1
00:00:00,000 --> 00:00:04,640
Do you ever get the feeling that marketing is like right on the edge of

2
00:00:04,640 --> 00:00:10,480
like a total transformation? Like we're going to go from you know just

3
00:00:10,480 --> 00:00:14,120
like throwing stuff at the wall and seeing what it takes to being able to

4
00:00:14,120 --> 00:00:18,800
like actually hit the bullseye every single time. So that's what we're going

5
00:00:18,800 --> 00:00:25,760
to dive into today. The world of AI and LLMs and marketing. So I had some really

6
00:00:25,760 --> 00:00:31,880
interesting insights to share as a data driven marketer. So one of my recent

7
00:00:31,880 --> 00:00:40,320
articles, top 10 AI and LLM trends which were transforming marketing in 2024. This

8
00:00:40,320 --> 00:00:45,240
is something I wrote recently and I'm excited to share about that today. So

9
00:00:45,240 --> 00:00:51,880
what's fascinating is that you know there's like a lot of shift happening

10
00:00:51,880 --> 00:00:58,120
these days with everybody switching to AI. So it's not just being data

11
00:00:58,120 --> 00:01:02,800
driven, that's just not like the anything that's going to cut it anywhere. We are

12
00:01:02,800 --> 00:01:11,960
going to need to be constantly evolving, constantly shifting. It's about being

13
00:01:11,960 --> 00:01:18,700
insight driven. So you keep calling us as data driven but it has to be insight

14
00:01:18,700 --> 00:01:26,560
driven is the more accurate word to be using. So let's just talk

15
00:01:26,560 --> 00:01:33,720
about AI right? And AI is not just something that just processes numbers, it

16
00:01:33,720 --> 00:01:39,000
understands the why behind the numbers, helps you give you more clarity of why

17
00:01:39,000 --> 00:01:46,320
those numbers are. So to help you connect the dots, let me help

18
00:01:46,320 --> 00:01:50,080
you understand what does an insight driven marketing campaign actually look

19
00:01:50,080 --> 00:01:55,640
like in practice. So imagine this, you're scrolling to your inbox right and you

20
00:01:55,640 --> 00:02:00,320
see an email from your favorite coffee shop for example. Here's the kicker,

21
00:02:00,320 --> 00:02:07,200
it's not just advertising their latest latte, it's suggesting a new

22
00:02:07,200 --> 00:02:13,520
drink based on your previous orders. So right, what does that mean? It's

23
00:02:13,520 --> 00:02:17,840
something like it's very personalized right, like it's using your previous

24
00:02:17,840 --> 00:02:24,840
data to suggest you something new right. So let's just I mean the time of the

25
00:02:24,840 --> 00:02:30,120
day, even the local weather, that's something you know it's more personalized.

26
00:02:30,120 --> 00:02:34,760
It would know that information based on you know where you purchased from

27
00:02:34,760 --> 00:02:41,800
in the past as well. So that's really the power of AI digging into the data to

28
00:02:41,800 --> 00:02:46,880
predict what they'll actually want. So it's sort of just like sending an email

29
00:02:46,880 --> 00:02:51,120
blast right to everyone, like an email broadcast to everyone with the same

30
00:02:51,120 --> 00:02:57,560
generic message. You are actually personalizing what you're sending to

31
00:02:57,560 --> 00:03:02,600
your subscriber. So really tailoring your experience to each

32
00:03:02,600 --> 00:03:09,400
individual subscriber or customer that you have. That's kind of a next

33
00:03:09,400 --> 00:03:15,440
personalization here. So this kind of personalization which used to be really

34
00:03:15,440 --> 00:03:20,560
unthinkable before, you need to time and resource constraints and obviously

35
00:03:20,560 --> 00:03:27,200
something like AI was not available before or not easily accessible. So people

36
00:03:27,200 --> 00:03:31,640
are now able to think in these directions. Now it's all achievable thanks

37
00:03:31,640 --> 00:03:37,200
to this enhancements in generative AI and enlarged language models

38
00:03:37,200 --> 00:03:46,560
right and these technologies can analyze you know huge fairly massive sets of

39
00:03:46,560 --> 00:03:52,760
data of customer information basically and it can understand patterns, what are

40
00:03:52,760 --> 00:03:59,600
the customer preferences and then automatically just generate highly

41
00:03:59,600 --> 00:04:05,120
tailored content from these emails and social media post to even get to the

42
00:04:05,120 --> 00:04:10,280
better product recommendations. So we're not just talking about AI that knows

43
00:04:10,280 --> 00:04:16,880
that you want a coffee but it also understands that you like a certain kind

44
00:04:16,880 --> 00:04:22,800
of like iced coffee with like oat milk or like almond milk with an extra shot

45
00:04:22,800 --> 00:04:29,960
right. Like your matcha latte with like four shots of vanilla foam. That's

46
00:04:29,960 --> 00:04:38,840
something that you know it understands now. So that's really awesome

47
00:04:38,840 --> 00:04:45,800
right and it can use that knowledge. AI can use that knowledge to basically

48
00:04:45,800 --> 00:04:51,920
like you know craft a message that feels like it was written just for you and

49
00:04:51,920 --> 00:04:57,760
even if millions of other people are getting a kind of a similar version of

50
00:04:57,760 --> 00:05:04,000
that campaign right it would feel unique to you because there's a lot of

51
00:05:04,000 --> 00:05:09,640
personalization happening here. It's like having a very dedicated marketing team

52
00:05:09,640 --> 00:05:16,520
for every single customer and that's amazing and that's where we are with

53
00:05:16,520 --> 00:05:21,520
things these days and this is and this is where it gets interesting too from a

54
00:05:21,520 --> 00:05:27,040
business perspective. AI can map out the entire customer journey in a way that

55
00:05:27,040 --> 00:05:34,160
was previously impossible. So let's say I'm trying to picture one of these you

56
00:05:34,160 --> 00:05:38,160
know murder boards you see detective movies right with the strings

57
00:05:38,160 --> 00:05:42,880
connecting all the clues but instead of the clues it's customer interaction. So

58
00:05:42,880 --> 00:05:49,480
it's like a mind map. Yeah you know it's kind of like a it's like a map of things

59
00:05:49,480 --> 00:05:56,100
right like how you're progressing from journey A to I mean from A to Z and that

60
00:05:56,100 --> 00:06:01,440
whole journey is getting shown and yeah for lack of better terminology the

61
00:06:01,440 --> 00:06:05,880
murder board was like a nice way to display things but that's because I'm

62
00:06:05,880 --> 00:06:13,960
also into murder movies basically in these crime movies so that's why. But how

63
00:06:13,960 --> 00:06:19,200
does it actually translate into actionable insights for marketers? How

64
00:06:19,200 --> 00:06:24,920
do they actually use that information? So let me talk about that as well right. So

65
00:06:24,920 --> 00:06:31,600
imagine being able to pinpoint the exact moment of customer abandonce their card.

66
00:06:31,600 --> 00:06:37,240
All specific web page where they lose interest like AI can identify this

67
00:06:37,240 --> 00:06:41,160
friction points within the customer journey highlighting areas for improvement.

68
00:06:41,160 --> 00:06:46,160
So instead of guessing what's working and what's not, they now have a

69
00:06:46,160 --> 00:06:51,120
data-driven roadmap for optimizing the entire customer experience right. For

70
00:06:51,120 --> 00:06:56,840
example say company notices through the AI powered customer journey mapping that

71
00:06:56,840 --> 00:07:01,200
a large percentage of customers abandon their online shopping cards after

72
00:07:01,200 --> 00:07:06,440
viewing shopping course right. So yeah I mean how many times has that happened

73
00:07:06,440 --> 00:07:12,120
like you're trying to go you see a product you like you know maybe you

74
00:07:12,120 --> 00:07:19,840
entered through a Google ad maybe you're looking for sneakers, you're looking for

75
00:07:19,840 --> 00:07:26,320
sneakers and for men and you're also giving like a size of sneakers like a

76
00:07:26,320 --> 00:07:30,960
wide sneakers might fit. So I have wide feet so that's why I'm just thinking

77
00:07:30,960 --> 00:07:38,520
about that first. But that being said like so you're going into this you're

78
00:07:38,520 --> 00:07:47,840
clicking into this ad and yeah you're basically just like you go into all the

79
00:07:47,840 --> 00:07:52,600
funnel steps and eventually you are landing on the checkout page where you're

80
00:07:52,600 --> 00:07:57,480
trying to pay and they tell you oh you know there's a shipping charge associated

81
00:07:57,480 --> 00:08:02,520
with it. What do you do? Most likely you're gonna drop because there are

82
00:08:02,520 --> 00:08:07,240
other companies which offer yes free shipping and you probably go for that

83
00:08:07,240 --> 00:08:15,360
right. Yeah I mean but the company would notice something like this trend

84
00:08:15,360 --> 00:08:20,520
happening with the AI powered customer journey. So it will give you like more

85
00:08:20,520 --> 00:08:24,720
insights like oh you know what your customers are dropping, large percentage

86
00:08:24,720 --> 00:08:28,640
of your customers are dropping during their shipping process. Yes you can do

87
00:08:28,640 --> 00:08:32,720
it with data as well but like there's so many things that you're constantly

88
00:08:32,720 --> 00:08:37,760
thinking about and you want somebody else to also do that thinking for you

89
00:08:37,760 --> 00:08:42,280
and that's where AI comes in. So you're basically you're armed now with that

90
00:08:42,280 --> 00:08:48,240
insight and now you could also experiment like do any B-test with

91
00:08:48,240 --> 00:08:55,200
free shipping and see you know if that sticks and you can maybe adjust the

92
00:08:55,200 --> 00:08:59,440
price or highlight faster shipping options to see how that also impacts

93
00:08:59,440 --> 00:09:07,080
conversion. So it's like having like a complete top view, a bird's eye view

94
00:09:07,080 --> 00:09:13,640
to be to say another word of the entire customer journey right and but instead

95
00:09:13,640 --> 00:09:19,280
of just observing you can actually tweak things in real time to keep

96
00:09:19,280 --> 00:09:24,000
people moving smoothly towards their purchase and that real time

97
00:09:24,000 --> 00:09:30,280
aspect is key to another major trend right. It highlights the

98
00:09:30,280 --> 00:09:35,520
rise of AI powered dashboards. So that's where I'm talking about right. You are

99
00:09:35,520 --> 00:09:41,200
seeing it in real time with staging and you're able to react to it quickly.

100
00:09:41,200 --> 00:09:46,320
That's where AI powered dashboards come into play. So they always sound cool in

101
00:09:46,320 --> 00:09:51,760
theory but I admit the term kind of makes me think of endless spreadsheets

102
00:09:51,760 --> 00:09:59,240
and just data overload and I'm sure like some people who are like not as

103
00:09:59,240 --> 00:10:04,200
familiar they might think just that with not as familiar with the data aspect

104
00:10:04,200 --> 00:10:08,920
they might be thinking that's intimidating in dashboards but

105
00:10:08,920 --> 00:10:15,240
and that can be intimidating for any new person. So how are AI powered

106
00:10:15,240 --> 00:10:22,520
dashboards different? I mean to explain in simpler terms let's just

107
00:10:22,520 --> 00:10:27,320
forget about spreadsheets for a second right. Imagine like a very sleek

108
00:10:27,320 --> 00:10:32,080
interface that pulls in data from all your marketing channels, the social media,

109
00:10:32,080 --> 00:10:37,400
even your customer service interactions and then uses AI to distill it

110
00:10:37,400 --> 00:10:41,880
all into clear action and insights. So instead of just drowning in the data

111
00:10:41,880 --> 00:10:45,040
you're getting a clear picture of what's resonating with the customers, what's

112
00:10:45,040 --> 00:10:52,840
falling flat right and where the opportunities lie. All in real time and

113
00:10:52,840 --> 00:10:57,560
this isn't just about tweaking your marketing copy or adjusting your ad

114
00:10:57,560 --> 00:11:04,240
spend. Just imagine like seeing a sudden spike in your negative sentiment around

115
00:11:04,240 --> 00:11:08,480
the new product launch. How would you react right? I mean obviously that

116
00:11:08,480 --> 00:11:16,120
that would be demoralizing. Now with AI what you're going to do is like maybe

117
00:11:16,120 --> 00:11:20,800
you can with the AI you're analyzing your real-time feedback. Basically AI is

118
00:11:20,800 --> 00:11:25,200
doing that real-time feedback analysis from your social media and your online

119
00:11:25,200 --> 00:11:31,920
interviews and you can exactly look at the issue. You can pinpoint a specific

120
00:11:31,920 --> 00:11:36,960
issue which is causing that negative sentiment and adapt your strategy

121
00:11:36,960 --> 00:11:46,680
before it becomes like a full-blown PR crisis right. So it's like having a

122
00:11:46,680 --> 00:11:50,160
crystal ball that shows you not just what's happening but why it's happening.

123
00:11:50,160 --> 00:11:55,760
That's impressive I think. But this level of insight, this ability to personalize

124
00:11:55,760 --> 00:12:01,720
its scale. If I think more about it, it feels like we're also stepping

125
00:12:01,720 --> 00:12:09,000
into a whole new territory, ethically speaking right. And I don't shy away from

126
00:12:09,000 --> 00:12:16,640
that like the ethical considerations. I do talk about it in my article. It is

127
00:12:16,640 --> 00:12:20,840
data privacy is like a very important aspect these days with the rise of AI

128
00:12:20,840 --> 00:12:28,760
and they want to be transparent about that in this era of AI driven

129
00:12:28,760 --> 00:12:36,520
marketing. Because while hyper personalization with your coffee sounds

130
00:12:36,520 --> 00:12:44,960
really awesome with your matcha tea sounds really great right. It is a

131
00:12:44,960 --> 00:12:52,960
bit creepy that AI knows so much about you. It's like if I were to even

132
00:12:52,960 --> 00:12:58,680
stop and think about how much like a company would actually know about me.

133
00:12:58,680 --> 00:13:07,400
That can be scary at times. So brands are like walking that tight rope right. How

134
00:13:07,400 --> 00:13:12,400
much information do they want to collect? So now how do you leverage the power of

135
00:13:12,400 --> 00:13:19,200
AI to create these incredible personalized experiences without making

136
00:13:19,200 --> 00:13:23,400
the customers feel like their privacy is being invaded. It's a fine line to walk

137
00:13:23,400 --> 00:13:28,800
for sure. So where do we even begin to draw that line? What are some of the

138
00:13:28,800 --> 00:13:37,280
guidelines or the best practices? You know one of my research has shown it's

139
00:13:37,280 --> 00:13:42,200
a big part of transparency. Companies need to be upfront with their customers

140
00:13:42,200 --> 00:13:47,320
about how they're using their data and give them control over their privacy

141
00:13:47,320 --> 00:13:52,960
preferences. So no more buying those data usage policies and a wall of legal

142
00:13:52,960 --> 00:13:58,900
jargon that nobody really reads right. You ever come across something that

143
00:13:58,900 --> 00:14:03,080
you're trying to accept the terms and conditions of anything. So much legal

144
00:14:03,080 --> 00:14:07,120
stuff there and nobody's even reading that. You just either make a

145
00:14:07,120 --> 00:14:12,560
purchase or you're just trying to do something quick right. That's a

146
00:14:12,560 --> 00:14:19,560
waste. So customers are more likely to trust brands that are open about the AI

147
00:14:19,560 --> 00:14:27,280
practices. So you need to be very clear to the customer what kind of information

148
00:14:27,280 --> 00:14:35,120
you're tracking so then they know what they're signing up for. And maybe that

149
00:14:35,120 --> 00:14:39,440
gives you more leverage maybe because not all companies will do that and

150
00:14:39,440 --> 00:14:48,040
people would get like a... So yeah like I said it's just be open as a company. So

151
00:14:48,040 --> 00:14:53,760
you know the customers will also know more about you. They are more likely to

152
00:14:53,760 --> 00:14:57,920
trust you, buy more from you because you're more open. That means more sales.

153
00:14:57,920 --> 00:15:05,080
You can even raise the prices because of your transparency. Give them that

154
00:15:05,080 --> 00:15:11,280
opportunity to opt out of certain types of data collection or personalization

155
00:15:11,280 --> 00:15:16,200
should they not want it. This is where something like GDPR comes into play

156
00:15:16,200 --> 00:15:21,880
right. Setting important standards for data privacy and the seed of AI. So it's

157
00:15:21,880 --> 00:15:26,880
not just about technology. It's about building trust like I mentioned.

158
00:15:26,880 --> 00:15:32,560
It becomes even more critical the trust that is when these AI powered

159
00:15:32,560 --> 00:15:43,360
experiences become more sophisticated and more integrated into our lives. So it

160
00:15:43,360 --> 00:15:49,720
brings me to other fascinating aspect like how AI is poised to completely

161
00:15:49,720 --> 00:15:55,440
shape the basically reshape the customer experience. So we've talked about AI

162
00:15:55,440 --> 00:16:02,920
working its magic behind the scenes in marketing but let's bring it's you know

163
00:16:02,920 --> 00:16:08,760
let's bring the focus for the average person listening. How AI is going to

164
00:16:08,760 --> 00:16:14,040
change my experience as a customer. That's the million dollar question right.

165
00:16:14,040 --> 00:16:18,640
Like this is where it gets really interesting because AI can make our

166
00:16:18,640 --> 00:16:24,920
interactions with brands not just more efficient but like way more enjoyable.

167
00:16:24,920 --> 00:16:31,800
So if I were to paint you a picture what does that look like in the real world. So

168
00:16:31,800 --> 00:16:38,040
think about time you've been stuck like you know chatting with the really

169
00:16:38,040 --> 00:16:43,680
frustrating customer service chatbot. Try to get answers from it and you know

170
00:16:43,680 --> 00:16:47,000
you've seen those customer service chatbots you interacted with it probably

171
00:16:47,000 --> 00:16:54,880
at some point and you know had a great reaction great interaction right. Feels

172
00:16:54,880 --> 00:16:59,640
like you're just talking to like a brick wall you just like oh I'm sorry I don't

173
00:16:59,640 --> 00:17:05,840
understand oh I'm sorry you don't understand. It's the worst. Those days are

174
00:17:05,840 --> 00:17:09,880
numbered with these large language orders we're talking about AI powered

175
00:17:09,880 --> 00:17:15,120
chat assistants that don't just get your problem they also try to

176
00:17:15,120 --> 00:17:20,080
solve it and they do it fast often without ever having to pull in like a

177
00:17:20,080 --> 00:17:27,600
natural human. So yeah you had those interactions with like oh this

178
00:17:27,600 --> 00:17:32,760
customer tries to when basically you as a customer are trying to interact

179
00:17:32,760 --> 00:17:38,200
with the chatbot and the agent joins in because the chatbot is not able to

180
00:17:38,200 --> 00:17:43,800
answer some of these questions. Yeah those days are definitely numbered.

181
00:17:43,800 --> 00:17:55,040
AI will solve those problems. So I've had multiple cases where I've had to like

182
00:17:55,040 --> 00:18:03,480
repeat what I'm saying a hundred times. We'd be on phone or on chat it's really a

183
00:18:03,480 --> 00:18:08,720
pain and other things to do so does everybody else who does interact with

184
00:18:08,720 --> 00:18:14,040
the chatbot or like a voice assistant on the like a actual human assistant on the

185
00:18:14,040 --> 00:18:28,600
other line. So there's been frustrating times but good news those days are

186
00:18:28,600 --> 00:18:37,520
going to be behind us soon. Let's even go deeper than that like remember how we

187
00:18:37,520 --> 00:18:41,840
were talking about AI mapping out your customer journey the whole

188
00:18:41,840 --> 00:18:46,720
customer journey. Now imagine your chatbot has that map right it knows what you

189
00:18:46,720 --> 00:18:51,320
brought what you looked at even what you asked the customer service about before.

190
00:18:51,320 --> 00:18:56,680
It could know what you need before you even ask making everything smooth and

191
00:18:56,680 --> 00:19:04,880
easy. It's like having a personal assistant who knows all your stuff and

192
00:19:04,880 --> 00:19:09,280
it goes way beyond just customer service. Imagine you're shopping online for some

193
00:19:09,280 --> 00:19:14,480
new shoes right and AI could be working in the background. The website actually

194
00:19:14,480 --> 00:19:19,560
changes how it looks based on what you brought before which you've clicked on.

195
00:19:19,560 --> 00:19:29,800
So AI is more like personalized. It's more personalized to you on the platform

196
00:19:29,800 --> 00:19:40,440
as well. So the website itself is like adapting to you right and that's crazy.

197
00:19:41,400 --> 00:19:46,200
It's the future and get this like it could it could go beyond just suggesting

198
00:19:46,200 --> 00:19:50,680
stuff you might want. It could nail your size perfectly maybe even warn you about

199
00:19:50,680 --> 00:19:55,800
potential issues based on what other customers said. This is like the future

200
00:19:55,800 --> 00:20:00,000
of shopping right now and the best part is this is just the beginning right. It

201
00:20:00,000 --> 00:20:05,600
could only get more finely tuned more intuitive basically a seamless part of

202
00:20:05,600 --> 00:20:12,040
our lives as the tech even gets smarter. We'll just talk about the elephant in the

203
00:20:12,040 --> 00:20:16,080
room just because AI can do all of this doesn't mean it should. As a responsible

204
00:20:16,080 --> 00:20:26,040
tier transparency trust like I spoke about before those are non-negotiable and just

205
00:20:26,040 --> 00:20:31,560
like to reinforce what I said before companies do have to be very upfront

206
00:20:31,560 --> 00:20:41,520
about what they are using AI for. Give customers at any point the control to

207
00:20:41,520 --> 00:20:50,440
opt out if they want to and be ethical about everything. It's a very fine line

208
00:20:50,440 --> 00:20:56,720
that you're walking on. AI that makes things better without going fully creepy

209
00:20:56,720 --> 00:21:04,440
and invasive right and it's about finding that sweet spot where AI just

210
00:21:04,440 --> 00:21:09,720
works. You barely even notice it it's just there making things easier more

211
00:21:09,720 --> 00:21:20,960
enjoyable. So yeah I mean thank you my name is Mukundan again and thank you

212
00:21:20,960 --> 00:21:25,280
this has been incredible. We've gone from just using data to actually getting

213
00:21:25,280 --> 00:21:30,280
insights from it. From marketing to the masters to hyper personalized

214
00:21:30,280 --> 00:21:35,880
experiences clunky old chatbots to AI that's practically reading our minds and

215
00:21:35,880 --> 00:21:40,440
like I mentioned before this is just the tip of the iceberg we're going to see

216
00:21:40,440 --> 00:21:44,640
even wilder wooden back for stuff in the coming years. Whereas we headed to this

217
00:21:44,640 --> 00:21:50,200
uncharted territory it's not just on the brands it's on all of us. We get to

218
00:21:50,200 --> 00:21:57,440
decide how comfortable we are with all of this right. So next time you see one

219
00:21:57,440 --> 00:22:01,680
of these super personalized ads or a website that seems to know you better

220
00:22:01,680 --> 00:22:07,600
than you know yourself ask yourself is this legit is it actually useful and

221
00:22:07,600 --> 00:22:16,200
most importantly how do I feel about how they knew that about me right. It's a

222
00:22:16,200 --> 00:22:21,040
conversation we all need to be part of because in the end the best marketing

223
00:22:21,040 --> 00:22:25,920
the kind that lasts is going to be like a two-way street. It's about balancing the

224
00:22:25,920 --> 00:22:31,680
cool new tech we're doing what's right. Personalization that respects our

225
00:22:31,680 --> 00:22:39,520
privacy and that's a wrap on this deep dive into AI and LLMs in marketing. I

226
00:22:39,520 --> 00:22:45,080
hope you enjoyed it. Stay curious stay informed and until next time keep those

227
00:22:45,080 --> 00:22:56,520
questions coming. Thank you.