:09:12.779 --> 00:09:15.779
a marketing director, et cetera. I think what

00:09:15.779 --> 00:09:18.220
it did was it gave me a completely different

00:09:18.220 --> 00:09:21.279
perspective than someone who is coming from more

00:09:21.279 --> 00:09:24.480
of a traditional marketing background or PR or

00:09:24.480 --> 00:09:26.779
something like that, because I was looking, I

00:09:26.779 --> 00:09:29.120
was really results driven and I was able to kind

00:09:29.120 --> 00:09:32.159
of do a lot of the analysis I wanted to do myself,

00:09:32.679 --> 00:09:34.340
where a lot of other marketers sometimes have

00:09:34.340 --> 00:09:38.279
to wait for a data team to deliver on something

00:09:38.279 --> 00:09:40.419
and then like hope that the person got their

00:09:40.419 --> 00:09:42.399
business context question and everything like

00:09:42.399 --> 00:09:44.870
that. So that's really defined my career. As

00:09:44.870 --> 00:09:48.470
I got into the latter part of working for companies,

00:09:48.649 --> 00:09:50.450
I realized that I really wanted to specialize

00:09:50.450 --> 00:09:54.049
into data and be an analyst, like I started as

00:09:54.049 --> 00:09:57.049
well, and evaluate both marketing and product

00:09:57.049 --> 00:10:01.629
data as my focus. But yeah, starting with that,

00:10:01.970 --> 00:10:03.970
now you would call it almost like a data science

00:10:03.970 --> 00:10:07.460
background of having R. was really helpful in

00:10:07.460 --> 00:10:09.600
that because I was able to do things super efficiently

00:10:09.600 --> 00:10:12.019
to things that like with unwieldy spreadsheets

00:10:12.019 --> 00:10:14.679
etc. I already kind of had some skill sets there

00:10:14.679 --> 00:10:17.100
to get that done not that like I would say it's

00:10:17.100 --> 00:10:19.100
like the most important skill but it's one of

00:10:19.100 --> 00:10:22.259
the important skills. Yeah it definitely helps

00:10:22.259 --> 00:10:25.940
and you know I always have learned that you know

00:10:25.940 --> 00:10:29.580
marketing and data are joined at the hip because

00:10:29.580 --> 00:10:32.779
you don't you can't work one without the other

00:10:32.779 --> 00:10:38.809
and you know It's always good when, for me, it's

00:10:38.809 --> 00:10:41.210
always good to partner with a person who has

00:10:41.210 --> 00:10:43.070
had some data background on the marketing side,

00:10:43.070 --> 00:10:50.549
so it makes my job easier. For sure. Yeah, you're

00:10:50.549 --> 00:10:54.129
a thought leader, like I said, at the intersection

00:10:54.129 --> 00:10:57.860
of digital marketing and data. And I know you

00:10:57.860 --> 00:11:01.139
think a lot about how companies and clients should

00:11:01.139 --> 00:11:04.320
approach data. Can you please share your thoughts

00:11:04.320 --> 00:11:08.460
on it and give us some perspective? Yeah, I mean,

00:11:08.879 --> 00:11:11.159
sorry, go ahead. Was there anything else? OK.

00:11:12.120 --> 00:11:15.519
Yeah. So this is alluding a little bit to something

00:11:15.519 --> 00:11:19.039
that's going to happen later this year. But I

00:11:19.039 --> 00:11:23.480
do have a unique philosophy. And I think as time

00:11:23.480 --> 00:11:25.080
has gone on, I've gotten a little more specific

00:11:25.080 --> 00:11:28.940
about it. And what it is, is I think that I want

00:11:28.940 --> 00:11:33.100
people to feel good about data. And so what that

00:11:33.100 --> 00:11:37.600
ends up meaning is, as you look in our kind of

00:11:37.600 --> 00:11:41.419
like data flooded world right now, even though

00:11:41.639 --> 00:11:43.960
There's a ton of data being generated by pretty

00:11:43.960 --> 00:11:46.500
much every company, just overwhelming amounts.

00:11:47.539 --> 00:11:49.600
Not everyone feels super great about it. We don't

00:11:49.600 --> 00:11:52.419
really use it to its fullest potential. And there's

00:11:52.419 --> 00:11:56.679
a lot of waste or spillage of data where we spend

00:11:56.679 --> 00:11:59.259
a lot of time collecting it. We're storing it

00:11:59.259 --> 00:12:01.460
and everything, and we're not using it. And I

00:12:01.460 --> 00:12:04.820
think what that really comes down to is that

00:12:04.820 --> 00:12:08.299
there's like these other factors. Data is the

00:12:08.299 --> 00:12:11.220
tool and what we're running into is there's a

00:12:11.220 --> 00:12:14.559
lack of trust in data and organization and people

00:12:14.559 --> 00:12:16.120
don't feel good about it when someone's sharing

00:12:16.120 --> 00:12:18.039
something and they don't feel like it represents

00:12:18.039 --> 00:12:22.120
the company and things like that. what my approach

00:12:22.120 --> 00:12:25.080
is in digital marketing or marketing in general,

00:12:25.299 --> 00:12:27.960
where that really can create a conflict because

00:12:27.960 --> 00:12:31.960
a lot of times marketing is about feelings. Data

00:12:31.960 --> 00:12:37.080
comes in and a traditional analyst can be so

00:12:37.080 --> 00:12:39.580
exacting in how they want something, and those

00:12:39.580 --> 00:12:43.820
two can sometimes cause some conflict. My approach

00:12:43.820 --> 00:12:48.200
is really like, working to make the data quality

00:12:48.200 --> 00:12:50.840
as good as possible, being really pragmatic about

00:12:50.840 --> 00:12:52.940
what you're collecting and if it makes sense

00:12:52.940 --> 00:12:55.139
to collect that. And then on the other hand,

00:12:55.200 --> 00:12:58.600
focus a lot on using what you have and using

00:12:58.600 --> 00:13:01.679
the most useful data. And what really you need

00:13:01.679 --> 00:13:03.700
to do first to do that is to make sure you have

00:13:03.700 --> 00:13:08.460
a culture that's ready to receive data. Yeah,

00:13:08.460 --> 00:13:12.559
that's a great point. And you really have to,

00:13:12.559 --> 00:13:18.690
there's a lot of... intricacies when it comes

00:13:18.690 --> 00:13:25.669
to data. Well, what does having data mean? Especially

00:13:25.669 --> 00:13:29.309
in the digital world, data is never accurate

00:13:29.309 --> 00:13:33.909
and it will be the accuracy or you won't be able

00:13:33.909 --> 00:13:37.169
to collect everything that you want. You go more

00:13:37.169 --> 00:13:41.090
so going forward because of privacy laws and

00:13:41.090 --> 00:13:43.830
everything and it should be that way. because

00:13:43.830 --> 00:13:46.809
more power to the consumers, right? But being

00:13:46.809 --> 00:13:50.710
on the data side, I'm like, no, I wish I had

00:13:50.710 --> 00:13:54.789
all the data. But yeah, I think you passed on

00:13:54.789 --> 00:14:03.129
a point. In terms of our goal, if our goal is

00:14:03.129 --> 00:14:06.169
to get to intelligence, there are precursors

00:14:06.169 --> 00:14:09.450
to it, which is you've got to have the right

00:14:09.450 --> 00:14:12.509
infrastructure to collect the data and then...

00:14:12.460 --> 00:14:15.679
you have to have adoption by the customers or

00:14:15.679 --> 00:14:21.500
the users. And then you have to get the feedback

00:14:21.500 --> 00:14:24.120
from them and then start developing intelligence.

00:14:24.620 --> 00:14:29.679
So there are a few steps to getting to the intelligence

00:14:29.679 --> 00:14:34.000
or, you know, be it testing and marketing, or

00:14:34.000 --> 00:14:38.120
maybe in my world, be developing a machine learning

00:14:38.120 --> 00:14:41.190
model that needs to be adopted and then provided

00:14:41.190 --> 00:14:44.690
some feedback, only then we can tune it to get

00:14:44.690 --> 00:14:50.649
to some... Although it sounds like some people

00:14:50.649 --> 00:14:53.250
think of it like magic, it's just iterative process

00:14:53.250 --> 00:15:01.090
of getting to a good phase. Yeah, you made some

00:15:01.090 --> 00:15:05.450
great points there. Yeah, I followed you on LinkedIn

00:15:05.450 --> 00:15:11.700
and I know... GA4 is big on your mind. And can

00:15:11.700 --> 00:15:15.200
you please tell our audience what is Google Analytics

00:15:15.200 --> 00:15:19.600
and how companies leverage it to go ahead? Yeah,

00:15:19.700 --> 00:15:21.980
that's a great one. I think sometimes we skip

00:15:21.980 --> 00:15:23.860
over explaining something like that and just

00:15:23.860 --> 00:15:26.620
assume everybody knows. But Google Analytics

00:15:26.620 --> 00:15:30.830
is the most popular. platform for tracking behavioral

00:15:30.830 --> 00:15:34.570
data and marketing data for websites and apps.

00:15:35.409 --> 00:15:38.429
It's one of the only free solutions. I'm putting

00:15:38.429 --> 00:15:40.669
a little asterisk by that because there's some

00:15:40.669 --> 00:15:44.309
things changing. What it does is you put a tag

00:15:44.309 --> 00:15:46.600
on your website. And by tag, it's a little bit

00:15:46.600 --> 00:15:48.639
of code that goes into the header. And what it

00:15:48.639 --> 00:15:51.200
does is start to monitor what people do on your

00:15:51.200 --> 00:15:53.620
site. So when they hit the site, where do they

00:15:53.620 --> 00:15:56.779
go afterwards? Where did they come from? And

00:15:56.779 --> 00:15:58.740
gets that general intelligence. And it's really

00:15:58.740 --> 00:16:01.679
easy to implement, too. So almost every site

00:16:01.679 --> 00:16:04.100
can implement Google Analytics and get something

00:16:04.100 --> 00:16:08.519
out of that data. So that's it. my super simplified

00:16:08.519 --> 00:16:11.100
version of what it does. Obviously, there's a

00:16:11.100 --> 00:16:12.720
little bit more that goes into it for bigger

00:16:12.720 --> 00:16:16.700
companies, but that's the general premise. That's

00:16:16.700 --> 00:16:20.659
awesome. Yeah. It's super impressive the way

00:16:20.659 --> 00:16:24.399
that those little tags can tell us so much about

00:16:24.399 --> 00:16:27.759
our customers, especially on the behavioral side,

00:16:28.259 --> 00:16:30.980
where they come in, the final metrics and all

00:16:30.980 --> 00:16:36.059
that. But overarching, overarchingly in terms

00:16:36.059 --> 00:16:40.220
of types of data. I think of it in terms of geographical

00:16:40.220 --> 00:16:43.600
data when it comes to customers and then demographics

00:16:43.600 --> 00:16:47.360
a little more nuanced and also psychography,

00:16:47.820 --> 00:16:51.779
which is behavioral. So Google Analytics is super

00:16:51.779 --> 00:16:54.820
helpful for us to understand the behavior of

00:16:54.820 --> 00:16:59.960
our customers within our apps or websites and

00:16:59.960 --> 00:17:04.140
stuff like that. So thanks for helping us understand

00:17:04.140 --> 00:17:08.980
that. I know you use a couple more tools. Are

00:17:08.980 --> 00:17:11.579
there any alternatives to Google Analytics that

00:17:11.579 --> 00:17:13.799
you can tell us about? Yeah, there certainly

00:17:13.799 --> 00:17:17.720
are. One of my personal favorites is Adobe Analytics.

00:17:18.059 --> 00:17:20.480
That's on the enterprise level for sure, but

00:17:20.480 --> 00:17:22.559
that's kind of, I would say, kind of the two

00:17:22.559 --> 00:17:26.779
big. players in the market. And then of course

00:17:26.779 --> 00:17:29.740
there's like a whole suite of more specialized

00:17:29.740 --> 00:17:33.119
what I call best of breed tools. So this is something

00:17:33.119 --> 00:17:34.819
that's once again changing a little bit right

00:17:34.819 --> 00:17:37.339
now, but previously there's been tools that are

00:17:37.339 --> 00:17:39.500
a little more focused towards apps. If you're

00:17:39.500 --> 00:17:41.940
a company that's more focused on having an app

00:17:41.940 --> 00:17:44.279
and you know maybe your website isn't your primary

00:17:44.279 --> 00:17:47.400
focus, then there's tools like Mixpanel and Amplitude

00:17:47.400 --> 00:17:51.680
that are also good options. And now with the

00:17:51.690 --> 00:17:53.990
They've been around for a little longer than

00:17:53.990 --> 00:17:56.349
this, but now with the switch of GA4, there's

00:17:56.349 --> 00:17:58.069
definitely a lot of pain points around that.

00:17:58.089 --> 00:18:00.730
There's also some other pain points about Google

00:18:00.730 --> 00:18:03.789
within Europe. There's some other options like

00:18:03.789 --> 00:18:06.170
Piwik Pro. I'm not sure if I'm pronouncing it

00:18:06.170 --> 00:18:07.569
right. I think I've never heard it pronounced

00:18:07.569 --> 00:18:10.849
out loud. And there's some other few free solutions

00:18:10.849 --> 00:18:12.750
that are also starting to become a little bit

00:18:12.750 --> 00:18:14.990
more prominent. So there are a lot of options.

00:18:15.069 --> 00:18:18.900
And then there's also... some more custom solutions.

00:18:19.440 --> 00:18:21.779
So if you were a really advanced company and

00:18:21.779 --> 00:18:23.680
you wanted something that was completely your

00:18:23.680 --> 00:18:27.440
own, you could be using something like Snowplow

00:18:27.440 --> 00:18:30.680
and Snowflake together. So there's also some

00:18:30.680 --> 00:18:33.480
solutions like that as well. So it's definitely

00:18:33.480 --> 00:18:36.339
a complicated landscape, but I forget the exact

00:18:36.339 --> 00:18:37.799
number, so I don't want to quote it exactly.

00:18:38.440 --> 00:18:41.640
But there's a vast majority of sites, especially

00:18:41.640 --> 00:18:44.640
small to medium businesses, are on Google Analytics.

00:18:45.150 --> 00:18:48.690
That's awesome. Yeah, Google Analytics is definitely

00:18:48.690 --> 00:18:53.970
the biggest player in the market. I know there's

00:18:53.970 --> 00:18:57.970
a big shift that Google Analytics is going through.

00:18:58.410 --> 00:19:01.450
Can you talk about that and what was it originally

00:19:01.450 --> 00:19:04.769
called and what did that do and what's the next

00:19:04.769 --> 00:19:08.849
phase? Yeah, so the current version of Google

00:19:08.849 --> 00:19:11.089
Analytics that's been around for a few years

00:19:11.089 --> 00:19:15.900
now is called Universal Analytics. was announced,

00:19:16.279 --> 00:19:17.859
I forget exactly when it was, it feels like it

00:19:17.859 --> 00:19:19.839
was a really long time ago, I've been slogging

00:19:19.839 --> 00:19:22.240
through these migrations, is Google Analytics

00:19:22.240 --> 00:19:27.140
4. And that transformation, and this is colored

00:19:27.140 --> 00:19:29.220
by my opinion completely, but someone who's more

00:19:29.220 --> 00:19:32.259
familiar with also some of those other analytics

00:19:32.259 --> 00:19:35.700
tools, is that Google is catching on to moving

00:19:35.700 --> 00:19:38.980
towards event -based tracking. So prior, they

00:19:38.980 --> 00:19:41.940
were session -based, and really what the difference

00:19:41.940 --> 00:19:44.880
is there is like, event based tracking is stronger

00:19:44.880 --> 00:19:48.079
for combining multiple properties. So if you

00:19:48.079 --> 00:19:51.579
have a website and an app, it was really difficult

00:19:51.579 --> 00:19:53.119
with Google Analytics. And I thought that they

00:19:53.119 --> 00:19:56.559
were really weak in it was measuring app analytics.

00:19:56.900 --> 00:20:00.440
They just weren't really set up to be event focused,

00:20:00.440 --> 00:20:03.759
which is what apps are. And so I think that's

00:20:03.759 --> 00:20:06.039
one of their major moves. I think there's a lot

00:20:06.039 --> 00:20:08.460
of other pressures of why Google Analytics is

00:20:08.460 --> 00:20:10.900
changing. And one little like stepping back a

00:20:10.900 --> 00:20:14.579
bit is that the way this new version is working

00:20:14.579 --> 00:20:16.900
compared to some of the other version updates

00:20:16.900 --> 00:20:19.940
that Google has done in the past, you have to

00:20:19.940 --> 00:20:23.019
completely re -tag your implementation and switch

00:20:23.019 --> 00:20:25.119
over. So it's really a tool migration, which

00:20:25.119 --> 00:20:28.220
is brutal because any time that you do an analytics

00:20:28.220 --> 00:20:30.539
tool migration or any other technical tool migration,

00:20:31.339 --> 00:20:34.230
that's a big... a strategic undertaking and like

00:20:34.230 --> 00:20:37.369
a big cost undertaking for companies. So Google

00:20:37.369 --> 00:20:39.450
announced that they're going to do this and that

00:20:39.450 --> 00:20:41.849
every company is expected to go on this journey.

00:20:42.329 --> 00:20:44.710
And whether they were ready or not to go on to

00:20:44.710 --> 00:20:47.029
do that sort of strategic project, they're going

00:20:47.029 --> 00:20:50.569
to have to whether they choose to go to GA4 or

00:20:50.569 --> 00:20:53.470
whether they choose to move on to another analytics

00:20:53.470 --> 00:20:57.349
tool. And so We're still hearing bits and pieces

00:20:57.349 --> 00:20:59.769
about what their final vision is. I honestly

00:20:59.769 --> 00:21:01.769
couldn't speak to it. It would all be conjecture.

00:21:02.269 --> 00:21:04.730
But I know it's partially because of some of

00:21:04.730 --> 00:21:06.809
the privacy pressures that they're switching

00:21:06.809 --> 00:21:10.069
the way their tool functions. Personally, like

00:21:10.069 --> 00:21:12.329
I said, I think it's because they're trying to

00:21:12.329 --> 00:21:15.549
move to a better interface. They've really operated

00:21:15.549 --> 00:21:17.730
off of these static reports that are very marketer

00:21:17.730 --> 00:21:20.369
friendly, but their competitors like Adobe Analytics

00:21:20.369 --> 00:21:23.670
have had a lot more flexibility for some time

00:21:23.670 --> 00:21:27.250
of making visualizations, catering things, having

00:21:27.250 --> 00:21:30.410
it be a lot more in the flow analysis. So they're

00:21:30.410 --> 00:21:32.190
definitely moving towards that and they're trying

00:21:32.190 --> 00:21:35.769
to apply some more intelligence to that. That's

00:21:35.769 --> 00:21:37.349
another thing that they've been a little slow

00:21:37.349 --> 00:21:41.259
on the uptick. for compared to Adobe is anomaly

00:21:41.259 --> 00:21:44.539
detection and things like that. So I know that

00:21:44.539 --> 00:21:47.220
was a really broad explanation, but we're going

00:21:47.220 --> 00:21:51.440
through a strategic technical migration, everybody

00:21:51.440 --> 00:21:54.740
who was on Google Analytics, to a tool that's

00:21:54.740 --> 00:21:57.440
going to function pretty differently. Yeah. I

00:21:57.440 --> 00:22:01.819
mean, it's definitely a seismic shift for companies.

00:22:02.900 --> 00:22:07.160
And a lot of companies are depending so heavily

00:22:07.160 --> 00:22:10.759
on Google Analytics that they have to do it.

00:22:10.900 --> 00:22:16.500
And I think if anybody is interested, July 1st

00:22:16.500 --> 00:22:19.279
is the deadline? Yeah, July 1st is the deadline

00:22:19.279 --> 00:22:22.539
for any non -Google Analytics 360, so the premium

00:22:22.539 --> 00:22:25.720
version. And then I believe it's October 1st

00:22:25.720 --> 00:22:28.619
for the Google Analytics 360 customers. So if

00:22:28.619 --> 00:22:30.720
you're paying for the premium version, you have

00:22:30.720 --> 00:22:34.200
a little bit more time. I don't know if it's

00:22:34.200 --> 00:22:37.140
really a fair... a bit more time because anyone

00:22:37.140 --> 00:22:39.220
who's on 360, their implementation is pretty

00:22:39.220 --> 00:22:41.339
complex. So I feel like they could use a few

00:22:41.339 --> 00:22:45.680
more months than just that, but yes. Yeah. And

00:22:45.680 --> 00:22:50.299
also the cost difference is huge between free

00:22:50.299 --> 00:22:58.779
and 360. We looked into it and then it was definitely

00:22:58.779 --> 00:23:06.210
a lot more expensive. Yeah, just takeaways for

00:23:06.210 --> 00:23:12.009
myself as well. GA4 is focused on more about

00:23:12.009 --> 00:23:15.309
app analytics and also event -based analytics,

00:23:15.329 --> 00:23:19.170
and it gives us a lot of fun metrics, and you

00:23:19.170 --> 00:23:23.910
can customize the way that you could track different

00:23:23.910 --> 00:23:27.890
buttons, scrolls, different objects within your

00:23:27.890 --> 00:23:37.259
web pages or app pages. very marketing and data

00:23:37.259 --> 00:23:42.460
geeks dreamland, in my opinion. Yeah, and definitely

00:23:42.460 --> 00:23:47.480
a little bit, the more business users and not

00:23:47.480 --> 00:23:51.539
as data -focused marketers are going to face

00:23:51.539 --> 00:23:53.299
a reckoning, that's for sure. It's going to be

00:23:53.299 --> 00:23:56.140
a very different way of getting information out

00:23:56.140 --> 00:23:59.359
of the tool. Yeah, and one of the things that

00:23:59.359 --> 00:24:06.630
I've been preparing my executives and my teams

00:24:06.630 --> 00:24:13.390
for is that your historical data, you won't be

00:24:13.390 --> 00:24:15.769
able to just continue on with it. There will

00:24:15.769 --> 00:24:19.470
be some differences because the way that these

00:24:19.470 --> 00:24:23.210
two versions are looking at the data is entirely

00:24:23.210 --> 00:24:26.069
different. Like we said, session -based versus

00:24:26.069 --> 00:24:29.890
event -based. So if you've had a thousand sessions

00:24:29.890 --> 00:24:35.039
in the past, a month, you could see 1 ,500 or

00:24:35.039 --> 00:24:42.779
800. It really depends. Yeah. I've been seeing

00:24:42.779 --> 00:24:45.119
that already and hopefully I have been because,

00:24:45.180 --> 00:24:47.059
by the way, if you're listening to this and you

00:24:47.059 --> 00:24:48.839
haven't migrated yet, you're in trouble because

00:24:48.839 --> 00:24:51.460
you only have a few months and you're probably

00:24:51.460 --> 00:24:53.420
not going to finish the migration in time. I

00:24:53.420 --> 00:24:56.720
still think you should start now. But yeah, there's

00:24:56.720 --> 00:24:59.519
going to be discrepancies and also just differences

00:24:59.519 --> 00:25:03.630
that aren't really like an issue, but I'm also

00:25:03.630 --> 00:25:07.849
seeing that like When we start these implementations,

00:25:07.990 --> 00:25:09.869
sometimes the tracking doesn't really do what

00:25:09.869 --> 00:25:11.930
we expected. And there's a lot of pieces like

00:25:11.930 --> 00:25:14.849
that. So have been going through that with clients

00:25:14.849 --> 00:25:18.750
right now where we think that we got parity and

00:25:18.750 --> 00:25:20.670
then we're looking, okay, is this a difference

00:25:20.670 --> 00:25:22.650
in how they track it? Or is this that something

00:25:22.650 --> 00:25:25.190
is not going right with the tagging? And that's

00:25:25.190 --> 00:25:27.569
a lot of work to go in and make sure that those

00:25:27.569 --> 00:25:30.410
are correct. And right now, ideally you started

00:25:30.410 --> 00:25:33.569
your migration a year ago or something like that,

00:25:33.569 --> 00:25:35.650
especially if you're a big company. And you've

00:25:35.650 --> 00:25:38.130
been spending the last few months and will be

00:25:38.130 --> 00:25:40.450
spending the next few months just continuing

00:25:40.450 --> 00:25:43.529
to take inventory of that. Yeah, absolutely.

00:25:43.890 --> 00:25:47.150
And I think running both these versions in parallel

00:25:47.150 --> 00:25:50.049
and making sure that you're comfortable with

00:25:50.049 --> 00:25:53.069
the Delta. And then you could probably, once

00:25:53.069 --> 00:25:57.690
GAUA goes away, you could kind of extrapolate.

00:25:58.799 --> 00:26:02.759
your data backwards and get year over year comparisons

00:26:02.759 --> 00:26:08.299
just to keep your sanity. It's a lot of work

00:26:08.299 --> 00:26:11.619
and yeah, if you've not started your migration,

00:26:12.140 --> 00:26:14.079
definitely reach out to Mary Beth. She can help

00:26:14.079 --> 00:26:16.859
you out. Yeah, we have a few slots left. I'll

00:26:16.859 --> 00:26:18.660
say that that's the other warning that I have.

00:26:21.399 --> 00:26:23.980
that all of the agencies that do this are going

00:26:23.980 --> 00:26:26.680
to be really heavy hit. We do have some, as of

00:26:26.680 --> 00:26:28.819
recording this on May 3rd, I don't know when

00:26:28.819 --> 00:26:31.420
it's going to get published, we do have some

00:26:31.420 --> 00:26:33.200
slots available and we're always happy to help

00:26:33.200 --> 00:26:36.099
and we won't shame you too much. But it's definitely

00:26:36.099 --> 00:26:37.700
something where like, if you don't have the resources

00:26:37.700 --> 00:26:40.279
in house, it's like, we're almost getting to

00:26:40.279 --> 00:26:41.900
the point where we won't bother dual tagging.

00:26:42.039 --> 00:26:44.420
But what we do, that's I think it's an important

00:26:44.420 --> 00:26:46.880
call out, is almost no one on Google Analytics

00:26:46.880 --> 00:26:50.839
has like a data dictionary, right? Operating

00:26:50.839 --> 00:26:53.160
a data dictionary is a list of all the tags that

00:26:53.160 --> 00:26:56.140
you have, all the properties, what they do. And

00:26:56.140 --> 00:26:58.259
so what we're finding we always have to do is

00:26:58.259 --> 00:27:00.700
go in and create that data dictionary first.

00:27:00.720 --> 00:27:03.359
So even if you don't dual tag, because otherwise

00:27:03.359 --> 00:27:05.700
we don't really know what tags we need to make

00:27:05.700 --> 00:27:09.400
in Google Analytics 4. So that's something that

00:27:09.400 --> 00:27:11.519
is just something that you're gonna have to do.

00:27:11.519 --> 00:27:13.180
And it's a great way to work once you have that

00:27:13.180 --> 00:27:16.640
put together too. Yeah, that's a great point.

00:27:16.960 --> 00:27:20.130
And also there's some differences like We don't

00:27:20.130 --> 00:27:26.150
have segments anymore in GA4. I use segments

00:27:26.150 --> 00:27:28.670
all the time to kind of just slap that in there.

00:27:28.769 --> 00:27:32.490
And then I've got all the filters ready. So can

00:27:32.490 --> 00:27:36.269
you explain to us what segments were and what

00:27:36.269 --> 00:27:38.990
are we losing? Yeah. And I mean, I think there's

00:27:38.990 --> 00:27:40.630
like some slight ways that there are some ways

00:27:40.630 --> 00:27:43.000
that we're doing it, but it's not as. Nice in

00:27:43.000 --> 00:27:45.619
some ways. So segments and segmentation is when

00:27:45.619 --> 00:27:48.599
you create a subset of the data to do and primarily

00:27:48.599 --> 00:27:51.799
the way I use it is to do comparisons. So in

00:27:51.799 --> 00:27:55.119
Google Analytics Universal, you could make one

00:27:55.119 --> 00:27:58.319
of my favorites is a purchaser versus someone

00:27:58.319 --> 00:28:01.440
who abandoned cart or someone who's a non purchaser.

00:28:01.720 --> 00:28:04.059
And what you were able to do is that Google Analytics

00:28:04.059 --> 00:28:06.859
had been a lot of static reports where you could

00:28:06.859 --> 00:28:09.819
then look at those two segments or those two

00:28:09.819 --> 00:28:13.269
groups of behavioral you know the behaviors or

00:28:13.269 --> 00:28:16.970
people etc and compare them and say okay well

00:28:16.970 --> 00:28:20.130
with my people who didn't make a purchase I can

00:28:20.130 --> 00:28:22.109
see that the bounce rate is a lot higher and

00:28:22.109 --> 00:28:24.069
they came from these different sources so what

00:28:24.069 --> 00:28:25.569
are we doing with those sources like what do

00:28:25.569 --> 00:28:29.849
we need to improve and so Now, they're changing,

00:28:30.069 --> 00:28:32.150
and it's not quite as easy to do that side -by

00:28:32.150 --> 00:28:35.650
-side comparison. You can add and filter, et

00:28:35.650 --> 00:28:38.569
cetera, but at least right now, we'll see how

00:28:38.569 --> 00:28:40.369
they end up changing it, because every day I

00:28:40.369 --> 00:28:42.390
log into Google Analytics 4, I swear that they've

00:28:42.390 --> 00:28:44.930
changed something else. But it's definitely going

00:28:44.930 --> 00:28:46.910
to be a slightly different way of looking at

00:28:46.910 --> 00:28:49.390
it, and I'm curious what it's going to look like.

00:28:49.670 --> 00:28:51.509
Like I said, I'm familiar with Adobe Analytics,

00:28:51.529 --> 00:28:53.880
and they've... they have a similar interface

00:28:53.880 --> 00:28:56.779
and they found a way to do that. But, you know,

00:28:56.880 --> 00:28:59.200
it's going to be really interesting to see how

00:28:59.200 --> 00:29:01.779
they address that because that's a core function

00:29:01.779 --> 00:29:06.119
of Google Analytics. Yeah, for sure. And also

00:29:06.119 --> 00:29:12.079
in our company, the way that we are doing kind

00:29:12.079 --> 00:29:16.259
of overcoming some of those is we're actually

00:29:16.259 --> 00:29:21.400
using their API through Google BigQuery to, you

00:29:21.400 --> 00:29:24.859
know, export all the data into our warehouse,

00:29:24.920 --> 00:29:28.559
Snowflake, or any other database, and then you

00:29:28.559 --> 00:29:34.279
could build your own segmentation and reporting

00:29:34.279 --> 00:29:40.299
outside of GA's interface. One of the advantages

00:29:40.299 --> 00:29:43.619
to doing that is you can match your GA data with

00:29:43.619 --> 00:29:47.480
your internal data or third -party data, so it

00:29:47.480 --> 00:29:49.579
gives you a lot more flexibility, but then it's

00:29:49.579 --> 00:29:53.099
a lot more investment. infrastructure and all

00:29:53.099 --> 00:29:57.079
that. But depending on what phase the company

00:29:57.079 --> 00:30:03.200
is in, there are multiple options. It can be

00:30:03.200 --> 00:30:06.279
a lot, but it's super interesting for anybody

00:30:06.279 --> 00:30:08.759
that likes data and marketing and everything.

00:30:10.180 --> 00:30:12.920
Yeah, certainly. And I think that's another good

00:30:12.920 --> 00:30:16.480
call out, really, about GA4, is that at least

00:30:16.480 --> 00:30:18.640
the way that they're currently explaining it'll

00:30:18.640 --> 00:30:22.940
work, you do kind of need to use BigQuery. And

00:30:22.940 --> 00:30:25.160
what that means is, yes, there's a lot of infrastructure

00:30:25.160 --> 00:30:26.700
and everything, and then you need somebody who

00:30:26.700 --> 00:30:29.779
knows SQL, and BigQuery's weird version of SQL,

00:30:29.880 --> 00:30:33.920
too. So that's a whole other thing, where some

00:30:33.920 --> 00:30:36.180
companies may have been relying on analysts that

00:30:36.180 --> 00:30:40.700
come from a less like And this is not a bad thing,

00:30:40.880 --> 00:30:43.019
by the way, like a less technical background

00:30:43.019 --> 00:30:45.440
where they're really good contextual analysts

00:30:45.440 --> 00:30:48.079
and business analysts, but they didn't need to

00:30:48.079 --> 00:30:50.900
use SQL before. So now we're going into this

00:30:50.900 --> 00:30:54.160
place where to get the full usage out of your

00:30:54.160 --> 00:30:56.519
data that you were before with Google Analytics,

00:30:56.779 --> 00:30:58.740
you kind of need to consider whether you're going

00:30:58.740 --> 00:31:02.740
to upscale somebody, bring new talent in, et

00:31:02.740 --> 00:31:06.400
cetera. Yeah, it's definitely a lot to consider.

00:31:06.809 --> 00:31:11.710
and it's thrown on companies and they never asked

00:31:11.710 --> 00:31:16.170
for it, but here we are. Yeah, we'll see. I think

00:31:16.170 --> 00:31:19.549
it's a pretty interesting play, but I can't really

00:31:19.549 --> 00:31:22.710
comment on whatever the master plan is for Google,

00:31:22.789 --> 00:31:25.190
so I'm always interested to see what their next

00:31:25.190 --> 00:31:29.509
move will be. Yeah. I mean, it's a great tool

00:31:29.509 --> 00:31:32.109
that they offer for free, so I can't complain

00:31:32.109 --> 00:31:37.519
too much, but definitely something that Yeah,

00:31:37.619 --> 00:31:42.220
July 1st deadline, so make sure you've got your

00:31:42.220 --> 00:31:51.640
stuff in place. Awesome. So I know there are

00:31:51.640 --> 00:31:56.380
a few things when it comes to metrics and optimization.

00:31:56.460 --> 00:31:59.359
In the digital marketing space, A B testing and

00:31:59.359 --> 00:32:04.740
optimizations are super important. One of the

00:32:04.740 --> 00:32:09.220
things that We talked about initially, once you

00:32:09.220 --> 00:32:11.759
have your infrastructure, you need to have adoption.

00:32:13.500 --> 00:32:16.440
One of the things that help with adoption and

00:32:16.440 --> 00:32:20.259
also with the feedback is how you measure your

00:32:20.259 --> 00:32:23.160
tests or what you're doing. So what are some

00:32:23.160 --> 00:32:26.880
of your thoughts on measuring when it comes to

00:32:26.880 --> 00:32:30.099
measuring the impact of your A -B tests or optimizations

00:32:30.099 --> 00:32:34.160
that you run? Oh, this is such a juicy topic.

00:32:35.990 --> 00:32:38.690
So there's, and this is a really fun one too,

00:32:38.849 --> 00:32:41.829
because there's been, I would say, a divergence

00:32:41.829 --> 00:32:44.849
in the A -B testing world. And there's kind of,

00:32:44.849 --> 00:32:46.990
there's growth marketers and then there's more

00:32:46.990 --> 00:32:49.410
statistics focused people. And my philosophy

00:32:49.410 --> 00:32:53.660
is that. We need to try and make tests as scientifically

00:32:53.660 --> 00:32:56.420
sound as possible. Marketing is already a whole

00:32:56.420 --> 00:32:58.420
mess, and our data is already a whole mess, and

00:32:58.420 --> 00:33:00.779
it's not as accurate. And you have to accept

00:33:00.779 --> 00:33:02.720
a little bit that we're never going to get as

00:33:02.720 --> 00:33:07.720
accurate as a hard science. But we need to be

00:33:07.720 --> 00:33:11.079
unrelenting in keeping those elements that make

00:33:11.079 --> 00:33:13.980
things more sound. So what a few of those are

00:33:13.980 --> 00:33:19.259
is I think that any test should be pre -tested

00:33:19.259 --> 00:33:22.599
so that you can calculate how long it should

00:33:22.599 --> 00:33:24.859
run for for you to be able to see an uplift and

00:33:24.859 --> 00:33:26.880
whether you even have enough traffic to test

00:33:26.880 --> 00:33:30.299
that particular thing. So sometimes, and this

00:33:30.299 --> 00:33:32.019
is especially for like small to medium businesses,

00:33:32.259 --> 00:33:34.789
you want to... test on your conversion rate,

00:33:34.809 --> 00:33:37.069
for example, that's always, you know, get this

00:33:37.069 --> 00:33:39.950
term conversion rate optimization. But if your

00:33:39.950 --> 00:33:42.190
conversion rate is only like one or two percent

00:33:42.190 --> 00:33:45.089
and you don't get that much traffic, it's literally

00:33:45.089 --> 00:33:47.549
impossible for you to see a true statistical

00:33:47.549 --> 00:33:50.309
uplift during that time. So I think being really

00:33:50.309 --> 00:33:53.819
honest and strict with that is like element number

00:33:53.819 --> 00:33:56.519
one. And element number two is to never let a

00:33:56.519 --> 00:33:58.819
marketer go in and look at that test before it's

00:33:58.819 --> 00:34:00.640
done running and tell you whether it's winning

00:34:00.640 --> 00:34:03.339
or not, because you need to wait. So if you pre

00:34:03.339 --> 00:34:06.380
-test and it says it'll take seven days for statistical

00:34:06.380 --> 00:34:08.659
significance, then you wait seven days, then

00:34:08.659 --> 00:34:11.619
you look at the results on day seven. And if

00:34:11.619 --> 00:34:13.320
you look at it and it says there's no difference,

00:34:13.860 --> 00:34:17.679
then you know that you have rejected your hypothesis.

00:34:18.139 --> 00:34:20.619
it doesn't make a difference. A lot of times

00:34:20.619 --> 00:34:22.920
what we do instead is say, oh, let's just let

00:34:22.920 --> 00:34:26.780
it run a little longer. So that's one core area

00:34:26.780 --> 00:34:31.179
of it. And then I also think being really cautious

00:34:31.179 --> 00:34:33.440
about those thresholds, too, as much as possible

00:34:33.440 --> 00:34:36.460
of what do you need to see as an uplift? Because

00:34:36.460 --> 00:34:40.760
that's completely another thing. If you see a

00:34:40.760 --> 00:34:43.840
5 % uplift in conversion rate, That can make

00:34:43.840 --> 00:34:45.900
a really significant difference for your business.

00:34:46.579 --> 00:34:49.519
And it makes it, in some ways, a little bit easier

00:34:49.519 --> 00:34:50.920
because you're like, OK, well, I only need to

00:34:50.920 --> 00:34:53.360
see a 5 % uplift. But for some other metrics,

00:34:53.760 --> 00:34:57.599
a 5 % uplift might be negligible. So you might

00:34:57.599 --> 00:34:59.599
be wanting to see something more to consider

00:34:59.599 --> 00:35:02.539
it a winning test. So those are two elements.

00:35:03.059 --> 00:35:05.179
I think the other part that gets a little bit

00:35:05.179 --> 00:35:07.199
more sticky, and I've seen it done a ton of different

00:35:07.199 --> 00:35:11.469
ways, is how you calculate out the diminishing

00:35:11.469 --> 00:35:14.670
returns of operationalizing a test. I know that

00:35:14.670 --> 00:35:16.769
sounded like a lot of really big complicated

00:35:16.769 --> 00:35:20.289
words, but what that means is, you run the test

00:35:20.289 --> 00:35:23.929
on your website, it wins, you maybe even do some

00:35:23.929 --> 00:35:25.550
other confirmations and you're really clear.

00:35:25.730 --> 00:35:29.110
Whatever feature we added is a value add to the

00:35:29.110 --> 00:35:32.829
site. Then you have to go through development

00:35:32.829 --> 00:35:36.429
and get that put into place. And then you need

00:35:36.429 --> 00:35:38.610
to evaluate, well, how much impact did that really

00:35:38.610 --> 00:35:43.369
make? And it's such a fun thing because it's

00:35:43.369 --> 00:35:46.769
not a perfectly controlled test anymore. And

00:35:46.769 --> 00:35:48.309
like I said, I've seen it done a lot of different

00:35:48.309 --> 00:35:50.309
ways. I've seen people play around with modeling,

00:35:50.630 --> 00:35:53.210
like a diminishing return so that they'll get

00:35:53.210 --> 00:35:55.989
that perceived uplift over a few months and then

00:35:55.989 --> 00:35:58.809
it starts to trickle down. So there's been some

00:35:58.809 --> 00:36:00.510
kind of advanced models like that. Some people

00:36:00.510 --> 00:36:04.250
just make a guess of like a flat number. So that's

00:36:04.250 --> 00:36:06.230
where like very interested to hear like your

00:36:06.230 --> 00:36:08.150
listeners, if they have any opinions of how they've

00:36:08.150 --> 00:36:10.070
seen it done that they really like it. But I

00:36:10.070 --> 00:36:12.550
think those are some of the core elements that

00:36:12.550 --> 00:36:16.269
need to go into testing that make a big difference

00:36:16.269 --> 00:36:19.329
in the quality of what you're running. That's

00:36:19.329 --> 00:36:22.570
yeah. There's a lot to unpack in your answer.

00:36:22.869 --> 00:36:29.380
There's a lot of super useful nuggets. If I made

00:36:29.380 --> 00:36:34.119
some notes, as you were talking, be scientific

00:36:34.119 --> 00:36:38.099
in your approach. Test needs to be pre -tested.

00:36:38.599 --> 00:36:41.420
So you define what your tests need to be and

00:36:41.420 --> 00:36:45.239
then get a gut check for it. And you test on

00:36:45.239 --> 00:36:49.159
volume and having some statistically significant

00:36:49.159 --> 00:36:55.539
volume. If you don't have decent enough volume,

00:36:55.820 --> 00:36:59.800
there's no way you could run a scientific test

00:36:59.800 --> 00:37:04.199
on it. One of the key points that you said was

00:37:04.199 --> 00:37:11.039
this is more behavioral. You need to let the

00:37:11.039 --> 00:37:17.719
S run its course and not meddle with the lever

00:37:17.719 --> 00:37:21.440
as soon as you see something. Don't be jittery.

00:37:21.610 --> 00:37:25.809
don't react to anything that you see on day one,

00:37:25.809 --> 00:37:29.349
day two. You've got to run its course. And especially

00:37:29.349 --> 00:37:34.250
with marketing, and especially if you're doing

00:37:34.250 --> 00:37:38.050
paid media and everything, it's not just what

00:37:38.050 --> 00:37:41.570
you're doing, it's also the algorithms of Google

00:37:41.570 --> 00:37:44.809
and the social media platforms that they need,

00:37:44.809 --> 00:37:48.250
those need to learn what you're trying to do.

00:37:48.380 --> 00:37:54.980
and then adapt. And so it takes a course of a

00:37:54.980 --> 00:37:57.579
couple of weeks to even some months depending

00:37:57.579 --> 00:38:02.420
on how complex the models are. I've seen this

00:38:02.420 --> 00:38:08.699
so many times that there's competing interests.

00:38:09.380 --> 00:38:13.739
One is everybody understands intuitively that

00:38:13.739 --> 00:38:18.050
we need to wait. but there's always a board meeting

00:38:18.050 --> 00:38:22.070
coming up or there's a month -end reporting that,

00:38:22.590 --> 00:38:25.849
you know, you start off with the same mind and

00:38:25.849 --> 00:38:28.309
then all of a sudden, it's end of the month,

00:38:28.409 --> 00:38:33.210
you've got to push on your paid media spend to

00:38:33.210 --> 00:38:38.750
bump up the production. So it's a lot of... It

00:38:38.750 --> 00:38:43.710
takes a lot of patience and... courage in a lot

00:38:43.710 --> 00:38:48.909
of ways to keep that going. So yeah, that was

00:38:48.909 --> 00:38:51.090
a great answer, a lot to unpack there. You also

00:38:51.090 --> 00:38:55.309
talked about diminishing returns and ROIs. There

00:38:55.309 --> 00:38:58.309
will be a certain point that a successful test

00:38:58.309 --> 00:39:02.769
will no longer give you the same ROI or even

00:39:02.769 --> 00:39:06.409
if you keep tuning it out, it'll stop at one

00:39:06.409 --> 00:39:09.750
point. So you've got to be, you don't get attached

00:39:09.750 --> 00:39:15.440
to. success too much. So yeah, that was a great

00:39:15.440 --> 00:39:23.820
answer. Thank you. What are some of the key things

00:39:23.820 --> 00:39:30.659
to measure when it comes to Google Analytics,

00:39:31.039 --> 00:39:33.179
and so you talked about funnel metrics a little

00:39:33.179 --> 00:39:36.320
bit. What are funnel metrics and what are some

00:39:36.320 --> 00:39:38.559
of the key metrics that we want to measure when

00:39:38.559 --> 00:39:43.199
it comes to websites? We'll just take the website

00:39:43.199 --> 00:39:47.059
for now, because that's a loaded question. Yeah,

00:39:47.059 --> 00:39:50.000
sure. So I'll start with what are funnel metrics,

00:39:50.980 --> 00:39:54.389
especially on any sort of... It's any site that

00:39:54.389 --> 00:39:57.710
you have a goal in mind that you want a visitor

00:39:57.710 --> 00:40:01.349
to do, but we can operate with e -commerce as

00:40:01.349 --> 00:40:03.750
an example because it's the core one we usually

00:40:03.750 --> 00:40:07.489
think about. A funnel is any time that someone's

00:40:07.489 --> 00:40:09.730
taking several different steps on a site and

00:40:09.730 --> 00:40:12.630
you're watching what those drop -offs are between

00:40:12.630 --> 00:40:15.769
your key steps. Most traditionally, we think

00:40:15.769 --> 00:40:20.239
about a shopping funnel. landing on a certain

00:40:20.239 --> 00:40:23.900
page, going into your category pages. So that's

00:40:23.900 --> 00:40:26.159
like where you have the long list of items, clicking

00:40:26.159 --> 00:40:29.440
into products and then adding to cart. And then

00:40:29.440 --> 00:40:32.019
there's checkout funnels that get really detailed

00:40:32.019 --> 00:40:34.500
into the different steps that you have. So once

00:40:34.500 --> 00:40:36.699
someone's in the cart and then they start your

00:40:36.699 --> 00:40:39.719
checkout process. Where are there any certain

00:40:39.719 --> 00:40:43.380
spots where you are seeing a drop off that you

00:40:43.380 --> 00:40:45.639
don't expect? There's always going to be leakage

00:40:45.639 --> 00:40:48.920
in any sort of funnel. But those, you know, and

00:40:48.920 --> 00:40:50.820
the funnels can be broader than that, too. Maybe

00:40:50.820 --> 00:40:52.800
there's other actions that you want people to

00:40:52.800 --> 00:40:54.659
take or you know that they take in a certain

00:40:54.659 --> 00:40:59.500
order. And you want to evaluate how each of those

00:40:59.500 --> 00:41:01.500
steps are going and whether you're expecting

00:41:01.500 --> 00:41:04.800
the kind of leakage that you see or not. And

00:41:04.800 --> 00:41:08.420
like, when it comes to what to measure, what

00:41:08.420 --> 00:41:11.519
metrics to measure. There's both kind of some

00:41:11.519 --> 00:41:15.280
core metrics that I think are important for any

00:41:15.280 --> 00:41:18.320
sort of website based business to look at. And

00:41:18.320 --> 00:41:20.360
then there's the whole conversation about KPIs.

00:41:20.380 --> 00:41:23.159
And I think that a lot of people interchangeably

00:41:23.159 --> 00:41:25.780
use those two terms, but to me they're very different.

00:41:26.320 --> 00:41:29.320
And a KPI is like, There's only a few of them

00:41:29.320 --> 00:41:31.159
that you should have as a business, and they

00:41:31.159 --> 00:41:33.639
are the things that you are completely informing

00:41:33.639 --> 00:41:37.360
your strategy on. Generally, I like to do them

00:41:37.360 --> 00:41:39.960
in a pairing. I can't take credit for that. That's

00:41:39.960 --> 00:41:46.780
an Avanash thing. But what I would say some key

00:41:46.780 --> 00:41:50.719
metrics are is, and this one's changing, so right

00:41:50.719 --> 00:41:52.440
now it's bounce rate, but it's going to turn

00:41:52.440 --> 00:41:54.780
into engagement rate. But what you're trying

00:41:54.780 --> 00:41:57.699
to understand from this metric is, It's kind

00:41:57.699 --> 00:42:00.599
of like that you gross principle. So does someone

00:42:00.599 --> 00:42:02.719
go to your site and go, this isn't for me and

00:42:02.719 --> 00:42:06.920
immediately leave and you want to see whether

00:42:06.920 --> 00:42:09.059
that number is where you expect it to be. There's

00:42:09.059 --> 00:42:11.289
always going to be some people that go. hit a

00:42:11.289 --> 00:42:14.070
site and then leave immediately. But that's one

00:42:14.070 --> 00:42:15.809
that's like a great health gauge of pretty much

00:42:15.809 --> 00:42:18.130
any type of website. If you have a really, really

00:42:18.130 --> 00:42:20.550
high bounce rate, then you might be running into

00:42:20.550 --> 00:42:22.949
some issues where your site's not loading fast

00:42:22.949 --> 00:42:25.170
enough. Something about the content when first

00:42:25.170 --> 00:42:28.250
people come isn't what they were expecting. So

00:42:28.250 --> 00:42:31.880
that's one of the core ones. Some other ones

00:42:31.880 --> 00:42:34.639
that are on the opposite side of the funnel are

00:42:34.639 --> 00:42:37.420
obviously transactions. Conversion rate, we've

00:42:37.420 --> 00:42:40.559
already talked a lot about. The universal analytics

00:42:40.559 --> 00:42:43.920
definition is sessions by transactions. So you're

00:42:43.920 --> 00:42:46.820
looking at the ratio of the number of sessions

00:42:46.820 --> 00:42:48.980
that are happening, how many transactions are

00:42:48.980 --> 00:42:52.980
there. And then before that is add to cart. Add

00:42:52.980 --> 00:42:54.860
to cart is a really core one for e -commerce

00:42:54.860 --> 00:42:58.079
websites because it's like an intent to buy.

00:42:58.429 --> 00:43:01.289
but it's still pretty far away from actually

00:43:01.289 --> 00:43:05.349
purchasing. E -commerce sites and consumers currently

00:43:05.349 --> 00:43:09.269
can tend to use your ad to cart as a wish list,

00:43:09.550 --> 00:43:12.809
but it's one of the first intents to purchase

00:43:12.809 --> 00:43:14.789
that's a little more serious than just browsing

00:43:14.789 --> 00:43:17.349
products. I think that those are some of the

00:43:17.349 --> 00:43:21.869
broad ones, but what one you focus on improving

00:43:21.869 --> 00:43:24.630
is really specific to what your goals are in

00:43:24.630 --> 00:43:30.639
the business. That makes a lot of sense. Yeah,

00:43:30.639 --> 00:43:33.119
there's so many things that you could focus on,

00:43:33.119 --> 00:43:35.800
but yeah, you've got to define your pathway and

00:43:35.800 --> 00:43:40.400
then start measuring from there on. And also

00:43:40.400 --> 00:43:44.280
for some of the most advanced companies with

00:43:44.280 --> 00:43:49.900
advanced infrastructure, they also combine some

00:43:49.900 --> 00:43:52.920
of the marketing data within their own internal

00:43:52.920 --> 00:43:56.179
data and other third party data. You can match

00:43:56.179 --> 00:43:59.670
your data from Google Analytics with your paid

00:43:59.670 --> 00:44:03.710
media, and also your customer care, your email

00:44:03.710 --> 00:44:07.530
metrics, and a bunch of different things, which

00:44:07.530 --> 00:44:13.750
is where I start geeking out. In my mind, there

00:44:13.750 --> 00:44:17.590
are kind of three levels. You have your campaign

00:44:17.590 --> 00:44:23.150
metrics for any measurement, and then you've

00:44:23.150 --> 00:44:26.929
got your engagement metrics, which is Campaign

00:44:26.929 --> 00:44:29.449
metrics are what is the campaign doing, how many

00:44:29.449 --> 00:44:33.269
visits, how many add to cart and all that. Engagement

00:44:33.269 --> 00:44:40.190
is after your customer purchases, what do they

00:44:40.190 --> 00:44:46.130
do? Do they use your product? How many times

00:44:46.130 --> 00:44:48.849
do they use your product? How well or how many

00:44:48.849 --> 00:44:51.170
products do they use and all that. And then you

00:44:51.170 --> 00:44:54.730
have performance metrics, which ties to the KPIs

00:44:54.730 --> 00:44:57.599
that you talked about. performance metrics in

00:44:57.599 --> 00:45:04.900
my mind are aligned with your company's goals.

00:45:05.599 --> 00:45:08.539
It could be acquisitions goals, retention goals,

00:45:08.659 --> 00:45:11.079
and stuff like that. What are some of your thoughts

00:45:11.079 --> 00:45:19.360
on it? Are companies doing it this way or is

00:45:19.360 --> 00:45:22.480
it just a few companies that are thinking in

00:45:22.480 --> 00:45:29.730
these terms? It's a good question. And I think

00:45:29.730 --> 00:45:32.090
it's really important to distinguish them. It

00:45:32.090 --> 00:45:33.889
doesn't really matter how you're going to slice

00:45:33.889 --> 00:45:36.110
it. I think there's definitely a few different

00:45:36.110 --> 00:45:41.570
practices. I like that one. It is just getting

00:45:41.570 --> 00:45:46.269
that distinguishing factor between this is something

00:45:46.269 --> 00:45:49.400
that is measured for this reason, here's who

00:45:49.400 --> 00:45:52.340
looks at it, and then all the way up to this

00:45:52.340 --> 00:45:56.679
is what we're being judged on for our bonuses

00:45:56.679 --> 00:46:00.400
and for the board of directors is looking at

00:46:00.400 --> 00:46:04.320
these numbers. I've seen it done quite a few

00:46:04.320 --> 00:46:06.900
different ways. I've also seen it done really

00:46:06.900 --> 00:46:10.429
divided by And this is one of the ways that I

00:46:10.429 --> 00:46:12.989
divide it in a governance thing. I wrote a governance

00:46:12.989 --> 00:46:15.730
article recently about this, like who is looking

00:46:15.730 --> 00:46:19.070
at the metrics and analyzing it. So like your

00:46:19.070 --> 00:46:22.329
definition that like performance would be like.

00:46:22.480 --> 00:46:24.880
board of directors, C -level directors are looking

00:46:24.880 --> 00:46:27.179
at it. And then there's some really tactical

00:46:27.179 --> 00:46:30.159
metrics that an analyst is going to look at.

00:46:30.619 --> 00:46:33.300
I've also seen people talk about when they need

00:46:33.300 --> 00:46:36.519
to be evaluated, like how often. There's some

00:46:36.519 --> 00:46:39.820
of those metrics that even the evaluation of

00:46:39.820 --> 00:46:43.039
them should be automated. And that's really an

00:46:43.039 --> 00:46:45.739
opportunity for optimizing them using machine

00:46:45.739 --> 00:46:47.199
learning or things like that. Those are ones

00:46:47.199 --> 00:46:49.360
that you don't really want to waste an analyst's

00:46:49.360 --> 00:46:52.449
time on a day -to -day basis. there's another

00:46:52.449 --> 00:46:55.190
bucket that's things that need to be evaluated

00:46:55.190 --> 00:46:58.510
weekly or monthly or really monitored, and then

00:46:58.510 --> 00:46:59.969
so on and so forth. There might be something

00:46:59.969 --> 00:47:02.730
that's really worthwhile for a deep dive analysis

00:47:02.730 --> 00:47:06.230
once a year. So there's so many different ways

00:47:06.230 --> 00:47:08.349
to slice it, which is not like I'm like, here's

00:47:08.349 --> 00:47:12.849
an easy framework. And to answer that second

00:47:12.849 --> 00:47:15.210
part about, do I see companies using this? The

00:47:15.210 --> 00:47:17.690
answer is kind of yes and no. I see a lot of

00:47:17.690 --> 00:47:22.300
companies misusing some of these and really The

00:47:22.300 --> 00:47:26.639
biggest frustration or pain point I see is, especially

00:47:26.639 --> 00:47:29.599
on the leadership level, misinterpreting a metric

00:47:29.599 --> 00:47:33.139
to a KPI. And they say, we have this KPI dashboard,

00:47:33.139 --> 00:47:35.639
you want to see it? And it has like... 300 different

00:47:35.639 --> 00:47:37.679
metrics on it. I'm like, none of these are KPIs

00:47:37.679 --> 00:47:39.760
because you don't know, you know, there's like

00:47:39.760 --> 00:47:41.800
some little cartoon about it that has like a

00:47:41.800 --> 00:47:43.139
dashboard like that. And they're like, what is

00:47:43.139 --> 00:47:45.519
this one trending down? It says, it's our understanding

00:47:45.519 --> 00:47:48.719
of all these metrics. So, you know, I think that

00:47:48.719 --> 00:47:51.239
that's like the pain point there is that I think

00:47:51.239 --> 00:47:53.679
that there's different differing levels of maturity

00:47:53.679 --> 00:47:56.860
about understanding some of these things. And

00:47:56.860 --> 00:47:59.119
all companies are in different places, even like

00:47:59.119 --> 00:48:01.659
in the same industry. But that's the one where

00:48:01.659 --> 00:48:04.619
I see the most frequent misunderstanding. I see

00:48:04.619 --> 00:48:08.940
very few companies focusing down their KPIs properly

00:48:08.940 --> 00:48:13.059
by company -wide, function -wide, et cetera,

00:48:13.099 --> 00:48:16.039
to really focus people's efforts. Yeah, that's

00:48:16.039 --> 00:48:21.760
a great point. What you said kind of reminded

00:48:21.760 --> 00:48:25.380
me of another framework that one of my mentors

00:48:25.380 --> 00:48:29.500
talked about. You have your stuff that your executives

00:48:29.500 --> 00:48:32.500
look at, and then your managers look at, and

00:48:32.500 --> 00:48:35.960
then your technical teams look at. They're different

00:48:35.960 --> 00:48:39.260
metrics, and you have to treat them differently.

00:48:39.539 --> 00:48:43.139
And then, like you said, the time horizons that

00:48:43.139 --> 00:48:46.739
they need to be looked at is different from each

00:48:46.739 --> 00:48:54.679
other. That's a good point to remember. And also,

00:48:55.340 --> 00:48:58.690
with this podcast, we're trying to make it more

00:48:58.690 --> 00:49:05.590
accessible. This is more advanced and we don't

00:49:05.590 --> 00:49:10.389
really, not every company really needs all the

00:49:10.389 --> 00:49:12.889
infrastructure and need to be measuring everything

00:49:12.889 --> 00:49:16.409
in order to be successful. You can start with

00:49:16.409 --> 00:49:18.530
marketing, you can start with the final metrics

00:49:18.530 --> 00:49:22.349
you talked about and then slowly develop. and

00:49:22.349 --> 00:49:25.670
grow into the other KPIs and performance metrics

00:49:25.670 --> 00:49:31.650
that we touched on. Yeah, definitely. I think

00:49:31.650 --> 00:49:33.570
that sometimes people think that they have to

00:49:33.570 --> 00:49:35.489
build it all at once and they have to be flying

00:49:35.489 --> 00:49:37.750
their rocket ship when they're still waddling

00:49:37.750 --> 00:49:42.170
around in their diapers. And there's so much

00:49:42.170 --> 00:49:46.250
that you can do with just, say, the... data that

00:49:46.250 --> 00:49:49.250
you have from your advertising platforms. You

00:49:49.250 --> 00:49:51.429
can do a lot with just that, or you can do a

00:49:51.429 --> 00:49:53.309
lot with just customer feedback. And I think

00:49:53.309 --> 00:49:57.349
that's a really good call out that it feels sometimes

00:49:57.349 --> 00:50:00.789
like everybody has to have their own data link

00:50:00.789 --> 00:50:02.869
or data mesh and all these fancy terms and do

00:50:02.869 --> 00:50:04.730
all these things and have machine learning and

00:50:04.730 --> 00:50:08.550
AI. And the answer is you don't. All you need

00:50:08.550 --> 00:50:13.099
to do is to execute on your data. better than

00:50:13.099 --> 00:50:16.099
your competitors. And if your competitors are

00:50:16.099 --> 00:50:18.519
other small startups, then really what that means

00:50:18.519 --> 00:50:21.059
is you're leveraging market intelligence a little

00:50:21.059 --> 00:50:25.800
bit smarter. looking at your forecasting a little

00:50:25.800 --> 00:50:28.280
bit smarter. Yes, even really small companies

00:50:28.280 --> 00:50:30.880
can be doing really smart forecasting to make

00:50:30.880 --> 00:50:33.159
it a little bit more sustainable, to have the

00:50:33.159 --> 00:50:35.559
right amount of inventory, to predict where they're

00:50:35.559 --> 00:50:38.440
going to be in a year. There's things like that

00:50:38.440 --> 00:50:40.780
that really you don't need that much data to

00:50:40.780 --> 00:50:43.079
do. You just need the right people thinking about

00:50:43.079 --> 00:50:47.300
it the right way. Absolutely. You also passed

00:50:47.300 --> 00:50:52.949
on. misinterpreting metrics. That's something

00:50:52.949 --> 00:50:56.989
that I've got a lot of thoughts on. But there's

00:50:56.989 --> 00:51:06.230
two types of extremes. People who already have

00:51:06.230 --> 00:51:11.010
a story and then they just look at the data just

00:51:11.010 --> 00:51:17.010
to confirm or look at data with a lens that they

00:51:17.010 --> 00:51:20.889
already have. no matter what the data says, they

00:51:20.889 --> 00:51:24.650
have the same story. Call them wisdom analytics

00:51:24.650 --> 00:51:30.010
versus on the other side, you know, executives

00:51:30.010 --> 00:51:35.170
could be too jittery and look at a data set and

00:51:35.170 --> 00:51:38.510
a result and just change everything that they've

00:51:38.510 --> 00:51:40.769
been doing, change their course of action. So

00:51:40.769 --> 00:51:44.630
I think it's definitely a balance. And like you

00:51:44.630 --> 00:51:47.929
said, you've got to have the right frame of mind.

00:51:48.619 --> 00:51:51.800
to write people looking at their data with the

00:51:51.800 --> 00:51:55.780
right frame of mind is super important. Oh, yeah.

00:51:58.820 --> 00:52:01.539
I think it goes back to that feeling confident

00:52:01.539 --> 00:52:04.500
good about data as part of it. But then also,

00:52:04.500 --> 00:52:08.039
yeah, creating a culture, whatever you decide,

00:52:08.199 --> 00:52:09.679
it doesn't have to be the same at every company.

00:52:09.900 --> 00:52:14.739
That's like, we will look at this honestly. be

00:52:14.739 --> 00:52:16.599
realistic, and then also having people who can

00:52:16.599 --> 00:52:19.119
really temper something scary for executives.

00:52:19.440 --> 00:52:20.920
Because I think that's something that's sometimes

00:52:20.920 --> 00:52:23.480
missing is that we go in and we start to be like,

00:52:23.500 --> 00:52:24.760
we're going to be data driven. We're going to

00:52:24.760 --> 00:52:26.760
do so much with data. And then you start giving

00:52:26.760 --> 00:52:28.820
a lot of information to executives that aren't

00:52:28.820 --> 00:52:31.000
familiar with it. And it stresses them out because

00:52:31.000 --> 00:52:33.980
no one is being that translator between an analyst

00:52:33.980 --> 00:52:36.780
who's going to look at it super objectively or

00:52:36.780 --> 00:52:39.980
like really clinically to the executive where

00:52:39.980 --> 00:52:41.500
they're like, what does this mean? And they're

00:52:41.500 --> 00:52:43.300
jumping to conclusions where if they have like

00:52:43.300 --> 00:52:45.679
a data leader, they're going to be like, well,

00:52:45.780 --> 00:52:47.659
you know, here's kind of the nuances to this.

00:52:48.000 --> 00:52:51.340
Here's where you need to make a decision. Yeah,

00:52:51.340 --> 00:52:53.179
that is so much important. And that's something

00:52:53.179 --> 00:52:55.400
you talk about and I'm super passionate about

00:52:55.400 --> 00:52:59.440
too. Like I've started to write in my blogs,

00:53:00.000 --> 00:53:04.019
like the storytelling aspect to data is super

00:53:04.019 --> 00:53:08.420
key. And you got to have a contextual understanding

00:53:08.420 --> 00:53:11.579
of what the data is telling you because this

00:53:11.639 --> 00:53:17.340
same result or same set of data, 10 people can

00:53:17.340 --> 00:53:20.139
see it and interpret it 10 different ways. But

00:53:20.139 --> 00:53:23.000
you really have to have somebody who understands

00:53:23.000 --> 00:53:26.400
the product, the company, and then the data as

00:53:26.400 --> 00:53:30.059
well to tie it all together and not just an analyst

00:53:30.059 --> 00:53:37.320
looking at objective analysis of the data. Super

00:53:37.320 --> 00:53:42.670
helpful. I think that kind of There's a good

00:53:42.670 --> 00:53:49.010
segue into our topic on leadership. At InsightLine

00:53:49.010 --> 00:53:53.449
Analytics, you cover all aspects of data, infrastructure,

00:53:54.630 --> 00:53:59.449
marketing, and even results and measurements

00:53:59.449 --> 00:54:03.670
and everything. How do you assess what stage

00:54:03.670 --> 00:54:08.090
a client is at and what your work is going to

00:54:08.090 --> 00:54:11.949
be? When you go in, how do you assess? That's

00:54:11.949 --> 00:54:14.769
a good question. I think there's a few different

00:54:14.769 --> 00:54:17.250
things. One that we're currently actually working

00:54:17.250 --> 00:54:21.610
on to be even more measurable in that is creating

00:54:21.610 --> 00:54:24.250
a data maturity model. There's a few digital

00:54:24.250 --> 00:54:26.429
maturity models out there, but there's not as

00:54:26.429 --> 00:54:30.849
many super focused on data. That's one thing

00:54:30.849 --> 00:54:34.860
that's to come. soon. So I'm kind of like letting

00:54:34.860 --> 00:54:38.480
the cat out of the bag about that. But as experienced

00:54:38.480 --> 00:54:42.360
people in this field, it's something that like

00:54:42.360 --> 00:54:44.300
myself and like the other senior analysts can

00:54:44.300 --> 00:54:47.460
see right away. There's like a few key aspects

00:54:47.460 --> 00:54:51.360
that indicate where someone's at. And so that

00:54:51.360 --> 00:54:53.199
part of it's coming from experience now where

00:54:53.199 --> 00:54:55.940
I kind of know, depending on even what a company

00:54:55.940 --> 00:54:58.199
is approaching us with. where they're probably

00:54:58.199 --> 00:55:00.400
going to be at in their maturity. And then we

00:55:00.400 --> 00:55:02.099
start to get a little bit more into the details

00:55:02.099 --> 00:55:03.760
as we get there. Of course, you don't want to

00:55:03.760 --> 00:55:06.280
make any snap judgments of, wow, this company

00:55:06.280 --> 00:55:10.579
is really immature or really mature. But there's

00:55:10.579 --> 00:55:12.360
some indicators depending on, well, what are

00:55:12.360 --> 00:55:14.619
they valuing right now? What's really important

00:55:14.619 --> 00:55:17.420
to them? What are they super focused on? And

00:55:17.420 --> 00:55:20.739
then There's also sometimes where we go into

00:55:20.739 --> 00:55:24.000
this a lot deeper, which is we'll conduct an

00:55:24.000 --> 00:55:26.239
entire data strategy project where we're going

00:55:26.239 --> 00:55:30.440
in and mapping out their marketing and data tech

00:55:30.440 --> 00:55:33.920
stack and comparing that to what we consider

00:55:33.920 --> 00:55:36.780
industry standard. And we're evaluating like

00:55:36.780 --> 00:55:39.960
their whole team structure and doing some needs

00:55:39.960 --> 00:55:41.940
assessments of like, okay, well, they're here

00:55:41.940 --> 00:55:44.039
and they need this, but they don't have it yet.

00:55:44.400 --> 00:55:46.800
And so that's like the really comprehensive way

00:55:46.800 --> 00:55:48.929
that we dig in. to it, especially with bigger

00:55:48.929 --> 00:55:52.690
complex organizations. And that's where we get

00:55:52.690 --> 00:55:56.949
a lot of information and are able to hold up

00:55:56.949 --> 00:55:59.150
the mirror for an organization and say, well,

00:55:59.190 --> 00:56:01.710
this is what we see as far as where you're at.

00:56:03.150 --> 00:56:06.750
So we have some pretty detailed ways to do it.

00:56:06.769 --> 00:56:09.929
And then there's also these initial, what is

00:56:09.929 --> 00:56:12.969
someone asking us for? There's a very big difference

00:56:12.969 --> 00:56:16.019
between a company that's asking for just like

00:56:16.019 --> 00:56:17.900
conversion rate optimization for their site.

00:56:18.320 --> 00:56:21.219
Hey, we really want to get more sales and we

00:56:21.219 --> 00:56:23.820
want to lift our conversion rate and make way

00:56:23.820 --> 00:56:26.719
more money. That's a very different level of

00:56:26.719 --> 00:56:28.199
maturity than someone who's coming to us and

00:56:28.199 --> 00:56:31.639
being like, We already have unified customer

00:56:31.639 --> 00:56:34.559
records, and we really want to do some cool stuff

00:56:34.559 --> 00:56:37.739
with that. We want to start doing machine learning,

00:56:37.780 --> 00:56:39.579
and we want to apply that to our advertising.

00:56:39.920 --> 00:56:41.800
Those are two very different conversations, and

00:56:41.800 --> 00:56:44.079
there's nothing wrong with either of them, but

00:56:44.079 --> 00:56:46.559
it is a big indicator of how mature they are.

00:56:47.380 --> 00:56:51.019
Absolutely, and it's super important for us as

00:56:51.019 --> 00:56:53.460
data experts and marketing experts to go in with

00:56:53.460 --> 00:56:55.739
an open mind. You touched on that. You don't

00:56:55.739 --> 00:57:01.239
want to shame anybody for where they are. And

00:57:01.239 --> 00:57:03.860
everybody goes through the same process. People

00:57:03.860 --> 00:57:06.559
are at different phases and different points

00:57:06.559 --> 00:57:10.099
in time. So I think, yeah, definitely keeping

00:57:10.099 --> 00:57:12.219
an open mind and understanding where they're

00:57:12.219 --> 00:57:19.480
at will be super key. And also from a data perspective,

00:57:19.739 --> 00:57:23.260
you did touch on this, like you have companies

00:57:23.260 --> 00:57:26.380
running reports off of Excel sheets. Yeah, that's

00:57:26.380 --> 00:57:32.280
one end. And then you next go into having a marketing

00:57:32.280 --> 00:57:34.179
infrastructure in place, data infrastructure

00:57:34.179 --> 00:57:37.099
in place, where you can get some of those funnel

00:57:37.099 --> 00:57:41.739
metrics. And also next, you could probably look

00:57:41.739 --> 00:57:46.579
at BI and automating reports with visualizations.

00:57:47.400 --> 00:57:55.210
So at this day, you can talk about looking at

00:57:55.210 --> 00:57:58.110
historical data and understanding patterns. Hey,

00:57:58.110 --> 00:58:02.769
what's my seasonality of the sales or retention

00:58:02.769 --> 00:58:06.510
or stuff like that? So pattern recognition, this

00:58:06.510 --> 00:58:10.869
is where I think my opinion is the biggest ROI

00:58:10.869 --> 00:58:13.929
because running reports off of Excel or just

00:58:13.929 --> 00:58:18.849
looking at one or two metrics wouldn't give you

00:58:18.849 --> 00:58:23.889
a larger perspective of to be able to get those

00:58:23.889 --> 00:58:27.869
patterns, recognize those patterns. So BI and

00:58:27.869 --> 00:58:31.030
visualization is super helpful at that point.

00:58:31.369 --> 00:58:36.250
And then you go into predictive. So until now,

00:58:36.289 --> 00:58:38.929
you looked at historical data. Now you can start

00:58:38.929 --> 00:58:42.550
building models to predict what will happen in

00:58:42.550 --> 00:58:45.150
the future based off of what has happened in

00:58:45.150 --> 00:58:50.230
the past. And then you also have these other

00:58:50.230 --> 00:58:53.000
phases where you have machine learning models,

00:58:53.280 --> 00:58:56.300
you've got your data science and AI models. So

00:58:56.300 --> 00:59:03.800
there's various phases to data maturity. So that's

00:59:03.800 --> 00:59:08.750
super key and super important. Oh, certainly.

00:59:09.230 --> 00:59:11.789
And I think for people who are listening, too,

00:59:12.349 --> 00:59:15.590
it never happens just quite that linearly, too.

00:59:15.590 --> 00:59:17.829
So that's a good call out. I've known companies

00:59:17.829 --> 00:59:20.590
that I would think are fairly mature. They're

00:59:20.590 --> 00:59:22.550
starting to do some predictive stuff, and they're

00:59:22.550 --> 00:59:24.630
getting playful with that. And they still have

00:59:24.630 --> 00:59:27.170
a few Excel sheets they're using to report to

00:59:27.170 --> 00:59:30.570
executives. So there's one thing there. But also

00:59:30.570 --> 00:59:35.949
would agree with you that that's where advanced

00:59:35.949 --> 00:59:39.090
data science kind of stuff starts to have a good

00:59:39.090 --> 00:59:41.550
ROI is always predict. I'm always going to say,

00:59:42.030 --> 00:59:44.150
do predictive stuff. And do predictive stuff

00:59:44.150 --> 00:59:45.909
that really influences the business if you're

00:59:45.909 --> 00:59:48.650
not predicting. And you can do some lighter versions

00:59:48.650 --> 00:59:50.969
before you even get there. Like, you should be

00:59:50.969 --> 00:59:53.610
forecasting. Even if you're still in Excel, you

00:59:53.610 --> 00:59:56.929
should be forecasting. But getting to the juiciness

00:59:56.929 --> 01:00:01.949
of churn prediction or things like that is really

01:00:01.949 --> 01:00:05.059
where you can highlight issues before they happen,

01:00:05.239 --> 01:00:07.800
and take advantage of opportunities before they

01:00:07.800 --> 01:00:11.940
happen. Yeah, that's a great point. And also,

01:00:12.059 --> 01:00:19.480
yeah, you have to be understanding of where you

01:00:19.480 --> 01:00:24.619
are. And not everybody has to do everything,

01:00:24.739 --> 01:00:27.400
but you have to just know where you are. And

01:00:27.400 --> 01:00:29.639
also, one of the things that you talk about a

01:00:29.639 --> 01:00:35.539
lot is be customer centric, even if you're working

01:00:35.539 --> 01:00:38.760
in a corporate environment, look at the business

01:00:38.760 --> 01:00:42.460
needs and then develop a solution and not have

01:00:42.460 --> 01:00:44.980
a solution in place just because you know how

01:00:44.980 --> 01:00:50.619
to do it, then there's no use for it. If the

01:00:50.619 --> 01:00:55.539
marketing team is not ready for it, then you

01:00:55.539 --> 01:01:00.840
can't really, I mean, I can't build a CDP solution.

01:01:01.130 --> 01:01:05.230
if the marketing team is still not ready to use

01:01:05.230 --> 01:01:08.929
event -based data and event -based marketing

01:01:08.929 --> 01:01:11.849
yet, if they are just sending out onboarding

01:01:11.849 --> 01:01:17.110
and retention emails and you're talking about

01:01:17.110 --> 01:01:20.309
engaging customers where they are and looking

01:01:20.309 --> 01:01:23.809
at the different pages of the app that they're

01:01:23.809 --> 01:01:31.000
looking at and changing the changing what you

01:01:31.000 --> 01:01:33.519
show based off of that. So basically, event -based.

01:01:34.019 --> 01:01:40.880
You've got to have your customer in mind before

01:01:40.880 --> 01:01:42.800
developing solutions. Sorry, that was a long

01:01:42.800 --> 01:01:46.800
-winded thing, but yeah. Definitely keep your

01:01:46.800 --> 01:01:50.260
customers in mind before developing your solution.

01:01:56.539 --> 01:02:01.400
We talked about infrastructure and data maturity.

01:02:02.000 --> 01:02:05.019
How do you handle data adoption and the literacy

01:02:05.019 --> 01:02:09.719
with your clients? With difficulty. No, just

01:02:09.719 --> 01:02:16.099
kidding. It's a long game. And those two things

01:02:16.099 --> 01:02:21.980
can be their own whole project strategic function

01:02:21.980 --> 01:02:26.280
of your company. it does take some time. And

01:02:26.280 --> 01:02:28.539
that's one of the reasons I really love when

01:02:28.539 --> 01:02:31.179
we have these opportunities to stay in a company

01:02:31.179 --> 01:02:33.400
for years and really help them develop this.

01:02:34.119 --> 01:02:37.280
But there's the two aspects. One, like you said,

01:02:37.480 --> 01:02:41.760
data literacy is something where I'll define

01:02:41.760 --> 01:02:43.820
the term really quickly, too, because it sounds

01:02:43.820 --> 01:02:46.659
kind of like, one, sometimes I feel like people

01:02:46.659 --> 01:02:48.000
could be like, well, that's insulting. I know

01:02:48.000 --> 01:02:50.960
how to. look at data or whatever. And then also

01:02:50.960 --> 01:02:53.480
it sounds fancy at the same time of like, it's

01:02:53.480 --> 01:02:56.500
another term like strategy or synergy or something

01:02:56.500 --> 01:02:59.139
like that. But what it really means is like,

01:02:59.139 --> 01:03:01.519
do people understand the data that your company

01:03:01.519 --> 01:03:05.059
creates? Do they feel comfortable reading it

01:03:05.059 --> 01:03:07.519
when it's presented to them in a deck? And do

01:03:07.519 --> 01:03:11.679
they trust it? And do they feel comfortable using

01:03:11.679 --> 01:03:14.679
what's coming out of their team to justify what

01:03:14.679 --> 01:03:18.639
they're doing? And so that can have a lot to

01:03:18.639 --> 01:03:23.099
unpack there, right? Because you need to evaluate

01:03:23.099 --> 01:03:26.199
first, like, what level of data literacy does

01:03:26.199 --> 01:03:29.960
everyone need? How will you train them? Because

01:03:29.960 --> 01:03:32.179
it's something that does require some training.

01:03:32.780 --> 01:03:34.420
And then the other parts of it, like I said,

01:03:34.539 --> 01:03:37.440
trust in data and understanding of what you collect

01:03:37.440 --> 01:03:40.300
and why, kind of bleeds into the second part,

01:03:40.320 --> 01:03:43.000
which is about communication and adoption of

01:03:43.000 --> 01:03:48.739
data. I think that communication of data quality

01:03:48.739 --> 01:03:51.139
and governance. Everything like that is something

01:03:51.139 --> 01:03:53.659
that should be operating at the highest level

01:03:53.659 --> 01:03:57.099
of the company. If possible, owned by a whole

01:03:57.099 --> 01:03:59.880
group like governance, for example, governance

01:03:59.880 --> 01:04:02.340
and maybe communications tied together actually,

01:04:03.320 --> 01:04:06.760
because it's where things can fall apart and

01:04:06.760 --> 01:04:09.940
also where you can get the strongest message.

01:04:10.320 --> 01:04:12.630
So this is something that If someone wanted to

01:04:12.630 --> 01:04:14.590
take a script and do best practice, that's what

01:04:14.590 --> 01:04:19.550
you would do. Because without it, you have a

01:04:19.550 --> 01:04:22.650
really common scenario, which is that there's

01:04:22.650 --> 01:04:25.829
often conflicts between departments, and data

01:04:25.829 --> 01:04:29.030
is used more as a weapon than it is as a tool

01:04:29.030 --> 01:04:30.789
to get to understand each other better and your

01:04:30.789 --> 01:04:34.909
customer better. Just super unfortunate. One

01:04:34.909 --> 01:04:37.070
of the ways that you get around that is by building

01:04:37.070 --> 01:04:39.610
a culture where you and common agreement about

01:04:39.610 --> 01:04:42.909
all the data. So making sure that you're really

01:04:42.909 --> 01:04:46.750
evaluating what is our source of truth? How does

01:04:46.750 --> 01:04:50.110
everybody use that? What are the agreed upon

01:04:50.110 --> 01:04:54.130
terms we use as a company? And how do we validate

01:04:54.130 --> 01:04:56.090
that and make sure that whatever we're communicating

01:04:56.090 --> 01:04:58.269
all lines up to it? And that's so much easier

01:04:58.269 --> 01:05:01.190
to do at a really high executive level and then

01:05:01.190 --> 01:05:04.610
get buy -in and communicate it over time. All

01:05:04.610 --> 01:05:06.989
of that I'm saying, it takes years to develop

01:05:06.989 --> 01:05:10.130
properly. So I summarized it and said, that's

01:05:10.130 --> 01:05:12.550
all you got to do. Ta -da. Now you have this

01:05:12.550 --> 01:05:14.670
great company and everybody loves each other.

01:05:14.690 --> 01:05:16.630
And there's no fighting between your IT team

01:05:16.630 --> 01:05:19.289
and your analytics team. That's not necessarily

01:05:19.289 --> 01:05:22.110
the truth. But it's something where it's like

01:05:22.110 --> 01:05:24.309
an iterative process of evaluating, where do

01:05:24.309 --> 01:05:27.269
we have gaps? And one trick would be wherever

01:05:27.269 --> 01:05:31.000
you see conflict, like frustration. you hear

01:05:31.000 --> 01:05:33.059
a repeated term like, I don't trust that data,

01:05:33.199 --> 01:05:35.460
that's where you can dig in to really understand

01:05:35.460 --> 01:05:39.780
what's going on. Yeah, that's super important.

01:05:40.840 --> 01:05:46.739
Yeah, definitely top -down approach is a better

01:05:46.739 --> 01:05:51.980
way to go. And also you kind of identify the

01:05:51.980 --> 01:05:56.099
pain points and areas of pain points and then

01:05:56.099 --> 01:06:00.400
you dig into it. to understand how you can get

01:06:00.400 --> 01:06:02.960
better adoption, you know, solve some of these

01:06:02.960 --> 01:06:06.280
data quality issues or data issues to get better

01:06:06.280 --> 01:06:14.000
adoption. Awesome. So we talked a bunch of about

01:06:14.000 --> 01:06:22.260
data. Your main role as a founder of your company

01:06:22.260 --> 01:06:28.880
is to lead your team and You talked a lot about

01:06:28.880 --> 01:06:32.059
working with the clients and everything. When

01:06:32.059 --> 01:06:36.739
it comes to managing the team, I believe you

01:06:36.739 --> 01:06:43.619
said you run a six -member team. You're a couple

01:06:43.619 --> 01:06:48.480
of years old as a company, so that's a good size.

01:06:49.159 --> 01:06:54.639
Is your team remote or do you have a hybrid model?

01:06:55.900 --> 01:07:00.119
You're in Colorado, correct? Yeah. We like to

01:07:00.119 --> 01:07:03.440
say we're a remote first company, but we actually

01:07:03.440 --> 01:07:06.199
do have a headquarters in Durango, Colorado,

01:07:06.239 --> 01:07:08.760
and we have four of us out of the office fairly

01:07:08.760 --> 01:07:12.380
regularly, but it's entirely up to each employee.

01:07:13.519 --> 01:07:16.519
We have some that are... We've actually had some

01:07:16.519 --> 01:07:20.039
shifts recently, but we had a few in Texas, and

01:07:20.039 --> 01:07:22.119
now pretty much all of it's moving to Colorado,

01:07:22.179 --> 01:07:25.579
which is interesting. And that's not on purpose

01:07:25.579 --> 01:07:27.360
or anything. It's just kind of how it's happened.

01:07:28.119 --> 01:07:31.400
But that's the model is that there's some of

01:07:31.400 --> 01:07:33.099
us that like to come into the office and live

01:07:33.099 --> 01:07:36.480
here, but also it's not necessarily like a major

01:07:36.480 --> 01:07:39.619
metro. So I don't have any expectations for people

01:07:39.619 --> 01:07:44.679
to relocate or anything like that. get together

01:07:44.679 --> 01:07:47.099
at least once a year, as long as everyone wants

01:07:47.099 --> 01:07:49.420
to. If someone said, nah, I just don't feel like

01:07:49.420 --> 01:07:51.539
doing that, we wouldn't hold it against them.

01:07:51.800 --> 01:07:54.219
And then we do a little bit of travel to our

01:07:54.219 --> 01:07:57.519
clients when it's really necessary and they really

01:07:57.519 --> 01:08:01.719
want us to. But we started the company in 2020,

01:08:02.480 --> 01:08:05.539
so really at the start of everything going remote

01:08:05.539 --> 01:08:08.460
for work. And it seems like it's a model that

01:08:08.460 --> 01:08:11.679
really works well for our team. Yeah, definitely.

01:08:14.219 --> 01:08:17.659
startups and especially in data and marketing

01:08:17.659 --> 01:08:22.859
and some of these areas. Remote work is definitely

01:08:22.859 --> 01:08:27.479
feasible and also works really well. How do you

01:08:27.479 --> 01:08:31.720
handle onboarding and building team culture and

01:08:31.720 --> 01:08:35.680
all that when people are scattered around the

01:08:35.680 --> 01:08:38.460
country and the globe? It's definitely a bit

01:08:38.460 --> 01:08:40.960
more difficult than when you could just get someone

01:08:40.960 --> 01:08:43.180
in the office and have them meet everyone in

01:08:43.180 --> 01:08:45.720
person. There's a few different things we do.

01:08:45.920 --> 01:08:49.460
We do actually try our best to have people's

01:08:49.460 --> 01:08:52.760
first week start in the office. So we'll fly

01:08:52.760 --> 01:08:56.760
them to Durango and so far we've done that with

01:08:56.760 --> 01:08:59.100
everyone and that's been a really good kind of

01:08:59.100 --> 01:09:01.399
start. That doesn't mean that everyone else is

01:09:01.399 --> 01:09:03.500
going to also be at the office at the same time

01:09:03.500 --> 01:09:05.479
now that we're getting a bit bigger, but it does

01:09:05.479 --> 01:09:08.319
give them a little bit of time to meet with their

01:09:08.319 --> 01:09:12.199
manager and really get to know the company. And

01:09:12.199 --> 01:09:14.359
then the other parts of the culture is I think

01:09:14.670 --> 01:09:17.149
Something that's really key when you're doing

01:09:17.149 --> 01:09:20.029
a remote first company is to intentionally create

01:09:20.029 --> 01:09:22.609
a culture. So we've done several iterations.

01:09:22.609 --> 01:09:25.970
We work with a business coach that also helps

01:09:25.970 --> 01:09:28.289
us with this, but we really look at like, well,

01:09:28.329 --> 01:09:33.689
what are we now? What do we value? What are we

01:09:33.689 --> 01:09:36.149
currently doing that we don't like and we want

01:09:36.149 --> 01:09:38.569
to stop doing? And so we really have intentionally.

01:09:39.550 --> 01:09:41.350
Decided that and I think that is always going

01:09:41.350 --> 01:09:43.729
to be colored by who the founder is but I'd like

01:09:43.729 --> 01:09:47.449
to think as it gets bigger it moves away much

01:09:47.449 --> 01:09:50.390
less from just like who I am as a person and

01:09:50.390 --> 01:09:52.329
it becomes a bit more of like who we all are

01:09:52.329 --> 01:09:55.010
as a person. But, you know, I think that that's

01:09:55.010 --> 01:09:57.069
that's part of it was like I had some core elements

01:09:57.069 --> 01:09:59.350
that I wanted. I wanted, you know, people who

01:09:59.350 --> 01:10:01.930
are interested in leadership and analytics. And

01:10:01.930 --> 01:10:04.470
so that's like a core function of our culture.

01:10:05.050 --> 01:10:07.729
We also wanted people who were really focused

01:10:07.729 --> 01:10:10.920
on radical candor is like. the fancy term for

01:10:10.920 --> 01:10:12.920
it, but like transparency, being really open

01:10:12.920 --> 01:10:16.340
with each other. And people who really love what

01:10:16.340 --> 01:10:18.619
they do and like are interested in it. And like

01:10:18.619 --> 01:10:20.680
those are kind of the three, three of the core

01:10:20.680 --> 01:10:23.460
pillars that we built the business on. And I

01:10:23.460 --> 01:10:26.439
think we look for that when we're hiring and

01:10:26.439 --> 01:10:28.439
we also look for people who are going to be a

01:10:28.439 --> 01:10:31.810
culture add. And And then it's kind of like up

01:10:31.810 --> 01:10:34.189
to everyone else, like how that's been created

01:10:34.189 --> 01:10:36.329
since then. Like, there's a lot of things where

01:10:36.329 --> 01:10:38.510
it's like, I realize now it's like a living breathing

01:10:38.510 --> 01:10:40.710
thing. And I see things that are part of our

01:10:40.710 --> 01:10:42.890
culture now that like I never intended that I

01:10:42.890 --> 01:10:45.789
really like. And so, you know, but they're, I

01:10:45.789 --> 01:10:48.229
think, starting with that intentionality really

01:10:48.229 --> 01:10:52.109
helped. That's awesome. Yeah, me being having

01:10:52.109 --> 01:10:57.729
some. goals in mind about the culture will definitely

01:10:57.729 --> 01:11:02.829
help in one, expanding that culture and also

01:11:02.829 --> 01:11:08.350
in hiring. You want to see that people coming

01:11:08.350 --> 01:11:12.310
in are a good fit to the team that you already

01:11:12.310 --> 01:11:19.229
have. That's super cool. Yep. I know you work

01:11:19.229 --> 01:11:22.449
some in the corporate world and also you've got

01:11:22.449 --> 01:11:25.930
your own, run your own business now. What are

01:11:25.930 --> 01:11:29.010
some of the pros and cons of the corporate world?

01:11:30.270 --> 01:11:32.329
I was hoping you would ask me this question.

01:11:33.510 --> 01:11:37.590
And I talk about this a lot. And I'll also answer

01:11:37.590 --> 01:11:39.449
it in the vein of like the difference between

01:11:39.449 --> 01:11:45.390
consulting too and being in house. The pros of

01:11:45.390 --> 01:11:49.229
being in corporate is that there is a lot of

01:11:49.229 --> 01:11:53.210
upward mobility and possibility in the same company.

01:11:53.289 --> 01:11:55.149
So if you really love a company, you have a lot

01:11:55.149 --> 01:11:59.189
of different places you can go. And you can learn

01:11:59.189 --> 01:12:02.649
a lot about politics. And I know that a lot of

01:12:02.649 --> 01:12:03.930
people think about that word and they're like,

01:12:04.010 --> 01:12:06.329
oh, I don't want to do company politics. But

01:12:06.329 --> 01:12:09.850
you this is a hard truth. As you get higher up

01:12:09.850 --> 01:12:11.529
in a company, if that's really what you want

01:12:11.529 --> 01:12:14.569
to do, you have to play the game and you can

01:12:14.569 --> 01:12:17.800
play it in a honest way and really be using it

01:12:17.800 --> 01:12:19.539
correctly, but you have to be advocating for

01:12:19.539 --> 01:12:22.659
yourself and your team, etc. You can't really

01:12:22.659 --> 01:12:26.439
learn that as well when you're either a consultant

01:12:26.439 --> 01:12:28.760
or working at a really small company. It's just

01:12:28.760 --> 01:12:32.520
a different dynamic. Some other benefits is that

01:12:32.520 --> 01:12:36.000
you get a lot more budget to play with and you

01:12:36.000 --> 01:12:39.180
get to see things through to the end. If you're

01:12:39.180 --> 01:12:41.579
a consultant, you pop in, you make your recommendations

01:12:41.579 --> 01:12:44.199
and you just hope that they'll take some of them.

01:12:44.760 --> 01:12:47.859
in -house you really get to see long -term following

01:12:47.859 --> 01:12:50.239
through of projects, getting that satisfaction

01:12:50.239 --> 01:12:53.520
of seeing the results and learning from things

01:12:53.520 --> 01:12:57.279
that don't go well. Cons would be kind of the

01:12:57.279 --> 01:12:59.060
almost the flip side of some of those things.

01:13:00.029 --> 01:13:04.369
If you are someone who really wants to be entrepreneurial,

01:13:04.789 --> 01:13:07.289
it's possible to be entrepreneurial in a company,

01:13:07.310 --> 01:13:09.229
but it really depends on the company and how

01:13:09.229 --> 01:13:11.970
they receive that. As someone who has always

01:13:11.970 --> 01:13:15.149
been that way, it's been received well sometimes

01:13:15.149 --> 01:13:18.229
and sometimes not so much. And so it can be something

01:13:18.229 --> 01:13:20.090
where it causes friction if you're someone who's

01:13:20.090 --> 01:13:22.630
really innovative in your ideas. And over time,

01:13:22.670 --> 01:13:24.909
that can get stifled is what happens. If you

01:13:24.909 --> 01:13:27.369
do choose to stay in corporate, you can start

01:13:27.369 --> 01:13:30.720
to get a little bit comfortable. The other thing

01:13:30.720 --> 01:13:32.600
with corporate is if you stay in the same company

01:13:32.600 --> 01:13:34.659
for a long time, you don't get to see as many

01:13:34.659 --> 01:13:38.539
different ways of doing things. If you're a consultant

01:13:38.539 --> 01:13:41.140
or if you maybe change or switch between corporate

01:13:41.140 --> 01:13:43.239
and smaller companies, you're going to see a

01:13:43.239 --> 01:13:45.880
lot of different personalities and different

01:13:45.880 --> 01:13:49.399
industries. completely different ways of operating.

01:13:49.800 --> 01:13:52.039
And all of that can add color to your experience

01:13:52.039 --> 01:13:55.260
moving forward. So I can probably go on, but

01:13:55.260 --> 01:13:56.899
I think those are some of the really core elements

01:13:56.899 --> 01:13:59.319
between the differences. And what I've found

01:13:59.319 --> 01:14:01.600
is I really like working as a consultant for

01:14:01.600 --> 01:14:04.500
corporate companies because I still get to enjoy

01:14:04.500 --> 01:14:07.460
a lot of those pros that I really like. But then

01:14:07.460 --> 01:14:10.840
also people are a little more accepting of whatever

01:14:10.840 --> 01:14:12.619
wild ideas I'm coming in with, because they're

01:14:12.619 --> 01:14:15.479
like, ah, she's just a consultant. So that's

01:14:15.479 --> 01:14:17.260
at least where I found my sweet spot, but I think

01:14:17.260 --> 01:14:20.899
it really depends on the person. Yeah, it really

01:14:20.899 --> 01:14:23.180
depends on the person and the company that you're

01:14:23.180 --> 01:14:25.720
in, like you said. You touched on a couple of

01:14:25.720 --> 01:14:29.819
things that were interesting to me. I mean, the

01:14:29.819 --> 01:14:32.279
politics can be either pro or a con, depending

01:14:32.279 --> 01:14:36.239
on how you play it, or also how you don't like

01:14:36.239 --> 01:14:43.289
it. The way that I think about it is just dealing

01:14:43.289 --> 01:14:47.409
with people. And in the corporate world, you

01:14:47.409 --> 01:14:51.189
deal with the same people for a long period of

01:14:51.189 --> 01:14:54.529
time. So you've got to have that long -term mindset.

01:14:54.989 --> 01:14:59.710
Hey, if you have a couple of projects, you lose

01:14:59.710 --> 01:15:03.659
some, you win some, but... when you're in the

01:15:03.659 --> 01:15:07.279
consulting world, it's more about the customer's

01:15:07.279 --> 01:15:09.960
needs first, and you've got to make sure that

01:15:09.960 --> 01:15:13.159
they're comfortable with everything. And also,

01:15:13.600 --> 01:15:17.000
you touched on people who are entrepreneurial,

01:15:17.819 --> 01:15:21.579
how they manage in the corporate world. One of

01:15:21.579 --> 01:15:24.039
the things is, like you said, depends on the

01:15:24.039 --> 01:15:28.239
company. One of the things is smaller companies

01:15:28.239 --> 01:15:33.109
are usually open to give you more flexibility

01:15:33.109 --> 01:15:38.109
versus big tech or really large companies. Most

01:15:38.109 --> 01:15:40.529
of the times, they fit you in a box and then

01:15:40.529 --> 01:15:44.689
your role is defined. You do it and unless you

01:15:44.689 --> 01:15:48.189
move to a different role, you can't do too much

01:15:48.189 --> 01:15:57.979
about it. Following up on our previous question,

01:15:58.880 --> 01:16:02.199
what can startups and corporations learn from

01:16:02.199 --> 01:16:05.460
each other? Another one that I think is really

01:16:05.460 --> 01:16:10.420
fun to think about, I think that startups can

01:16:10.420 --> 01:16:13.319
learn how to stick to a strategy set strategy

01:16:13.319 --> 01:16:16.000
and stick to a strategy from corporate companies.

01:16:17.380 --> 01:16:20.020
I think that that's a key element of startups

01:16:20.020 --> 01:16:22.699
that are more successful is if they're able to

01:16:22.840 --> 01:16:26.579
get at least a bit close to the level at which

01:16:26.579 --> 01:16:28.800
some of these really big corporations are able

01:16:28.800 --> 01:16:32.279
to choose a strategy and really structure everything

01:16:32.279 --> 01:16:34.579
they do based on it. That's something that startups

01:16:34.579 --> 01:16:37.600
can certainly learn from corporate. I think that

01:16:37.600 --> 01:16:41.399
corporate can learn from startups that even though

01:16:41.399 --> 01:16:43.380
they have this whole established way of doing

01:16:43.380 --> 01:16:46.300
things. sometimes you can blow that up. And you

01:16:46.300 --> 01:16:48.520
should, and you should listen to people who have

01:16:48.520 --> 01:16:52.399
unique ideas and challenge the status quo. And

01:16:52.399 --> 01:16:54.819
that's where you're going to be able to innovate

01:16:54.819 --> 01:16:58.699
compared to competitors and be a lot more flexible.

01:16:59.039 --> 01:17:02.840
And that's one thing that you see with corporate.

01:17:03.159 --> 01:17:06.319
Like right now, I mean, we even see it in the

01:17:06.319 --> 01:17:08.859
kind of closing and Bed Bath and Beyond struggling,

01:17:08.899 --> 01:17:14.220
for example. Companies that fail to get excited

01:17:14.220 --> 01:17:17.100
about whatever is happening in the world and

01:17:17.100 --> 01:17:19.800
in the business world and like play around with

01:17:19.800 --> 01:17:23.500
it and try new things even if they're still executing

01:17:23.500 --> 01:17:26.979
within their strategy are the ones that succeed

01:17:26.979 --> 01:17:29.880
and ones that really don't adapt fast enough

01:17:29.880 --> 01:17:33.020
are going to fail and startups know that. Startups

01:17:33.020 --> 01:17:35.039
always know that they need to be agile and that's

01:17:35.039 --> 01:17:37.619
how they out compete. corporate companies sometimes

01:17:37.619 --> 01:17:40.279
too, is that they're more willing to like go

01:17:40.279 --> 01:17:42.279
bang into something new and they can pivot really

01:17:42.279 --> 01:17:44.380
quickly. So I think that those are some of the

01:17:44.380 --> 01:17:47.960
core areas that can really help. And I think

01:17:47.960 --> 01:17:49.640
the other part that you kind of touched on a

01:17:49.640 --> 01:17:53.800
bit that I'd like to pull into is even though

01:17:53.800 --> 01:17:57.020
corporate wants to have their boxes for every

01:17:57.020 --> 01:18:01.109
one of their employees, Allowing for some like,

01:18:01.229 --> 01:18:03.470
and I forget the term, maybe we can post it in

01:18:03.470 --> 01:18:06.390
your podcast notes if you remember it, but it's

01:18:06.390 --> 01:18:09.109
like when you're an entrepreneur within a company.

01:18:09.869 --> 01:18:13.369
And I think that you should encourage that. Intrapreneur,

01:18:13.369 --> 01:18:16.250
you're right. Yes, thank you. Yeah, so you should

01:18:16.250 --> 01:18:19.609
encourage some intrapreneurial activities, whether

01:18:19.609 --> 01:18:21.970
you have innovation functional groups and things

01:18:21.970 --> 01:18:24.789
like that. That's something that you don't want

01:18:24.789 --> 01:18:29.920
to lose. the fun and creativity of that, even

01:18:29.920 --> 01:18:32.340
if you need a lot of structure in your company.

01:18:33.420 --> 01:18:36.960
Yeah, no, that's a really good answer. Yeah,

01:18:38.420 --> 01:18:42.039
both sides can learn a lot from each other. And

01:18:42.039 --> 01:18:46.180
it's depending on where you are as an individual,

01:18:46.340 --> 01:18:49.979
you could choose one or the other or switch between

01:18:49.979 --> 01:18:56.560
the two. Awesome. Let's dive into some of the

01:18:56.560 --> 01:19:00.399
challenges in general. What are some of the challenges

01:19:00.399 --> 01:19:04.859
you face in the digital marketing world? We know

01:19:04.859 --> 01:19:09.520
we talked about GA4 and all that. So what are

01:19:09.520 --> 01:19:11.699
some of the other challenges or if you want to

01:19:11.699 --> 01:19:16.000
expand on GA4, that's cool too. Oh, well, it's

01:19:16.000 --> 01:19:18.039
kind of more focused on digital marketing since

01:19:18.039 --> 01:19:21.800
that's how you phrased it too. I think some other

01:19:21.800 --> 01:19:30.020
challenges are How to properly show impact and

01:19:30.020 --> 01:19:33.079
understand what's working and what's not is like

01:19:33.079 --> 01:19:35.960
one of the biggest pain points of marketing ever.

01:19:36.460 --> 01:19:38.420
It's not specific just to digital marketing,

01:19:38.640 --> 01:19:40.739
but with digital marketing, it actually makes

01:19:40.739 --> 01:19:42.819
it almost harder and more frustrating in some

01:19:42.819 --> 01:19:46.550
ways because we think that we know. a lot more

01:19:46.550 --> 01:19:49.609
about how the uplift is and what it's doing for

01:19:49.609 --> 01:19:53.090
us, but it can really misguide us. So that's

01:19:53.090 --> 01:19:56.829
a huge pain point is that with attribution, which

01:19:56.829 --> 01:19:59.970
I still think is important and we need to figure

01:19:59.970 --> 01:20:02.729
out how we do it, but the way that attribution

01:20:02.729 --> 01:20:06.779
has been It's so hard to discern. We're rewarding

01:20:06.779 --> 01:20:09.880
digital marketing over a lot of other things.

01:20:11.039 --> 01:20:13.640
So digital marketing, we just take the whole

01:20:13.640 --> 01:20:16.520
pie and we say that digital marketing has gotten

01:20:16.520 --> 01:20:19.119
all these slices for where all of our traffic

01:20:19.119 --> 01:20:21.420
came from. But if you think about the big picture,

01:20:21.479 --> 01:20:25.140
that's not true at all. There's returning customers.

01:20:25.739 --> 01:20:28.329
There's like your brand presence, there's all

01:20:28.329 --> 01:20:29.930
the traditional marketing that you're doing,

01:20:30.170 --> 01:20:32.449
you're not actually factoring for any of that

01:20:32.449 --> 01:20:36.029
when you look at that pie chart in Google Analytics.

01:20:37.189 --> 01:20:40.529
So that's one that is really difficult and it's

01:20:40.529 --> 01:20:43.890
really difficult to properly explain to marketers.

01:20:45.489 --> 01:20:47.390
One of them you touched on at the beginning of

01:20:47.390 --> 01:20:52.710
this was about deprecation of cookies and additional

01:20:52.710 --> 01:20:55.600
privacy. restrictions and laws that are going

01:20:55.600 --> 01:20:58.500
into place. This is something that is difficult

01:20:58.500 --> 01:21:02.760
because we have become accustomed in digital

01:21:02.760 --> 01:21:06.100
marketing to operating in a certain way. And

01:21:06.100 --> 01:21:08.420
the industry is really flexible, so I think that

01:21:08.420 --> 01:21:12.220
they always will adjust. But now what it does

01:21:12.220 --> 01:21:14.560
is forces companies to be more strategic about

01:21:14.560 --> 01:21:18.359
how they collect data. They can't just collect

01:21:18.359 --> 01:21:20.880
whatever they want and use it however they want

01:21:20.880 --> 01:21:23.619
and store it however they want anymore. Now there's

01:21:23.619 --> 01:21:26.960
laws holding them accountable. And then also

01:21:26.960 --> 01:21:30.220
they have to be a lot more intentional about

01:21:30.220 --> 01:21:32.579
gathering the data that they want. It's not just

01:21:32.579 --> 01:21:34.340
going to fall into your lap. If you really want

01:21:34.340 --> 01:21:37.819
to know about your customers, you're going to

01:21:37.819 --> 01:21:40.279
have to ask them. And convincing companies to

01:21:40.279 --> 01:21:42.020
do that is another one that's difficult where

01:21:42.020 --> 01:21:44.859
it's like, hey, if you want zero party, first

01:21:44.859 --> 01:21:47.590
party data, this is what has to happen and it's

01:21:47.590 --> 01:21:49.510
a little more complicated than it has been in

01:21:49.510 --> 01:21:53.430
the last five or so years. Yeah, it can definitely

01:21:53.430 --> 01:21:58.310
get a little more infrastructure heavy. You want

01:21:58.310 --> 01:22:05.250
to be able to allow the infrastructure to, say,

01:22:05.670 --> 01:22:11.470
delete data upon request or build some more sophisticated

01:22:11.470 --> 01:22:15.069
tagging mechanisms so that you understand where

01:22:15.159 --> 01:22:19.460
what part of your funnel you're tracking and

01:22:19.460 --> 01:22:23.380
all that. So it definitely is evolving as we

01:22:23.380 --> 01:22:28.199
speak. So it's an interesting time to be in the

01:22:28.199 --> 01:22:33.020
industry. What are some of the challenges in

01:22:33.020 --> 01:22:39.020
the consulting world? Oh, there's a lot I think.

01:22:41.049 --> 01:22:43.430
Finding clients is always going to be a difficult

01:22:43.430 --> 01:22:46.369
thing. Finding the right clients that like what

01:22:46.369 --> 01:22:49.229
you do is like when you first start out in consulting

01:22:49.229 --> 01:22:52.130
and first establish yourself as an agency, it's

01:22:52.130 --> 01:22:55.369
like, it can feel like an overwhelming task.

01:22:55.930 --> 01:22:57.529
And, you know, I think, so that's like kind of

01:22:57.529 --> 01:22:59.710
more of a tactical one, but I think another one

01:22:59.710 --> 01:23:01.310
that's really relevant to what we're talking

01:23:01.310 --> 01:23:06.899
about is like balancing recommendation. understanding

01:23:06.899 --> 01:23:10.640
that you don't have the wealth of knowledge that

01:23:10.640 --> 01:23:15.100
someone in -house does. There's pros and cons

01:23:15.100 --> 01:23:18.260
to that. Part of it is that you can see things

01:23:18.260 --> 01:23:20.760
they can't, and on the other side, you don't

01:23:20.760 --> 01:23:24.199
know all of the reasons why they are the way

01:23:24.199 --> 01:23:28.539
they are. Being a good consultant and really

01:23:28.640 --> 01:23:31.739
supporting your clients is actually really fun

01:23:31.739 --> 01:23:34.659
for me. I don't know if everyone would agree,

01:23:34.739 --> 01:23:37.060
but it's really fun, but it's also very challenging

01:23:37.060 --> 01:23:39.060
because you're always trying to be like, well,

01:23:39.060 --> 01:23:41.319
I want to recommend things, but I want to make

01:23:41.319 --> 01:23:43.420
sure that I have enough context to do so. And

01:23:43.420 --> 01:23:46.960
then you also have to let go of any ownership

01:23:46.960 --> 01:23:50.020
of what happens next because you might not be

01:23:50.020 --> 01:23:52.239
there. And even if you're going to be there,

01:23:52.479 --> 01:23:54.300
like sometimes we're running data strategy projects

01:23:54.300 --> 01:23:57.380
or implementation, we can still only do so much.

01:23:57.579 --> 01:24:01.640
We can create the project plan, we can highlight

01:24:01.640 --> 01:24:04.420
any pitfalls, we can really try and drive things

01:24:04.420 --> 01:24:07.600
forward, but in the end it's the company's kind

01:24:07.600 --> 01:24:10.689
of responsibility and choice to like... how they're

01:24:10.689 --> 01:24:12.529
going to do that. So I think that's something

01:24:12.529 --> 01:24:14.810
that's always really challenging is to not get

01:24:14.810 --> 01:24:17.869
overly invested or feel frustrated when someone

01:24:17.869 --> 01:24:20.329
isn't taking your recommendation. But I also

01:24:20.329 --> 01:24:24.170
have been told by someone else who's really experienced

01:24:24.170 --> 01:24:25.850
with consulting that you definitely don't want

01:24:25.850 --> 01:24:28.050
them to follow 100 % of your recommendations

01:24:28.050 --> 01:24:30.189
because then you're 100 % responsible for what

01:24:30.189 --> 01:24:34.270
happens. So it's just like hoping that they take

01:24:34.270 --> 01:24:36.010
the pieces that are going to be most useful to

01:24:36.010 --> 01:24:37.689
them and that'll make a difference for them.

01:24:38.279 --> 01:24:42.060
Yeah, no you're just on a really some some really

01:24:42.060 --> 01:24:46.640
important points there Consulting versus in -house

01:24:46.640 --> 01:24:51.659
you definitely As a consultant you don't want

01:24:51.659 --> 01:24:54.899
to be Arrogant going in saying hey, I'm a consultant.

01:24:54.960 --> 01:24:57.640
I know what I'm talking about. Yes, you do know

01:24:57.640 --> 01:25:01.300
your area of expertise, but you don't have a

01:25:01.300 --> 01:25:04.239
lot of times you don't have the contextual knowledge

01:25:04.239 --> 01:25:09.199
on why Things are the way they are And also,

01:25:10.119 --> 01:25:14.500
we talked on this earlier in the podcast, but

01:25:14.500 --> 01:25:21.319
a lot of times the data does not give you the

01:25:21.319 --> 01:25:24.520
entire picture. There's a lot of history and

01:25:24.520 --> 01:25:28.859
stories behind why operations are operating the

01:25:28.859 --> 01:25:35.739
way they are. So it's really collaborating with

01:25:35.739 --> 01:25:39.449
clients. to understand as much context as you

01:25:39.449 --> 01:25:42.489
can is super helpful. And also letting go of

01:25:42.489 --> 01:25:46.850
ownership. That's I love that because if you're

01:25:46.850 --> 01:25:51.949
an in -house employee of a company, then you

01:25:51.949 --> 01:25:54.750
can own a certain piece of the business for a

01:25:54.750 --> 01:25:58.850
long time. But as a consultant, you build it

01:25:58.850 --> 01:26:02.489
and you hand it over and you're more like a parent

01:26:02.489 --> 01:26:08.319
needs to let go. Yeah, really cool. What are

01:26:08.319 --> 01:26:12.300
some of the criticisms you have from the marketing

01:26:12.300 --> 01:26:15.939
side about data teams and what can we do to improve?

01:26:16.699 --> 01:26:20.800
We can take a jab or two, give it to us. Yeah,

01:26:20.960 --> 01:26:23.000
I guess my first jab is that they're not big

01:26:23.000 --> 01:26:27.149
enough often. And I guess maybe that's a jab

01:26:27.149 --> 01:26:29.729
at leadership instead. So I'll have some specific

01:26:29.729 --> 01:26:32.689
jobs at data teams as well. But like we spend

01:26:32.689 --> 01:26:34.689
too much money on tools and not enough money

01:26:34.689 --> 01:26:38.890
on smart people. I think another job is that

01:26:38.890 --> 01:26:44.310
we sometimes focus too much on super technical

01:26:44.310 --> 01:26:47.390
data scientist type people. And that's really

01:26:47.390 --> 01:26:51.189
a trend right now is that data scientists sound

01:26:51.189 --> 01:26:54.810
sexier. than an analyst, but they're both valuable.

01:26:55.010 --> 01:26:57.510
And if you're a smaller company and you had to

01:26:57.510 --> 01:27:01.310
pick one, I would pick an analyst any day because

01:27:01.310 --> 01:27:05.550
you need people who understand both parts or

01:27:05.550 --> 01:27:08.590
multiple parts of the business. You need someone

01:27:08.590 --> 01:27:11.989
who can speak the language of IT or really understand

01:27:11.989 --> 01:27:13.729
a marketer when they're talking about a brand

01:27:13.729 --> 01:27:17.350
concept. Because like what you just said, there's

01:27:17.350 --> 01:27:19.789
so much additional context that needs to be layered

01:27:19.789 --> 01:27:24.699
with the data. And so I guess my next like punch

01:27:24.699 --> 01:27:30.779
would be that oftentimes on larger organizations,

01:27:30.840 --> 01:27:35.720
I don't think that I see enough leadership roles

01:27:35.720 --> 01:27:38.560
with data. Like I think that a chief data officer

01:27:38.560 --> 01:27:40.420
needs to become just kind of like a standard

01:27:40.420 --> 01:27:43.619
part of the C -suite. And I don't see that. And

01:27:43.619 --> 01:27:46.159
I also don't see that many, you know, director

01:27:46.159 --> 01:27:50.319
level and above people advocating for it. I don't

01:27:50.319 --> 01:27:52.640
see I don't see enough hunger in our current

01:27:52.640 --> 01:27:55.420
data teams to do that. I think that we sometimes,

01:27:55.539 --> 01:27:58.739
and here's my kick now, is that we act like order

01:27:58.739 --> 01:28:01.039
-takers, where we should be the ones driving

01:28:01.039 --> 01:28:04.840
the conversations instead. I think those are

01:28:04.840 --> 01:28:06.720
some of the ones. I'm sure I could keep going,

01:28:06.779 --> 01:28:09.020
but I think those are some of the most important

01:28:09.020 --> 01:28:11.859
ones that are on my mind today. Awesome. Yeah,

01:28:12.119 --> 01:28:16.300
definitely data teams need to be more proactive.

01:28:17.180 --> 01:28:22.619
And because we see a lot of... data from a lot

01:28:22.619 --> 01:28:26.819
of different angles. We have different perspectives

01:28:26.819 --> 01:28:30.859
and somebody that's in operations and in the

01:28:30.859 --> 01:28:36.640
business and who has developed certain blind

01:28:36.640 --> 01:28:40.640
spots for those data teams can definitely, like

01:28:40.640 --> 01:28:46.119
you said, data leadership needs to become more

01:28:46.119 --> 01:28:52.800
obvious. chief data officers and a VP, senior

01:28:52.800 --> 01:28:55.319
vice president, executives at the data level

01:28:55.319 --> 01:29:03.380
who can tell a good story and be a bridge between

01:29:03.380 --> 01:29:06.239
the business and the backend teams. It's really

01:29:06.239 --> 01:29:11.359
important. You made a great point on companies

01:29:11.359 --> 01:29:14.779
hiring data science folks because it's way more

01:29:14.779 --> 01:29:21.699
sexier. And yeah, that's so true. Because we

01:29:21.699 --> 01:29:24.600
talked about data maturity. So you have to get

01:29:24.600 --> 01:29:28.520
to that level of data maturity to start hiring

01:29:28.520 --> 01:29:33.060
data scientists. So you definitely need to have

01:29:33.060 --> 01:29:36.239
analysts first. And then if you want to build

01:29:36.239 --> 01:29:39.600
out your infrastructure, go hire data engineers

01:29:39.600 --> 01:29:43.180
first to make sure that you build out the infrastructure.

01:29:44.579 --> 01:29:49.399
Because 80 % per I wouldn't say 80%, but 60 to

01:29:49.399 --> 01:29:53.319
70 % of the job is to get data from different

01:29:53.319 --> 01:29:57.060
places, massage it in a way that they can work

01:29:57.060 --> 01:30:00.720
well together, and then present it in a way for

01:30:00.720 --> 01:30:04.420
analysis and model building. So that's all that

01:30:04.420 --> 01:30:08.760
the data engineers do. And then you hire a data

01:30:08.760 --> 01:30:12.479
scientist or machine learning engineers. So it's

01:30:12.479 --> 01:30:16.420
really about data maturity, but you're spot on

01:30:16.420 --> 01:30:19.899
when you say, you know, a new executive comes

01:30:19.899 --> 01:30:22.020
in and he's like, hey, why don't we hire data

01:30:22.020 --> 01:30:25.479
scientists, you know, PhDs and all that, but

01:30:25.479 --> 01:30:30.439
they don't have what they need to, you know,

01:30:30.600 --> 01:30:33.060
be good at their jobs. You're setting them for

01:30:33.060 --> 01:30:38.380
failure. Yeah. And I think also, we misunderstand,

01:30:38.560 --> 01:30:40.359
maybe this goes back to data literacy a bit,

01:30:40.500 --> 01:30:42.439
the difference between a data scientist and an

01:30:42.439 --> 01:30:44.840
analyst. And I think sometimes they hire a data

01:30:44.840 --> 01:30:46.739
scientist and then they expect them to do an

01:30:46.739 --> 01:30:50.720
analyst's job. And this PhD is like, look, that's

01:30:50.720 --> 01:30:52.500
not what I do. And they're like, well, why won't

01:30:52.500 --> 01:30:56.460
you just pull these metrics for me? And so yeah,

01:30:56.640 --> 01:30:59.869
it's another one where. I don't know if it's

01:30:59.869 --> 01:31:03.350
maybe like a branding issue of us, of our different

01:31:03.350 --> 01:31:05.869
roles or something like that. But a lot of times

01:31:05.869 --> 01:31:08.569
I've seen like a job description where they're

01:31:08.569 --> 01:31:10.789
like, you know, the title says data scientist

01:31:10.789 --> 01:31:12.430
and I read it. I'm like, you just want an analyst.

01:31:13.489 --> 01:31:15.310
Or, you know, they also, or sometimes they even

01:31:15.310 --> 01:31:16.909
have the description correct. And then that's

01:31:16.909 --> 01:31:19.829
not what they ask the person to do at all. Exactly.

01:31:20.770 --> 01:31:26.380
Exactly. Speaking of, I mean, kind of of moving

01:31:26.380 --> 01:31:32.479
towards digital marketing again. Digital data

01:31:32.479 --> 01:31:36.380
can never be perfect. We talk about regulations

01:31:36.380 --> 01:31:40.279
coming in place and all that. Even without that,

01:31:40.300 --> 01:31:45.779
there's always gaps. In a lot of people's minds,

01:31:46.979 --> 01:31:49.739
digital marketing data is infinite and you have

01:31:49.739 --> 01:31:53.060
everything you need, but it's not the case. There's

01:31:53.060 --> 01:31:56.720
a lot of gaps. I could pop on the website and

01:31:56.720 --> 01:32:00.439
then leave the session and go on my mobile phone

01:32:00.439 --> 01:32:03.119
and log in. So it's really hard to tie those

01:32:03.119 --> 01:32:06.760
two together and understand that I'm the same

01:32:06.760 --> 01:32:10.420
person. And also it usually is directionally

01:32:10.420 --> 01:32:13.020
correct. So how do you handle those situations?

01:32:13.579 --> 01:32:16.300
How do you explain those to your clients and

01:32:16.300 --> 01:32:20.260
executives? That's one where that like communication

01:32:20.260 --> 01:32:24.310
and data literacy comes in again. It's one of

01:32:24.310 --> 01:32:26.550
the first initiatives I almost always pick up

01:32:26.550 --> 01:32:28.869
with a new company is saying, all right, well,

01:32:29.029 --> 01:32:32.149
let's for real evaluate what our discrepancies

01:32:32.149 --> 01:32:35.329
are. Let me show you what the industry averages

01:32:35.329 --> 01:32:37.670
are too. And let's have some common language

01:32:37.670 --> 01:32:41.090
about why that's actually OK. Because the average

01:32:41.090 --> 01:32:43.470
discrepancy between, for example, Google Analytics

01:32:43.470 --> 01:32:46.390
transaction revenue data and your back end is

01:32:46.390 --> 01:32:50.069
usually like 20%. That's like a normal number.

01:32:51.079 --> 01:32:53.760
sometimes people hear those numbers, or also

01:32:53.760 --> 01:32:57.119
this is where these conversations or issues and

01:32:57.119 --> 01:32:59.640
conflict comes up as someone goes, well, I can't

01:32:59.640 --> 01:33:02.239
trust Google Analytics data because it's 20 %

01:33:02.239 --> 01:33:04.600
different than the financial data. And the financial

01:33:04.600 --> 01:33:06.880
analysts are like, how dare you use that data

01:33:06.880 --> 01:33:10.140
to interpret anything? It's not accurate. But

01:33:10.140 --> 01:33:14.380
the thing is, it's usually consistently 20 %

01:33:14.380 --> 01:33:17.149
different. So you actually still have a directional

01:33:17.149 --> 01:33:19.229
change. If your revenue goes up, it's going to

01:33:19.229 --> 01:33:21.310
go up in both and it's going to be pretty proportional.

01:33:22.750 --> 01:33:26.470
So it's like a long, longer term game. But the

01:33:26.470 --> 01:33:28.510
first thing that I usually do is create like

01:33:28.510 --> 01:33:31.850
a one sheet of like, here's where we're at. Here's

01:33:31.850 --> 01:33:33.449
how we're going to keep being accountable about

01:33:33.449 --> 01:33:35.380
this. We're going to check it every quarter or

01:33:35.380 --> 01:33:37.880
something like that. And here's what it really

01:33:37.880 --> 01:33:40.939
means. And then it continues to be like little

01:33:40.939 --> 01:33:43.579
pieces that I build on as I present to a client.

01:33:43.640 --> 01:33:47.899
I'll explain that often and keep going back to

01:33:47.899 --> 01:33:50.500
the point like, this data is valuable for these

01:33:50.500 --> 01:33:53.300
reasons. Here's why we should use it. And then

01:33:53.300 --> 01:33:56.739
I'm also someone who really understands the other

01:33:56.739 --> 01:33:59.199
side. I really love working with finance teams,

01:33:59.659 --> 01:34:03.850
really love working with more advanced. data

01:34:03.850 --> 01:34:06.229
engineer teams that are really starting to do

01:34:06.229 --> 01:34:08.289
some more advanced things that get you a little

01:34:08.289 --> 01:34:12.470
more closer to accuracy. So I totally understand

01:34:12.470 --> 01:34:14.890
where they're coming from and why they're like

01:34:14.890 --> 01:34:19.430
that. like to also honor when an analysis actually

01:34:19.430 --> 01:34:21.750
does need to be done in finance data, for example,

01:34:22.090 --> 01:34:25.449
because there are those differences. And so starting

01:34:25.449 --> 01:34:27.850
to educate on that and make sure that we are

01:34:27.850 --> 01:34:30.630
playing on the same team really helps because

01:34:30.630 --> 01:34:32.710
establishing that agreement really helps. Sitting

01:34:32.710 --> 01:34:34.470
down with finance and being like, look, I'm not

01:34:34.470 --> 01:34:36.449
trying to step on your toes. Here's why I do

01:34:36.449 --> 01:34:38.130
what I do. And here's where I'm going to focus

01:34:38.130 --> 01:34:42.050
on. But if it goes into this field, financial

01:34:42.050 --> 01:34:44.130
forecasting for the whole company or calculating

01:34:44.130 --> 01:34:46.829
profitability or cogs or something like that.

01:34:46.829 --> 01:34:48.710
I'm going to put that over the fence to you,

01:34:48.729 --> 01:34:52.069
but we can both benefit from each other. Yeah,

01:34:52.149 --> 01:34:57.189
absolutely. I think you touched on it. It's really

01:34:57.189 --> 01:35:02.590
about setting up expectations. Finance and accounting

01:35:02.590 --> 01:35:04.890
data is super accurate and that's what they are

01:35:04.890 --> 01:35:09.439
paid to do, be accurate. in the digital world

01:35:09.439 --> 01:35:13.060
and a lot of data, you know, customer care or

01:35:13.060 --> 01:35:21.420
email marketing, all these data is mostly directional

01:35:21.420 --> 01:35:25.340
and as long as you have those deviations and

01:35:25.340 --> 01:35:28.119
expectations set, like you said, I think as long

01:35:28.119 --> 01:35:31.699
as you're measuring the same set of data the

01:35:31.699 --> 01:35:35.220
same way over time, you wouldn't know if there's

01:35:35.220 --> 01:35:38.399
a directional change. or a direction of shift.

01:35:38.680 --> 01:35:41.239
And that's when you can make decisions based

01:35:41.239 --> 01:35:47.920
off of. And I think I heard somebody say, it's

01:35:47.920 --> 01:35:50.960
all OK as long as you're consistently wrong.

01:35:52.060 --> 01:35:55.460
Yeah. Yes. Yeah. I really like that. And it's

01:35:55.460 --> 01:35:59.899
very true. And I think that sometimes discrepancies,

01:36:00.140 --> 01:36:03.100
I even have a blog post about it, can be like,

01:36:03.369 --> 01:36:07.130
a paralyzer for actually doing things and an

01:36:07.130 --> 01:36:10.170
excuse for why you're not doing things. And I

01:36:10.170 --> 01:36:13.270
would rather have data that I think is good enough.

01:36:13.649 --> 01:36:16.329
And good enough is exactly that. It's good enough.

01:36:17.689 --> 01:36:20.170
And be making a bunch of decisions on it and

01:36:20.170 --> 01:36:24.130
looking at how that impacts things and driving

01:36:24.130 --> 01:36:27.170
things forward and really having good conversations

01:36:27.170 --> 01:36:29.750
too about, hey, what does this mean? How do we

01:36:29.750 --> 01:36:33.970
want to approach this? versus choosing to do

01:36:33.970 --> 01:36:37.210
nothing or something I've also seen companies

01:36:37.210 --> 01:36:41.149
do is go so conservative that they put all the

01:36:41.149 --> 01:36:45.210
power into finance or into an engineering team,

01:36:45.550 --> 01:36:48.829
where then it takes them so long to do an analysis

01:36:48.829 --> 01:36:52.210
that it's useless anyways. So it's really like

01:36:52.210 --> 01:36:55.590
the business world moves fast and we have to

01:36:55.590 --> 01:36:58.779
accept that we... know something, but we don't

01:36:58.779 --> 01:37:02.100
know everything. And it's better to be imperfect

01:37:02.100 --> 01:37:05.439
in what you're using than to just be guessing

01:37:05.439 --> 01:37:08.720
completely without data. Yeah, it's really a

01:37:08.720 --> 01:37:12.560
balance. Like we talked about earlier about being

01:37:12.560 --> 01:37:15.880
scientific and knowing exactly what you're doing

01:37:15.880 --> 01:37:20.159
in terms of your hypothesis and testing. But

01:37:20.159 --> 01:37:25.260
also on the flip side, you don't want to be You

01:37:25.260 --> 01:37:27.899
don't want to wait too long. You don't want to

01:37:27.899 --> 01:37:30.920
wait six months to get the result of a test.

01:37:31.739 --> 01:37:34.140
Like you said, the business world moves so fast

01:37:34.140 --> 01:37:38.229
that it becomes stale at that point. That's so

01:37:38.229 --> 01:37:40.529
true. I'm glad that you called that out because

01:37:40.529 --> 01:37:42.149
some people might listen to this and be like,

01:37:42.250 --> 01:37:45.409
oh my God, she just totally contradicted herself.

01:37:45.630 --> 01:37:48.670
But yeah, it's really, and I think that the way

01:37:48.670 --> 01:37:51.170
you do that is keep the business focus in mind.

01:37:51.550 --> 01:37:54.810
It costs me practically nothing to pre -test

01:37:54.810 --> 01:37:58.329
tests and be strict there. It's so easy. It doesn't

01:37:58.329 --> 01:38:01.289
add operational complexity or anything like that.

01:38:01.789 --> 01:38:04.630
But on the other hand, running all of our analysis

01:38:04.630 --> 01:38:06.710
through a database that only finance has and

01:38:06.710 --> 01:38:08.869
you have to ask for it instead of querying it

01:38:08.869 --> 01:38:12.609
yourself or whatever the issue is, does add operational

01:38:12.609 --> 01:38:14.710
complexity and makes it harder for you to do

01:38:14.710 --> 01:38:17.390
your job. So always thinking about where's that

01:38:17.390 --> 01:38:21.750
balance and what level of control needs to be

01:38:21.750 --> 01:38:24.409
there because it also depends a bit on where

01:38:24.409 --> 01:38:26.810
the data is going, what you're using it for.

01:38:28.399 --> 01:38:31.359
and those two things do coexist because you're

01:38:31.359 --> 01:38:33.720
always wanting to make it as accurate as possible,

01:38:34.479 --> 01:38:37.119
as make sure your tracking is as clean as possible

01:38:37.119 --> 01:38:39.819
and the data is clean as possible, but to a point.

01:38:40.100 --> 01:38:42.439
There's a point of diminishing returns. You want

01:38:42.439 --> 01:38:44.539
to get to that good enough point and then just

01:38:44.539 --> 01:38:47.039
let it go and check it periodically, maintain

01:38:47.039 --> 01:38:49.979
data quality, and then really push hard into

01:38:49.979 --> 01:38:54.920
using that data. Absolutely. It always comes

01:38:54.920 --> 01:38:58.979
down to balance. The business world is so nuanced

01:38:58.979 --> 01:39:06.119
and there are a lot of times opposing interests.

01:39:06.180 --> 01:39:11.560
So you just have to keep a balance so you don't

01:39:11.560 --> 01:39:16.939
miss out on opportunities as well as not take

01:39:16.939 --> 01:39:22.680
drastic decisions based off of wrong or small

01:39:22.680 --> 01:39:29.149
set of data. in significant data. It's been a

01:39:29.149 --> 01:39:33.270
long week, Mary Beth. It's only Wednesday. I

01:39:33.270 --> 01:39:39.289
know. I agree. Cool. Cool. Cool. All right. Let's

01:39:39.289 --> 01:39:47.170
talk AI and Chan GPT. We're in a revolution right

01:39:47.170 --> 01:39:53.489
now. I am invested, but I have a hard time keeping

01:39:53.489 --> 01:39:56.380
up with everything that's going on. I just hope

01:39:56.380 --> 01:40:01.659
that there's a lot of value to AI in this thing.

01:40:02.979 --> 01:40:08.319
I don't like all the hype around it. You've got

01:40:08.319 --> 01:40:12.579
to temper that down and make sure you provide

01:40:12.579 --> 01:40:16.640
value. Like you said, have customers in mind

01:40:16.640 --> 01:40:20.560
and not just AI for AI. What are your thoughts

01:40:20.560 --> 01:40:23.619
on current advancements in AI and chat GPD and

01:40:23.619 --> 01:40:26.859
everything? Well, I have to say I laughed because

01:40:26.859 --> 01:40:29.100
I'm in this group called Restaurant Technology

01:40:29.100 --> 01:40:30.899
Network where we workshop some things and we

01:40:30.899 --> 01:40:33.319
have this game in most of the working sessions

01:40:33.319 --> 01:40:36.000
of like, how long can we go before someone says

01:40:36.000 --> 01:40:39.060
something about chat GPT? So it took us quite

01:40:39.060 --> 01:40:42.079
a while before I got there. But yeah, you're

01:40:42.079 --> 01:40:43.939
going to love I just actually finished writing

01:40:43.939 --> 01:40:45.859
another article that's going to go on our website

01:40:45.859 --> 01:40:50.970
about this. But I agree. I think my number one

01:40:50.970 --> 01:40:53.229
statement is don't believe the hype, but there's

01:40:53.229 --> 01:40:56.729
some pretty interesting stuff going on. And another

01:40:56.729 --> 01:40:58.930
thing that I say as someone who's been in the

01:40:58.930 --> 01:41:01.789
data world for a while is like, well, like some

01:41:01.789 --> 01:41:04.050
of the ways that people are using the term AI

01:41:04.050 --> 01:41:07.050
is really just like kind of predictive analytics

01:41:07.050 --> 01:41:10.369
and like, it's not that cool. Like, so yeah,

01:41:10.369 --> 01:41:13.489
I think that there's some applications that are

01:41:13.489 --> 01:41:15.500
going to totally change the world, I hope. in

01:41:15.500 --> 01:41:19.220
a good way. But I also see in the business world

01:41:19.220 --> 01:41:24.380
that it seems like there's a lot of new startups.

01:41:24.920 --> 01:41:27.939
And I was just at a conference in Vegas a few

01:41:27.939 --> 01:41:30.439
months ago and really saw that it was just like

01:41:30.439 --> 01:41:33.079
people are really capitalizing on this change.

01:41:33.539 --> 01:41:37.859
I mean, I saw it a bit when like like cookies

01:41:37.859 --> 01:41:39.619
started to be deprecated and things like that

01:41:39.619 --> 01:41:41.159
too. There's a ton of people saying how they're

01:41:41.159 --> 01:41:43.560
gonna still be able to identify your customers

01:41:43.560 --> 01:41:46.560
all these fancy ways. So like happens every time.

01:41:47.500 --> 01:41:50.579
But I think for businesses, but you know, might

01:41:50.579 --> 01:41:52.960
be listening to this. My advice would be, yeah,

01:41:53.119 --> 01:41:55.800
think about what could really impact both your

01:41:55.800 --> 01:41:59.350
customer. and the operations of your company

01:41:59.350 --> 01:42:02.130
first. Like what will save time of smart people?

01:42:02.590 --> 01:42:04.710
You don't want it to use it to replace smart

01:42:04.710 --> 01:42:07.550
people or to replace specific jobs. What you

01:42:07.550 --> 01:42:10.210
want to do is make it so that those smart people

01:42:10.210 --> 01:42:12.670
can use even more of their time doing smart stuff.

01:42:13.250 --> 01:42:15.329
Like data scientists are a perfect example or

01:42:15.329 --> 01:42:19.590
like engineers. What if you could use AI to do

01:42:20.279 --> 01:42:23.619
80 % of the data cleansing and data quality checks

01:42:23.619 --> 01:42:25.939
that you're currently doing. Like imagine how

01:42:25.939 --> 01:42:28.500
much more time would be freed up to actually

01:42:28.500 --> 01:42:31.079
analyze data and actually make business decisions.

01:42:31.720 --> 01:42:33.939
So, you know, I think, I think that's like where

01:42:33.939 --> 01:42:37.439
I see is, is like, and I, and I have these, like

01:42:37.439 --> 01:42:39.619
my new post is about like the truths of data,

01:42:39.960 --> 01:42:43.100
like hard truths of AI and you know, that touches

01:42:43.100 --> 01:42:47.300
a few of them. It's just like, I always see this.

01:42:47.399 --> 01:42:49.100
We get all excited about stuff in the business

01:42:49.100 --> 01:42:52.359
world and we get all worked up about it. Sometimes

01:42:52.359 --> 01:42:54.279
it's a big deal and sometimes it's not, but we're

01:42:54.279 --> 01:43:00.539
super adaptable. I think that there's some interesting

01:43:00.539 --> 01:43:04.039
things, but they're not really like what I imagined

01:43:04.039 --> 01:43:06.420
AI would be. If you had told me that AI was going

01:43:06.420 --> 01:43:09.840
to become a hot topic in 2023, and you asked

01:43:09.840 --> 01:43:11.479
me that a few years ago, I don't think I would

01:43:11.479 --> 01:43:13.560
be describing something like chat GBT because

01:43:13.560 --> 01:43:16.819
to me, it's like, It's just like the next generation

01:43:16.819 --> 01:43:20.399
of Google search. Let's be honest. Yes, it has

01:43:20.399 --> 01:43:22.600
a lot of things colored into that, and it helps

01:43:22.600 --> 01:43:25.279
you do things, and it's really interesting, and

01:43:25.279 --> 01:43:29.279
there's so many applications. But really, I'm

01:43:29.279 --> 01:43:32.159
excited for what comes next, too. What is really

01:43:32.159 --> 01:43:35.199
going to be pushing the envelope? Because I think

01:43:35.199 --> 01:43:38.819
it's a long time coming, and there's been pieces

01:43:38.819 --> 01:43:40.720
of it that we've been using for a long time.

01:43:42.000 --> 01:43:44.960
Yeah, absolutely. Yeah, and so some of these

01:43:44.960 --> 01:43:48.539
large language models are definitely a shoe to

01:43:48.539 --> 01:43:52.520
leap in technical advancement, but for the business

01:43:52.520 --> 01:43:59.039
user, it's more of a more sophisticated chatbot

01:43:59.039 --> 01:44:02.279
in a lot of ways. Right. Yeah. Right. And there

01:44:02.279 --> 01:44:08.079
are these other advancements like auto GPT, which

01:44:08.079 --> 01:44:14.239
you don't have to prompt it every time. You just

01:44:14.239 --> 01:44:18.060
give it a goal and you ask it to book a hotel

01:44:18.060 --> 01:44:22.020
and go look at different hotels and look at time

01:44:22.020 --> 01:44:25.859
slots and then book a hotel for you and stuff

01:44:25.859 --> 01:44:27.640
like that. So there's some advancements there,

01:44:27.640 --> 01:44:35.680
but totally get your point on. We definitely

01:44:35.680 --> 01:44:39.239
have to allow it to mature and see where it goes.

01:44:39.699 --> 01:44:42.100
And also you talked about the hype cycle. You

01:44:42.100 --> 01:44:45.840
know, I think a lot of ways, I was at South by

01:44:45.840 --> 01:44:50.420
Southwest and AI was a big topic. And last year

01:44:50.420 --> 01:44:55.500
it was Web3 and crypto. So it was like replace

01:44:55.500 --> 01:44:58.520
Web3 and crypto with AI. That was kind of the

01:44:58.520 --> 01:45:03.659
theme this year. So I don't want that to happen

01:45:03.659 --> 01:45:06.500
that, you know, a lot of companies trying to

01:45:06.500 --> 01:45:10.989
capitalize on it, say AI for everything and then

01:45:10.989 --> 01:45:13.449
just people lose interest in it because there's

01:45:13.449 --> 01:45:19.510
a lot of ways that this is useful. Yeah, I guess.

01:45:20.130 --> 01:45:23.029
I'm less impressed with crypto, but I agree.

01:45:24.409 --> 01:45:27.409
It's something where I really do think it's the

01:45:27.409 --> 01:45:30.869
start of something new that's really going to...

01:45:30.939 --> 01:45:33.800
And I, at least in my hopes, and maybe I'm a

01:45:33.800 --> 01:45:35.319
bit idealistic, you know, there's some people

01:45:35.319 --> 01:45:38.640
who are all doom about it, but I don't really

01:45:38.640 --> 01:45:41.199
think we're there yet for it to be really doomy.

01:45:41.560 --> 01:45:44.739
But I have concerns about bias and models and

01:45:44.739 --> 01:45:47.439
things like that. But I think there's things

01:45:47.439 --> 01:45:50.859
like the healthcare system and global warming

01:45:50.859 --> 01:45:52.760
and things like that, that we can really transform

01:45:52.760 --> 01:45:54.460
the world with this. So I don't want people to

01:45:54.460 --> 01:45:59.470
lose interest. And so I think it's for people

01:45:59.470 --> 01:46:01.750
like us continuing the conversation in a really

01:46:01.750 --> 01:46:05.229
pragmatic way and starting to show, here's the

01:46:05.229 --> 01:46:08.609
real benefit. Here's some tangible ways that

01:46:08.609 --> 01:46:10.970
make a lot of sense. It's the same thing with

01:46:10.970 --> 01:46:13.789
predictive or prescriptive modeling. It's something

01:46:13.789 --> 01:46:16.770
that's so useful. So I'm sure that we'll identify

01:46:16.770 --> 01:46:21.229
pretty soon what those are in AI. And then making

01:46:21.229 --> 01:46:23.350
sure they don't go away because we can see the

01:46:23.350 --> 01:46:26.090
return on investment. So I'm super excited about

01:46:26.090 --> 01:46:28.829
that. like you said, we're just still scratching

01:46:28.829 --> 01:46:31.390
the surface there and there's a lot of maturity

01:46:31.390 --> 01:46:35.310
that we'll need to develop. But I also think

01:46:35.310 --> 01:46:38.890
what you're saying about companies, part of what

01:46:38.890 --> 01:46:41.909
I'm saying to people is like, please don't hire

01:46:41.909 --> 01:46:46.560
a see like a sea level like chief AI officer

01:46:46.560 --> 01:46:49.939
right now like I think it's a good time to start

01:46:49.939 --> 01:46:52.159
to understand that starts asking questions of

01:46:52.159 --> 01:46:54.659
people in your company already and if you don't

01:46:54.659 --> 01:46:56.619
have that chief data officer whatever you want

01:46:56.619 --> 01:46:59.180
to call that get that person first because that's

01:46:59.180 --> 01:47:02.560
more of like a holistic picture of everything

01:47:02.560 --> 01:47:05.579
data and so I think that if people are excited

01:47:05.579 --> 01:47:07.220
about AI and think that's the direction their

01:47:07.220 --> 01:47:09.260
company wants to go and if you think about some

01:47:09.260 --> 01:47:13.779
of the more advanced more data mature companies,

01:47:14.140 --> 01:47:17.020
they do use either machine learning or just really

01:47:17.020 --> 01:47:20.880
data -driven. I heard a talk from Stitch Fix

01:47:20.880 --> 01:47:26.199
at Shop Talk this spring. Even though they're

01:47:26.199 --> 01:47:28.000
an e -commerce business and things like that,

01:47:28.520 --> 01:47:31.439
they just have data as a core way they function.

01:47:31.640 --> 01:47:34.920
If you want to go that way, like assessing your

01:47:34.920 --> 01:47:38.579
whole data maturity and taking an honest look

01:47:38.579 --> 01:47:40.979
at yourself is a good first step because then

01:47:40.979 --> 01:47:42.880
you're going to really know what is the right

01:47:42.880 --> 01:47:46.239
decision when it comes to AI. No, that's a great

01:47:46.239 --> 01:47:48.840
point and I've been thinking about this a lot

01:47:48.840 --> 01:47:58.140
with a lot of talk about AI and data. What is

01:47:58.140 --> 01:48:02.880
AI and how is it different from modeling or data

01:48:02.880 --> 01:48:06.659
science aspects of this thing. I still have to

01:48:06.659 --> 01:48:09.359
think more about it, but the way where I am right

01:48:09.359 --> 01:48:14.300
now is data science is more about extracting

01:48:14.300 --> 01:48:17.600
intelligence from your existing data, historical

01:48:17.600 --> 01:48:21.359
data, and maybe predicting what will happen,

01:48:21.380 --> 01:48:25.699
but based off of what you have. AI, at least

01:48:25.699 --> 01:48:29.840
in its current iteration, is more of a productivity

01:48:29.840 --> 01:48:33.819
tool. that you can come and use it as a plugin

01:48:33.819 --> 01:48:39.659
to take your productivity to the next level.

01:48:41.180 --> 01:48:43.180
I would really agree with that. That's kind of

01:48:43.180 --> 01:48:45.739
how the way I would also think about it and had

01:48:45.739 --> 01:48:48.060
chewed on that a bit myself. So yeah, I would

01:48:48.060 --> 01:48:50.039
agree. I think it's really like there's ways

01:48:50.039 --> 01:48:52.800
that we can apply it to make things easier for

01:48:52.800 --> 01:48:56.619
ourselves. But it's not necessarily like I've

01:48:56.619 --> 01:48:59.539
seen some startups do things like, look, you

01:48:59.539 --> 01:49:03.619
can use this kind of AI chat bot thing and like

01:49:03.619 --> 01:49:05.699
ask it questions and it'll answer like it was

01:49:05.699 --> 01:49:08.300
your analyst. It'll pull data from like your

01:49:08.300 --> 01:49:12.739
data sources. I'm like, oh, like I or like that

01:49:12.739 --> 01:49:14.840
ones where they're like, oh, I'm going to replace

01:49:14.840 --> 01:49:18.600
your copywriter with AI. I'm like, like, you

01:49:18.600 --> 01:49:21.119
know, I think yes, if you really wanted to, but

01:49:21.119 --> 01:49:24.619
like, isn't there like, why can't you Why don't

01:49:24.619 --> 01:49:26.939
you make things 100 times more efficient for

01:49:26.939 --> 01:49:29.579
those different people and make it so that they

01:49:29.579 --> 01:49:31.699
can spend even more time answering your business

01:49:31.699 --> 01:49:35.359
questions and writing good copy versus some of

01:49:35.359 --> 01:49:37.600
the other more menial tasks that they need to

01:49:37.600 --> 01:49:41.399
do? Yeah. And then this goes back to our previous

01:49:41.399 --> 01:49:45.119
point on having context around what the data

01:49:45.119 --> 01:49:48.300
is saying without those subject matter experts

01:49:48.300 --> 01:49:50.800
in there to provide some context or understand

01:49:50.800 --> 01:49:57.560
some context. these AI models can't really, it

01:49:57.560 --> 01:50:02.319
can derive way different interpretations than

01:50:02.319 --> 01:50:08.460
what it really should be or what the data is

01:50:08.460 --> 01:50:12.140
actually telling you. Yeah, definitely. You need

01:50:12.140 --> 01:50:17.880
to have that institutional knowledge for copywriting.

01:50:18.640 --> 01:50:23.010
You can't just automate it. at least not in this

01:50:23.010 --> 01:50:28.630
iteration without trading the model. Very cool.

01:50:29.470 --> 01:50:30.930
And I'm not just saying that because I don't

01:50:30.930 --> 01:50:36.329
want to be out of a job. No, like you said, it

01:50:36.329 --> 01:50:42.529
makes copywriter's job a lot more easier, but

01:50:42.529 --> 01:50:47.500
it cannot replace copywriter's. Yeah, it's like

01:50:47.500 --> 01:50:49.680
the creative thing that like, you know, we are

01:50:49.680 --> 01:50:51.619
using it to kind of generate some art and things

01:50:51.619 --> 01:50:54.260
like that, but it's not necessarily like ever

01:50:54.260 --> 01:50:59.079
going to have that same raw energy and like ideas

01:50:59.079 --> 01:51:03.359
that like a person is going to have. Yeah, you

01:51:03.359 --> 01:51:06.619
could, I mean, I'm using chat GPT to write some

01:51:06.619 --> 01:51:09.500
of my blogs and stuff like that. The way that

01:51:09.500 --> 01:51:13.920
I think about it is if I can't think of any ideas,

01:51:14.539 --> 01:51:19.000
I'll use it to you know, give me a head start.

01:51:19.359 --> 01:51:24.119
If I have an idea and I am like 80 % there and

01:51:24.119 --> 01:51:27.020
use it to kind of fine tune it. So both sides,

01:51:27.479 --> 01:51:32.579
but it cannot, it can give you the 20 % edge,

01:51:32.579 --> 01:51:35.399
but you still have to do the 80 % of the work

01:51:35.399 --> 01:51:38.800
that you need to do. Otherwise it's, I mean,

01:51:38.800 --> 01:51:41.699
I could probably just ask, give a topic to ChatGPD

01:51:41.699 --> 01:51:45.520
and ask it to write an article, but it'll be

01:51:45.520 --> 01:51:53.420
so robotic. and they'll be lacking soul. I just

01:51:53.420 --> 01:51:56.100
saw an article about it and maybe we can also

01:51:56.100 --> 01:51:59.039
throw that in your podcast notes or whatever.

01:51:59.819 --> 01:52:02.439
But I saw some other interesting ways where I

01:52:02.439 --> 01:52:06.079
think what I'm considering using it for is I

01:52:06.079 --> 01:52:09.619
don't every day get into R scripts and things

01:52:09.619 --> 01:52:11.260
like that, but you can actually have it really

01:52:11.260 --> 01:52:13.800
help you build some of that stuff for yourself.

01:52:14.100 --> 01:52:17.229
And it's like, why not? And there's this one

01:52:17.229 --> 01:52:19.609
article that I just shared to my team that was

01:52:19.609 --> 01:52:25.550
about using a combination of R, ChatGBT, to help

01:52:25.550 --> 01:52:28.970
you do data quality analysis. And I thought that

01:52:28.970 --> 01:52:32.590
was a really interesting way to automate pieces

01:52:32.590 --> 01:52:35.609
of that, where before we might spend a lot more

01:52:35.609 --> 01:52:38.189
time tweaking the code and making sure everything's

01:52:38.189 --> 01:52:40.270
right and everything's pulling in through. And

01:52:40.270 --> 01:52:43.670
they can connect the dots a little bit for you.

01:52:44.170 --> 01:52:45.909
I haven't read the... the whole article completely,

01:52:45.970 --> 01:52:47.989
but I just like some interesting concepts about

01:52:47.989 --> 01:52:51.489
how that can make our jobs easier. Yeah, there's

01:52:51.489 --> 01:52:55.310
a lot of applications and different, like you

01:52:55.310 --> 01:52:58.250
said, different pieces of the workflow and not

01:52:58.250 --> 01:53:01.689
replacing the entire workflow. Speaking of which,

01:53:02.609 --> 01:53:05.850
are you using ChadGBD in any way in your workflow?

01:53:06.109 --> 01:53:10.449
And if so, how so? I mean, you know, I've played

01:53:10.449 --> 01:53:12.750
around with it a bit. Like you said, I've also

01:53:12.750 --> 01:53:14.609
kind of used it like before to be like, just

01:53:14.609 --> 01:53:16.930
kind of see like, okay, how can it help me write

01:53:16.930 --> 01:53:19.829
content? And, you know, and then I've looked

01:53:19.829 --> 01:53:21.350
started just like look around at some of those

01:53:21.350 --> 01:53:24.569
other ideas, like, well, how can we connect it

01:53:24.569 --> 01:53:26.609
in with like what we're currently doing? And

01:53:26.609 --> 01:53:28.529
like, maybe we can play with that. We haven't

01:53:28.529 --> 01:53:31.180
like I haven't found any particular workflows

01:53:31.180 --> 01:53:34.039
that are now part of my day -to -day. I've really

01:53:34.039 --> 01:53:36.000
dipped my toes in it, really played with the

01:53:36.000 --> 01:53:37.840
platform and spent a lot more time listening

01:53:37.840 --> 01:53:40.340
to what everyone else is saying about it. I'm

01:53:40.340 --> 01:53:43.680
super interested in all the discourse about the

01:53:43.680 --> 01:53:46.979
privacy of it and the IP and everything like

01:53:46.979 --> 01:53:50.960
that and what everyone's using it for. But I

01:53:50.960 --> 01:53:53.399
have done some of those things too that just

01:53:53.399 --> 01:53:56.819
make it really quick. I forget, I've done some

01:53:56.819 --> 01:54:00.529
things like... make me a recipe or create a plan

01:54:00.529 --> 01:54:02.770
for me. And like, I kind of use it like a little

01:54:02.770 --> 01:54:05.369
bit of an assistant. And I'm super interested

01:54:05.369 --> 01:54:07.250
in the idea of like, you know, some of the other

01:54:07.250 --> 01:54:09.609
ones that can really like iteratively like go

01:54:09.609 --> 01:54:12.250
figure something out for you. Because that's

01:54:12.250 --> 01:54:13.850
something that like, you know, right now I'm

01:54:13.850 --> 01:54:16.529
not once again, not replacing anyone's job, but

01:54:16.529 --> 01:54:18.670
it can make things a lot easier. But haven't

01:54:18.670 --> 01:54:20.590
found my like one thing that's like, okay, every

01:54:20.590 --> 01:54:22.350
day I have it open. I'm really still in like

01:54:22.350 --> 01:54:25.369
kind of play around mode. That makes sense. That

01:54:25.369 --> 01:54:29.090
makes sense. Yeah. There's definitely a lot of

01:54:29.090 --> 01:54:31.270
discourse about it and it's super interesting

01:54:31.270 --> 01:54:36.149
time to just learn about what's going on. What

01:54:36.149 --> 01:54:40.850
are some of the cool applications that you would

01:54:40.850 --> 01:54:44.229
like to see of, say, these large language models?

01:54:45.550 --> 01:54:49.109
I think the one I've mentioned in passing a bit

01:54:49.109 --> 01:54:54.640
is things such as improving And this is something

01:54:54.640 --> 01:54:58.520
I listened to another podcast recently about

01:54:58.520 --> 01:55:03.600
creating simulations. So I think it could be

01:55:03.600 --> 01:55:06.060
a way to make simulations a little easier. And

01:55:06.060 --> 01:55:08.020
what's interesting about that or what is really

01:55:08.020 --> 01:55:10.579
impactful would be thinking about some of the

01:55:10.579 --> 01:55:14.079
critical things that can save lives. hospital

01:55:14.079 --> 01:55:16.060
bed capacity was one of the examples in that

01:55:16.060 --> 01:55:18.960
podcast. I really want to see us take a look

01:55:18.960 --> 01:55:21.439
at some of those things or maybe traffic congestion,

01:55:21.619 --> 01:55:23.539
some of these problems that are super hard to

01:55:23.539 --> 01:55:27.779
solve for, and see if we can make simulations

01:55:27.779 --> 01:55:30.020
and models that are a little bit easier for us

01:55:30.020 --> 01:55:33.340
to conceptually see these big, big problems solved.

01:55:35.090 --> 01:55:36.890
That's something that like simulations already

01:55:36.890 --> 01:55:38.930
can do and they can train them really quickly.

01:55:39.350 --> 01:55:41.529
But I wonder with like these bigger language

01:55:41.529 --> 01:55:44.109
models and being able to just kind of dump a

01:55:44.109 --> 01:55:46.149
ton of input into it, like it'd be so cool if

01:55:46.149 --> 01:55:48.449
you could dump all that input, like almost like

01:55:48.449 --> 01:55:51.109
a whole data lakes worth, tell it to. cleanse

01:55:51.109 --> 01:55:53.869
it, normalize metrics, blend everything together,

01:55:54.069 --> 01:55:56.810
consider all the pitfalls, and then tell it the

01:55:56.810 --> 01:55:58.890
business question, and it's able to model things

01:55:58.890 --> 01:56:01.890
out for you. I think we're super far, far away

01:56:01.890 --> 01:56:03.510
from that right now, but that's where I think

01:56:03.510 --> 01:56:06.670
it will be super helpful. Because another thing

01:56:06.670 --> 01:56:09.840
from that... That podcast that was really interesting

01:56:09.840 --> 01:56:12.520
to me was thinking about how that could explain

01:56:12.520 --> 01:56:17.079
why we make some hard decisions in politics and

01:56:17.079 --> 01:56:19.479
in the government and things like that. If you

01:56:19.479 --> 01:56:23.520
really model it out and say, well, this is why,

01:56:24.340 --> 01:56:26.739
it can be really impactful. But really, anything

01:56:26.739 --> 01:56:30.159
that's going to positively impact people's lives

01:56:30.159 --> 01:56:31.899
is where I really want to see the first one,

01:56:31.899 --> 01:56:34.239
even though I sit here and I work for for -profit

01:56:34.239 --> 01:56:39.930
companies all the time. No, that's a great outlook.

01:56:40.050 --> 01:56:47.149
It definitely can help in a lot of areas. It

01:56:47.149 --> 01:56:50.890
can help a lot of people. I think Satya Nadella,

01:56:51.270 --> 01:56:56.430
Microsoft's CEO, talked about how language barriers

01:56:56.430 --> 01:57:00.189
can be broken and you don't need to know language.

01:57:00.250 --> 01:57:02.930
Google already does this with their Google Translate,

01:57:02.930 --> 01:57:07.840
but this could take it to the next level. You

01:57:07.840 --> 01:57:12.000
know, with ChatGBD's human -like speech abilities,

01:57:12.699 --> 01:57:15.460
I can definitely take it to the next level. Yeah,

01:57:15.460 --> 01:57:18.439
it'll be like Star Wars, right? Little translator

01:57:18.439 --> 01:57:20.640
in your ear, I'm so excited. That's one I didn't

01:57:20.640 --> 01:57:23.420
think of right now, but yes, that'll be really

01:57:23.420 --> 01:57:26.840
interesting. Like, getting to the like, you know,

01:57:27.000 --> 01:57:29.579
in the moment, being able to like translate directions

01:57:29.579 --> 01:57:33.020
or whatever, super cool. Yeah, it's super helpful

01:57:33.020 --> 01:57:36.159
and, you know. Like you said, you touched on

01:57:36.159 --> 01:57:40.220
this, but right now you could use it more like

01:57:40.220 --> 01:57:44.279
a personal assistant, free of cost right now.

01:57:44.300 --> 01:57:47.920
So that's really cool. We will see how this goes.

01:57:48.979 --> 01:57:51.399
No, appreciate your time. This was super fun.

01:57:52.399 --> 01:57:55.960
I learned a lot about data literacy, wrote down

01:57:55.960 --> 01:58:00.039
some notes, data literacy, customer focus, data

01:58:00.039 --> 01:58:05.789
maturity, and data... Everything. Well, thank

01:58:05.789 --> 01:58:07.590
you. I mean, I really love to hear someone else's

01:58:07.590 --> 01:58:10.390
opinion about ChatGVT, who's really like being

01:58:10.390 --> 01:58:13.250
an AI, like being more curious and pragmatic

01:58:13.250 --> 01:58:16.550
about it. So it's good to hear that some of the

01:58:16.550 --> 01:58:18.770
ideas that I have are validated. I'm not the

01:58:18.770 --> 01:58:21.170
only one thinking of it that way. So thank you

01:58:21.170 --> 01:58:22.970
so much for having me. It was so nice to catch

01:58:22.970 --> 01:58:26.210
up. I feel like it wasn't that long ago that

01:58:26.210 --> 01:58:28.449
we were having some meetings together internally.

01:58:29.390 --> 01:58:32.600
So we should keep doing this. Absolutely, absolutely.

01:58:32.739 --> 01:58:35.180
Thank you so much for your insights and we'll

01:58:35.180 --> 01:58:38.159
definitely leave your company's websites and

01:58:38.159 --> 01:58:40.800
LinkedIn and contact information with the podcast

01:58:40.800 --> 01:58:43.640
notes. So yeah, thank you so much. Appreciate

01:58:43.640 --> 01:58:45.960
you helping out. Thank you so much. I'll talk

01:58:45.960 --> 01:58:47.460
to you later. Thank you. Bye.