1
00:00:00,000 --> 00:00:20,560
Welcome everybody. So we're going to be talking canvas data today. So I guess just to, I guess

2
00:00:20,560 --> 00:00:27,440
we've got Cecilia here today. So I guess I'll introduce myself then I guess we could go around

3
00:00:27,440 --> 00:00:34,880
and just do a quick introduction to each other. So my name is Keegan and I've been in the cannabis

4
00:00:34,880 --> 00:00:41,600
space for a little while. My background is in economics and so it's fun to apply economics,

5
00:00:41,600 --> 00:00:50,160
data science to problems in the cannabis industry. And so I've coincidentally started a company,

6
00:00:50,160 --> 00:00:57,280
Canelytics, that does exactly that. And so we help people in the cannabis industry make sense of their

7
00:00:57,280 --> 00:01:07,040
data, wrangle data, and get value from their data. But enough of me, I'll let Charles and Paul and

8
00:01:07,040 --> 00:01:16,480
then you as well Cecilia introduce yourselves to each other. I'm Charles and I have 27 years of

9
00:01:16,480 --> 00:01:24,720
software development experience and I've been working in data science for the last year and a

10
00:01:24,720 --> 00:01:40,160
half and I guess Cecilia is gone so we can skip that. No, do go on Charles. I didn't know you

11
00:01:40,160 --> 00:01:48,000
had 27 years experience, that's quite a bit. Yeah, it's been a long, long time of programming.

12
00:01:53,760 --> 00:01:59,920
And if you can share, how goes your forays into machine learning?

13
00:01:59,920 --> 00:02:12,960
It's going pretty good. I've been making some progress in the Canel competition. I'm at 97%

14
00:02:13,680 --> 00:02:20,320
and the leaders are at 98, 99% so I just got to eek that little extra percentage out there.

15
00:02:20,800 --> 00:02:26,560
Good for you, that's good. Thanks. And I found out, actually somebody told me about this a couple

16
00:02:26,560 --> 00:02:40,160
months ago, and I tried it last night, there's this library called Pandas Profiling and you can

17
00:02:40,160 --> 00:02:46,880
give it a data frame and it will tell you about each one of the variables or each one of the

18
00:02:46,880 --> 00:02:52,960
columns like how many missing values there are, what percentage of, you know, what the most common

19
00:02:52,960 --> 00:03:00,880
values are in it. It will give you a correlation matrix. Wow. It just does all these things and

20
00:03:00,880 --> 00:03:08,640
it'll export it to an interactive HTML file. So you can bring it up in a browser and just

21
00:03:10,320 --> 00:03:16,720
click on different buttons and tabs and stuff and look at it. It's amazing.

22
00:03:16,720 --> 00:03:22,480
It sounds pretty powerful. It is.

23
00:03:22,480 --> 00:03:31,600
Welcome, Sean. We're just sort of introducing each other at the moment, or introducing ourselves.

24
00:03:32,240 --> 00:03:40,560
So my name is Keegan, this is Charles, Paul. We meet up regularly to talk about cannabis data.

25
00:03:40,560 --> 00:03:46,560
So we're happy to have you. Thank you. Can you hear me, guys?

26
00:03:46,560 --> 00:03:48,000
All right. Barely.

27
00:03:51,280 --> 00:03:52,720
How about that? Much better.

28
00:03:53,280 --> 00:04:03,040
Perfect. Well, nice to meet you all. Yeah, I guess I'll introduce myself. So my name is Sean.

29
00:04:03,040 --> 00:04:10,320
I've worked in the cannabis industry here in Denver and Northern Colorado

30
00:04:12,080 --> 00:04:18,960
for about seven years. And I just recently finished a data science program

31
00:04:20,160 --> 00:04:28,720
to kind of upskill myself. And so I'm just looking for any opportunity to apply, you know,

32
00:04:28,720 --> 00:04:33,280
my cannabis experience with this newfound skill of data science that I have.

33
00:04:36,400 --> 00:04:38,320
Well, you found the right group.

34
00:04:38,320 --> 00:04:42,880
Perfect. Is everyone in Colorado?

35
00:04:44,160 --> 00:04:52,640
We're scattered actually in various states. So I found a company, CanLinic,

36
00:04:52,640 --> 00:05:01,920
that's based out of Olympia, Washington. Charles is in Oregon, and Paul is in Michigan.

37
00:05:01,920 --> 00:05:07,920
So we sort of represent cannabis markets all across the country.

38
00:05:08,960 --> 00:05:16,000
We'll be in Colorado Springs over the July 4th week to visit some friends out there. So we'll

39
00:05:16,000 --> 00:05:20,080
be too far away from you, Sean. Yeah, nice. That's a good time to visit.

40
00:05:20,080 --> 00:05:20,720
Yeah.

41
00:05:32,000 --> 00:05:40,800
Can you, I guess, do you have like any topics of interest that are like, so what like are people

42
00:05:40,800 --> 00:05:54,000
talking about in Colorado? If I can ask. Yeah, Colorado, last year passed legislation for

43
00:05:54,800 --> 00:06:01,440
cannabis clubs, basically a bar you can go to and smoke up. And that's been the buzz lately.

44
00:06:01,440 --> 00:06:11,920
So I don't know how that plays in the analytics. I actually have some data from a dispenser I worked

45
00:06:11,920 --> 00:06:18,000
at that I wanted to see if I could get some advice from you guys on how to crunch these numbers,

46
00:06:18,000 --> 00:06:27,440
or if you know any algorithms to do some. My idea was to do like a sort of practice with some time

47
00:06:27,440 --> 00:06:33,200
series prediction and use it's about five months worth of sales data and see if I could use that

48
00:06:33,200 --> 00:06:37,520
five months. I know it's not perfect because not a full year is worth the sales data, but

49
00:06:38,160 --> 00:06:42,320
at least try and play with some algorithms to get some time series predictions.

50
00:06:44,560 --> 00:06:53,520
Yeah. Keegan, do you mind if I step in? Yeah, so Sean, I'm just actually wrapping up a master's

51
00:06:53,520 --> 00:06:58,240
program in data science from University of Wisconsin and I'm working on my capstone project

52
00:06:58,240 --> 00:07:05,520
right now. Thanks to Charles and Keegan, I was able to get hold of some Washington state sales data.

53
00:07:07,040 --> 00:07:14,800
What I'm doing right now is going to be using some market basket analysis on some sales

54
00:07:14,800 --> 00:07:24,640
information to see if I could make some sales recommendations from dispensaries to customers.

55
00:07:25,600 --> 00:07:33,120
What I really wanted to do was do a project on sequential pattern mining, which kind of is the

56
00:07:33,120 --> 00:07:37,680
same thing as market basket analysis where it makes recommendations, but it makes recommendations

57
00:07:37,680 --> 00:07:46,000
over time. I just found out yesterday, last night, unfortunately, I was going through the data and

58
00:07:47,920 --> 00:07:54,080
the sequential patterns I was looking to uncover was between the suppliers or the producers,

59
00:07:54,080 --> 00:07:58,640
the product producers and the dispensaries. So the product producers could say, okay,

60
00:07:58,640 --> 00:08:03,760
you bought these products from us. Maybe you could buy product X and Y as well.

61
00:08:03,760 --> 00:08:09,280
What I found out though is those transactions that are recorded in the Washington state data

62
00:08:10,000 --> 00:08:15,520
are what the producers are doing is they're recording their transactions in a big batch.

63
00:08:16,160 --> 00:08:20,480
So instead of every time there's a transaction, they record it, they just pile up all the

64
00:08:20,480 --> 00:08:27,360
transactions to enter them in one day. So I can't actually use that data. So the data is pretty bad

65
00:08:27,360 --> 00:08:35,760
in that respect. But I did notice that the sales transactions are definitely spread out over time

66
00:08:36,320 --> 00:08:41,760
and I can use a market basket analysis approach. So it sounds like you're trying to do something

67
00:08:41,760 --> 00:08:47,040
like with retail sales as well, but you want to do more of what? Trending, sales trending or

68
00:08:47,040 --> 00:08:54,160
something like that? Yeah, just kind of a way to see if I can use the data I have to

69
00:08:54,160 --> 00:09:01,120
kind of predict the next week's sales or the next month's sales. So some time series,

70
00:09:02,000 --> 00:09:07,920
like a halt winters or some seasonal type forecasting. Yeah, exactly. And unfortunately

71
00:09:07,920 --> 00:09:13,200
it's only like five months worth of data. So I can't get that full year spectrum in there,

72
00:09:13,200 --> 00:09:18,640
but it does include 420. So that kind of adds an interesting outlier in there.

73
00:09:18,640 --> 00:09:25,840
Okay, so I don't know what that means. What does 420 mean? Oh, 420, the weed holiday that everyone

74
00:09:25,840 --> 00:09:32,160
goes and buys up the most weed. Okay, I didn't even know about that. Yeah, he told him I'm green

75
00:09:32,160 --> 00:09:41,200
to the space, pun intended. Everyone has their big sale. Okay. Yeah, it's like the Black Friday of

76
00:09:41,200 --> 00:09:50,400
the weed industry. Gotcha. Okay. Oh, that's interesting. Cool. Yeah, I can pull that up if

77
00:09:50,400 --> 00:09:55,760
you guys give me a moment and then maybe you can give me your two cents on the data. Unfortunately,

78
00:09:55,760 --> 00:10:04,400
it doesn't specify product. It just has the dollar amounts and then the time and then

79
00:10:04,400 --> 00:10:10,400
I've redacted it because it had all the employee names. So I just changed the employee names to

80
00:10:10,400 --> 00:10:15,920
employee one, two, three, et cetera, et cetera. So is this from their point of sale system at the...

81
00:10:15,920 --> 00:10:23,440
Okay. Yeah. It's using metric. That's the standard here in Colorado for POS systems.

82
00:10:25,440 --> 00:10:28,880
But yeah, just give me a moment. I'll pull that up for us to check out.

83
00:10:28,880 --> 00:10:36,560
Just thoughts right off the bat. If you've got daily data, you could start uncovering...

84
00:10:37,840 --> 00:10:46,960
Some of my favorite things to look at are day of the week effects. So what day are you having the

85
00:10:46,960 --> 00:10:56,880
most sales? I have done some analysis like that and pinpointed which hours of the day you're

86
00:10:56,880 --> 00:11:04,560
going to do the hottest and which days of the month are hottest. So I have been able to do some

87
00:11:04,560 --> 00:11:12,880
analysis there, but at the end of the day, it just seems more like an HR tool to motivate employees

88
00:11:12,880 --> 00:11:20,240
to sell more or something. That's kind of the gist I get from it. It's not really like predictive.

89
00:11:20,240 --> 00:11:28,800
It's more like management tool or planning tool. Exactly. For essentially optimal staffing.

90
00:11:28,800 --> 00:11:40,880
Yeah, exactly. Yeah. The way I would go about it is you'd partition your man hours. So it's essentially

91
00:11:40,880 --> 00:11:50,400
if you have 12 employees, then they're each working 30, 40 hours a week. You would just have your

92
00:11:50,400 --> 00:11:56,880
total hours and then you'd almost want to match that with... If you're doing hourly, then you'd

93
00:11:56,880 --> 00:12:07,680
almost want to weight your staffing in accordance to your predicted sales spikes. If you're doing

94
00:12:07,680 --> 00:12:15,280
that correctly, then you'd have your total hours and then you'd almost want to match that with

95
00:12:15,280 --> 00:12:26,240
your predicted sales spikes. To the best you can. But if you can share, what were some of the...

96
00:12:26,240 --> 00:12:42,320
I'm trying to remember where I put that in the analysis. This is sort of what Paul and I talk

97
00:12:42,320 --> 00:12:51,840
about is there's so many opportunities here and I'm sure people have done them. So you're doing

98
00:12:51,840 --> 00:12:57,520
them yourselves, but there's not that much out there in the public. So there's not that much

99
00:12:57,520 --> 00:13:04,160
public research about, or at least that I've found. Point me in the direction if I'm wrong,

100
00:13:04,160 --> 00:13:15,200
but it's research about say the busiest times of the day. So I don't know. There's just lots

101
00:13:15,200 --> 00:13:21,920
of opportunities for you to get your research out there and people to get some eyeballs on it.

102
00:13:21,920 --> 00:13:27,280
What would you say is a good platform for getting that out there?

103
00:13:30,000 --> 00:13:33,280
Well, I may be biased, but at the moment,

104
00:13:33,280 --> 00:13:40,160
Candlitics is essentially trying to provide data to people in the cannabis industry. So

105
00:13:40,160 --> 00:13:48,400
it's still a seedling company. I'm always just a big fan of self-publishing. So if you have

106
00:13:48,400 --> 00:13:56,800
your own website, put it on your own website. GitHub. Yeah, I've done Medium and GitHub.

107
00:13:59,440 --> 00:14:07,040
And that's about it so far, but I'm always... Or LinkedIn. So I'm just always looking for

108
00:14:07,040 --> 00:14:14,640
new platform really. So that's essentially what Candlitics is. So if you have something

109
00:14:14,640 --> 00:14:20,720
in particular you want to get out there, we can send me a message and we can

110
00:14:22,240 --> 00:14:28,560
try to get your data and your analysis out there. Great. Yeah. And just to be fully transparent,

111
00:14:28,560 --> 00:14:39,280
my main motivation is to find a job in data science. Same here. Specifically in the cannabis

112
00:14:39,280 --> 00:14:47,360
industry or just any... In general, but I think my domain knowledge would be best supplied in

113
00:14:47,360 --> 00:14:55,520
the cannabis industry, but I know data science applies to anything really. Yeah. So for Charles

114
00:14:55,520 --> 00:15:01,680
and Sean, if you guys are both looking for work, I work for General Motors here in Southeast

115
00:15:01,680 --> 00:15:10,320
Michigan and we have this big push going on right now as far as data science analytics,

116
00:15:10,320 --> 00:15:16,240
because it's really driving all parts of the company. So I can probably send you guys a link

117
00:15:16,240 --> 00:15:23,840
to our job posting board. I can't guarantee anything. The good thing is now that we're

118
00:15:23,840 --> 00:15:30,880
doing a lot of our positions, especially positions like data science remotely. So take a look at the

119
00:15:30,880 --> 00:15:36,400
job postings. I can't guarantee anything, but I can always reach out to the hiring manager that

120
00:15:36,400 --> 00:15:41,440
you'd be applying to and put in a word for you. But yeah, I'll send that link to both of you if

121
00:15:41,440 --> 00:15:49,280
you're interested. Yeah. That would be awesome. Great. And I got those charts. You guys still

122
00:15:49,280 --> 00:15:57,840
want to take a look at them? You're welcome to share your screen if you're able. Okay. Yeah,

123
00:15:57,840 --> 00:16:04,720
I'm just not familiar with this one. The screen share is present now.

124
00:16:04,720 --> 00:16:13,280
Share your screen.

125
00:16:23,200 --> 00:16:26,560
Does it require permission or something?

126
00:16:26,560 --> 00:16:35,120
Okay.

127
00:16:47,760 --> 00:16:50,080
Does it seem like I need to require permissions?

128
00:16:50,080 --> 00:16:58,640
Yes. Yeah, on the share button, it's all blacked out. So it seems... Oh no, sorry. That's my fault.

129
00:17:00,320 --> 00:17:05,200
Okay. Is that working now? Yeah, we can see you, Sean.

130
00:17:05,920 --> 00:17:10,240
Okay. So how do I drive this?

131
00:17:10,240 --> 00:17:23,360
So is it moving? Yes. Okay. So this came from... I'll show you the data real quick. So this is kind

132
00:17:23,360 --> 00:17:31,680
of the raw format of the data. So we have our price, the tax, and then total price and total paid are

133
00:17:31,680 --> 00:17:40,080
the same numbers. So I eliminated one of those columns. Register, this column was kind of

134
00:17:40,080 --> 00:17:46,960
unnecessary data. So I eliminated that as well. And this is where I redacted the employee names

135
00:17:46,960 --> 00:17:54,160
from their actual name to E5 or E1, et cetera, et cetera. And this was the time of date. This was

136
00:17:54,160 --> 00:18:07,120
taken back in 2019 as of the first of the year, all the way to June 2nd. So roughly five months.

137
00:18:07,120 --> 00:18:14,560
And then this was the total gross revenue for the dispensary within that five-month time frame.

138
00:18:14,560 --> 00:18:20,480
Quick idea. I don't know how much of interest it is to you, but you could potentially

139
00:18:21,280 --> 00:18:28,080
go to the dispensary or a similar dispensary with this idea is you could basically try to...

140
00:18:28,080 --> 00:18:37,200
I think... Sorry, I was just thinking for a second. We'll have to maybe pinpoint the specific

141
00:18:37,200 --> 00:18:43,280
regression, but you could essentially run a regression and try to pinpoint essentially

142
00:18:43,280 --> 00:18:50,320
if any of the... Like, if you're going to run a regression, you could run a regression and

143
00:18:50,320 --> 00:18:59,360
you could essentially run a regression and try to pinpoint essentially if any of the salespeople

144
00:19:00,000 --> 00:19:11,440
have above average sales. And I did something like that. So I'll go into some of that right now.

145
00:19:11,440 --> 00:19:20,480
So this was my first initial analysis. This is in a Google Collab notebook running on Python.

146
00:19:22,160 --> 00:19:30,160
So the original data set, I'm changing some of the column names here. And then this is

147
00:19:30,960 --> 00:19:38,000
the pie chart indicating which employees had the most sales. Now, keep in mind that not all employees

148
00:19:38,000 --> 00:19:46,640
are strictly doing sales work like this E10 and E9. They were mostly front desk checking IDs and

149
00:19:46,640 --> 00:19:57,440
such. But as you can see, E7, E1, E2 had the lion's share of the sales. And then in another

150
00:19:59,120 --> 00:20:03,840
little analysis. So this is where I was able to pinpoint

151
00:20:03,840 --> 00:20:08,960
which hour of the day and which day of the month had the highest sales.

152
00:20:10,400 --> 00:20:20,800
So based on... And forgive me, it's been a while since I looked at this. But this would have...

153
00:20:20,800 --> 00:20:30,080
I believe this is where I pinpointed. So the Y column is... Oh, these numbers don't make sense now.

154
00:20:31,840 --> 00:20:39,840
Sorry about that, guys. Yeah, it's been a while since I looked at these. But what I was trying to

155
00:20:39,840 --> 00:20:50,560
do is I'm going to look at the numbers. So I'm going to look at the numbers. So I'm going to

156
00:20:50,560 --> 00:20:58,560
do it as where it's blackest, it's the highest average sale relative to all the other days.

157
00:20:59,520 --> 00:21:07,600
Here's the way that you could potentially simplify this to make it meaningful.

158
00:21:09,600 --> 00:21:13,840
Or more meaningful. Or meaningful in a different way. So I was thinking,

159
00:21:13,840 --> 00:21:23,520
you could almost lump the times of day. So the simplest would just be a dichotomous morning

160
00:21:23,520 --> 00:21:33,920
or afternoon. And that way you could compare the two. So you could see, okay, what's the average

161
00:21:33,920 --> 00:21:44,080
sales before 12 PM? And what's the average sales after 12 PM? I'm not sure.

162
00:21:47,040 --> 00:21:51,920
I mean, they may be the exact same, but it would be interesting to see if, I don't know, maybe

163
00:21:51,920 --> 00:21:59,200
average receipt goes up or down. And if so, by how much?

164
00:21:59,200 --> 00:22:06,160
Yeah, I had something like that where I compared each employee to the average ticket price,

165
00:22:06,160 --> 00:22:14,880
average sale price, or average total receipt. Just kind of comparing that way. So I can

166
00:22:14,880 --> 00:22:34,560
cruise around and see if I can find it. But yeah, that's kind of the extent of what I had.

167
00:22:34,560 --> 00:22:45,440
Just curious, Sean, do you know that you friends with the owner of the dispensary?

168
00:22:45,440 --> 00:22:57,120
So the, no, well, yes, but they sold. So I caught wind of that and I tried to

169
00:22:57,120 --> 00:23:06,960
get some financing together to kind of buy them out. And I actually got three million

170
00:23:07,840 --> 00:23:13,040
lined up to make an offer, made an offer, and they laughed in my face.

171
00:23:15,840 --> 00:23:20,240
Three million? Oh my gosh. I'm really lowballing you, I guess. I don't know.

172
00:23:20,240 --> 00:23:27,360
But yeah, they rejected my offer and I'm pretty sure what they got was closer to 10 million.

173
00:23:28,080 --> 00:23:35,120
Wow. So maybe you guys have a better perspective on this than I do, but I'm sure you do,

174
00:23:35,120 --> 00:23:39,920
actually. But when it comes to dispensaries, they're all cash businesses, right?

175
00:23:40,960 --> 00:23:46,080
Yeah, they're progressively, you know, debit cards and credit cards are working their way in

176
00:23:46,080 --> 00:23:51,360
there, but it's through loopholes. It's not really anything official.

177
00:23:51,360 --> 00:23:56,240
Dealing with that much cash on hand must be a real pain for the businesses to deal with, huh?

178
00:23:57,120 --> 00:24:07,040
Yeah, it had its own challenges. We had two gun safes on site, and that's where our petty cash

179
00:24:07,040 --> 00:24:16,320
and our product was stored. Even then, it got robbed twice while I was working there.

180
00:24:16,320 --> 00:24:25,520
I went there for over a year, but both those robberies were kind of like almost joke instances.

181
00:24:25,520 --> 00:24:30,720
Like one, they went into the grow room and then just literally dragged the plants out by their

182
00:24:30,720 --> 00:24:36,960
stems with a trail of dirt following them. And then they were like, oh, my gosh,

183
00:24:36,960 --> 00:24:43,200
they're down the block. And another one, they did a smash and grab after hours, broke through the

184
00:24:43,200 --> 00:24:51,120
front door, went into the blood tending room, and then just took all the concentrates and then tried

185
00:24:51,120 --> 00:24:56,560
to resell them. And then they were caught reselling them because they had all the initial RFID tags

186
00:24:56,560 --> 00:25:05,520
and all that stuff. So, it's a bunch of jokesters, really. But still, I mean, that kind of goes to

187
00:25:05,520 --> 00:25:13,360
show you that it's still a threat of safety. And I was thinking that would increase costs

188
00:25:13,360 --> 00:25:20,960
across the board, whether you get robbed or not, because now everybody has to spend a small fortune

189
00:25:20,960 --> 00:25:32,880
on security cameras, security doors. I think they use the armored cars for the transportation.

190
00:25:32,880 --> 00:25:43,600
So, there's all sorts of extra expenses involved. For context, this owner, she was really

191
00:25:44,880 --> 00:25:50,880
not privy to spending money on the business. She just wanted to collect as much,

192
00:25:53,600 --> 00:26:02,080
I guess, margin from the revenue as possible. So, anytime where she could skip out. I mean,

193
00:26:02,080 --> 00:26:09,360
we had security cameras all over the place. But when it came to reinforcing doors or anything

194
00:26:09,360 --> 00:26:15,520
like that, she didn't care. She just ate that minor cost compared to whatever costs that

195
00:26:15,520 --> 00:26:21,760
reinforced the doors. So, she's just selling out and leaving the industry or she thinks she can

196
00:26:21,760 --> 00:26:30,560
reinvent? Yeah. That's my impression. She got started when it was just the people who were

197
00:26:30,560 --> 00:26:38,400
growing in the basement that had the willingness to jump into this industry. And that crowd is

198
00:26:38,400 --> 00:26:43,520
kind of getting phased out now. And then the market's consolidating in Colorado to more

199
00:26:43,520 --> 00:26:52,800
chain type operations. But she is kind of like the mom and pop type of business owner and didn't

200
00:26:52,800 --> 00:26:59,440
really have ambitions to kind of grow it. Just kind of walk away and cash out and walk away.

201
00:26:59,440 --> 00:27:02,400
And she cashed out pretty well. I mean, you're not.

202
00:27:03,200 --> 00:27:08,720
Yeah. So, you talked about consolidation. I think we've talked about that on this call a

203
00:27:08,720 --> 00:27:17,200
couple of times, but that seems like that's inevitable, right? I mean, every industry

204
00:27:17,200 --> 00:27:23,920
goes through that phase. And the intersection of data and consolidation, I wonder what kind of

205
00:27:23,920 --> 00:27:27,040
opportunities I'm just kind of throwing this out as a brainstorming idea, but

206
00:27:27,040 --> 00:27:33,680
market consolidation and data, right? So, what could be some opportunities there? I mean,

207
00:27:33,680 --> 00:27:38,080
obviously every business, if you're going to get bought up by somebody else, you got to have your

208
00:27:38,080 --> 00:27:43,280
numbers, right? You got to have your typical finances in line, your POS and everything else.

209
00:27:43,280 --> 00:27:51,360
But can you guys think of anything that might be useful to large kind of, not corporate,

210
00:27:51,360 --> 00:27:55,680
but maybe corporate, yeah, entities that are going to consolidate and pull the industry together?

211
00:27:55,680 --> 00:28:04,240
Well, one data source that I wanted to play with was, at least in Colorado, all license owners are

212
00:28:04,240 --> 00:28:12,800
listed publicly. So, to take that list of license owners and see whether it has shrunk or whether

213
00:28:12,800 --> 00:28:21,920
there's more licenses, but fewer owners, just kind of get some preliminary analysis from that,

214
00:28:21,920 --> 00:28:26,480
just to see, kind of measure that consolidation that you're mentioning.

215
00:28:26,480 --> 00:28:29,920
Yeah. So, kind of have an idea of what's on the horizon or where it's trending. Yeah.

216
00:28:33,440 --> 00:28:38,640
I'm sure it's just the same in the other states. All license owners are listed on a state list

217
00:28:38,640 --> 00:28:48,320
somewhere. Yeah. It's hard to, I know in Michigan, it's not like Washington, right? Where Keegan was

218
00:28:48,320 --> 00:28:54,640
able to supply some of the state data for Washington, but it's, Keegan, you're saying

219
00:28:54,640 --> 00:29:00,080
that Washington state has like a very liberal Freedom of Information Act, and that's how

220
00:29:00,080 --> 00:29:07,280
the data came about. Yes. However, Colorado does do a good job about publishing data. And in fact,

221
00:29:07,280 --> 00:29:16,480
Yeah, we do. When, I have done a little analysis of Colorado data, it may be a little tricky, but

222
00:29:16,480 --> 00:29:23,280
so one thing you can do with their data, so they publish, I think the licensee lists monthly.

223
00:29:24,160 --> 00:29:33,120
So you can actually, you can track exits and entries. So you can say, oh, this person's

224
00:29:33,120 --> 00:29:44,080
entered. And then at a certain point, they've exited the market. So it's an interesting thing

225
00:29:44,080 --> 00:29:52,720
to track because say your licensees, say they increased by 50 in the month. Well, there may

226
00:29:52,720 --> 00:30:05,200
have been 80 entrants and 30 exits. And so in the Colorado data, you can actually track, you know,

227
00:30:05,200 --> 00:30:13,520
who's entering and who's leaving, you know, when they do it. It's not the easiest analysis in the

228
00:30:13,520 --> 00:30:20,480
world, but it is possible. Yeah, that would offer a lot of good insight to see what the average time

229
00:30:21,200 --> 00:30:29,120
or lifespan is for a license holder. At least the license holders that are seeing turnover versus

230
00:30:29,120 --> 00:30:39,440
the ones that are sticking around. Exactly. So I'll have to, maybe next week, I can share this

231
00:30:39,440 --> 00:30:44,080
analysis with you. So I'll share what I've done with Colorado. So I was just doing

232
00:30:45,440 --> 00:30:57,440
entries and exits for each month. And then it would be a little more detailed, but I think

233
00:30:57,440 --> 00:31:04,000
like you said, you could do essentially, I think in economics, they would call that survival analysis.

234
00:31:04,640 --> 00:31:11,840
So, you know, how long are these companies surviving on average?

235
00:31:15,840 --> 00:31:20,880
I think initially that would be, that would have been a, because in Colorado is kind of

236
00:31:20,880 --> 00:31:24,240
an opportunity for a slow growth type of business model.

237
00:31:24,240 --> 00:31:30,480
So I bet initially that was kind of a longer sort of timeframe. Whereas now

238
00:31:31,360 --> 00:31:36,800
you don't have, you know, two million at least in capital to start something up, then

239
00:31:37,600 --> 00:31:48,400
you kind of have to have to work. What about Keegan, with all your background in the lab

240
00:31:48,400 --> 00:31:54,560
testing environment, just curious to start up a lab, that must be a huge investment, right?

241
00:31:55,520 --> 00:32:03,200
Yes. So from what I've heard, it seems you're not going to really start up a lab for less than

242
00:32:03,200 --> 00:32:11,840
about, you know, a million dollars in investment. And the more I learn about labs, it seems that

243
00:32:11,840 --> 00:32:19,040
maybe be low. So maybe one to two million, just to get all your standard instruments

244
00:32:19,680 --> 00:32:28,800
and get them calibrated, get your team of scientists do testing. However, that may be,

245
00:32:29,600 --> 00:32:37,600
you know, peanuts compared to what some of these cultivators are spending on their companies.

246
00:32:37,600 --> 00:32:44,640
So that's one thing, like you'll see at CannaCon, there were like some, you know, people who were

247
00:32:44,640 --> 00:33:01,920
setting up some high dollar processing facilities, as well as cultivations. So I think the people who

248
00:33:01,920 --> 00:33:09,440
are seriously going after large scale production, I think they're at much higher costs, but it's

249
00:33:09,440 --> 00:33:19,200
definitely not trivial to set up a lab. And I would actually think just because of the some of the,

250
00:33:19,200 --> 00:33:25,680
you know, the hurdles you have to jump, you know, getting your scientific team, so you've got to get

251
00:33:25,680 --> 00:33:34,800
people with masters, preferably PhDs in chemistry or microbiology. So those are tough requirements

252
00:33:34,800 --> 00:33:46,000
to meet, even if they're not nominally cost. Yeah, yeah, that was a big thing in Oregon. We had

253
00:33:46,000 --> 00:33:51,840
legalization, and then they had all these really stringent testing requirements. But there were no

254
00:33:51,840 --> 00:33:56,400
labs. And then people thought, oh, I'll make a bunch of money. So they, you know, they bought a

255
00:33:56,400 --> 00:34:00,400
bunch of lab equipment. But yeah, then they couldn't find people to run the lab equipment.

256
00:34:00,960 --> 00:34:07,360
There was this huge backlog, and they had to sort of relax the requirements until the labs caught up

257
00:34:08,000 --> 00:34:13,200
to, you know, to be able to do the actual testing that they wanted.

258
00:34:14,400 --> 00:34:19,120
I mean, that's interesting, because as we talk through these different kind of pain points,

259
00:34:19,120 --> 00:34:25,840
there's an opportunity right there. So my brother-in-law, he does medical hiring, right,

260
00:34:25,840 --> 00:34:33,840
for nurses. And so it's a whole industry on this type of thing. Well, there's an industry,

261
00:34:33,840 --> 00:34:38,400
what you just stated there, Charles, there's a whole industry potentially there, right? How do you

262
00:34:38,400 --> 00:34:45,360
find the right people to man these labs? That's just amazing that all through this chain, this

263
00:34:45,360 --> 00:34:50,240
kind of value chain, there's all kinds of different things you could approach and, you know, try

264
00:34:50,240 --> 00:34:51,280
something with.

265
00:34:53,440 --> 00:35:02,560
Well, it honestly seems like every lab is looking for a good laboratory director, and or quality

266
00:35:02,560 --> 00:35:09,040
assurance manager, because it's almost one of the things where if you have a good enough quality

267
00:35:09,040 --> 00:35:15,440
assurance manager on the market, you may as well get two, if you can find them, because they're

268
00:35:15,440 --> 00:35:22,400
that valuable and hard to find. Same for scientific director slash laboratory directory. Like if

269
00:35:23,280 --> 00:35:31,520
there's two that you can get, like you'd almost want two. But like, you were all hitting on,

270
00:35:32,320 --> 00:35:35,280
it's tough to find them and connect the dots, because

271
00:35:35,280 --> 00:35:43,920
apparently if you've got a PhD in chemistry or, you know, there's a high demand for your skills.

272
00:35:43,920 --> 00:35:44,420
Yeah.

273
00:35:50,480 --> 00:35:57,520
And then the other thing with the cannabis industry is it's tough to find people that are,

274
00:35:58,320 --> 00:36:00,880
I guess, on board to work in the cannabis industry.

275
00:36:00,880 --> 00:36:07,520
So you find some people that are enthusiastic to work there. However, just to be frank,

276
00:36:07,520 --> 00:36:15,520
there's still a little shunned in the scientific community. So just, you know, from word of mouth,

277
00:36:15,520 --> 00:36:23,600
I have heard that, you know, it can be tricky to, to maybe move into other industries and other

278
00:36:23,600 --> 00:36:32,720
lines of work if your background is at a lab that test. I imagine as time changes, probably in the

279
00:36:32,720 --> 00:36:37,520
near future, right, especially as more and more money comes into this industry. I think this can

280
00:36:37,520 --> 00:36:45,200
probably be less and less of an issue. Yes, I speak to that. I have two resumes, one with my

281
00:36:45,200 --> 00:36:47,200
cannabis experience and one without.

282
00:36:47,200 --> 00:36:52,240
Well, if you're looking for something at GM, send them on without because there's a bunch of

283
00:36:52,240 --> 00:36:57,760
old timers in this, in this company. I swear these engineers, I get emails all the time that say,

284
00:36:57,760 --> 00:37:02,160
you know, somebody's heading out the door of retirement, you know, with 42 years of service,

285
00:37:02,160 --> 00:37:06,480
and I'm just blown away. I've been with the company six years, but these people have been

286
00:37:06,480 --> 00:37:12,800
here since forever. These old engineers, it's amazing. And I'm just, I'm just, I'm just,

287
00:37:12,800 --> 00:37:24,640
and then, yeah, so I, yeah, kind of got a feel for what I was

288
00:37:24,640 --> 00:37:29,680
trying to show you guys what was better. So I'll just clear that up real quick if you don't mind.

289
00:37:31,200 --> 00:37:38,900
Yeah, so for the Internship Luolun is all in control of CAMP, the

290
00:37:38,900 --> 00:37:42,000
CAP, and we used there, I guess what Will and El rode there for course,

291
00:37:42,000 --> 00:37:51,000
So this is the x-axis is a minute of the hour.

292
00:37:51,000 --> 00:37:57,000
And so the y is day of the month.

293
00:37:57,000 --> 00:38:09,000
So on let's say the 27th minute in the 27th day,

294
00:38:09,000 --> 00:38:15,000
which sounds weird, but the average ticket price was $86.

295
00:38:15,000 --> 00:38:23,000
So that'd be, I guess, a way to pinpoint, you know, target those times.

296
00:38:23,000 --> 00:38:29,000
So guessing. Oh, so never mind.

297
00:38:29,000 --> 00:38:33,000
That's the average of every minute within each hour.

298
00:38:33,000 --> 00:38:38,000
OK. Why don't you you can bin it like by the hour.

299
00:38:38,000 --> 00:38:42,000
Yeah, I think I was trying to do that in my function here.

300
00:38:42,000 --> 00:38:51,000
So the column is based on the aggregating function was the average price.

301
00:38:51,000 --> 00:38:57,000
I thought it was a little cleaner than that.

302
00:38:57,000 --> 00:39:03,000
I know sort of taking big averages does lose your granular data,

303
00:39:03,000 --> 00:39:11,000
but I just like to start with just a simple snapshot. So I love the hourly minutes.

304
00:39:11,000 --> 00:39:19,000
I think just looking at the date, the day of the week's would be useful

305
00:39:19,000 --> 00:39:27,000
and then then getting more granular as you go, because I think people have expectations that,

306
00:39:27,000 --> 00:39:36,000
oh, yes, probably Friday, Saturday are busy days. But, you know, to what extent?

307
00:39:36,000 --> 00:39:42,000
And like our Sunday's busy. Right.

308
00:39:42,000 --> 00:39:49,000
And one thing I wish one data data point that I wish we took was

309
00:39:49,000 --> 00:39:53,000
location of drivers license we looked at.

310
00:39:53,000 --> 00:40:00,000
Because from looking at all the IDs at every customer that walked in,

311
00:40:00,000 --> 00:40:04,000
I could, you know, I was working in dispensary in northern Colorado.

312
00:40:04,000 --> 00:40:10,000
So we were pretty close to Wyoming and, you know, just ballparking from the IDs that I saw.

313
00:40:10,000 --> 00:40:15,000
I would say, you know, 20 percent of our customers came, drove down from Wyoming,

314
00:40:15,000 --> 00:40:21,000
which is which cannabis is still illegal there, but they didn't care.

315
00:40:21,000 --> 00:40:27,000
And the cops weren't really pulling people over, driving back from Colorado into Wyoming.

316
00:40:27,000 --> 00:40:32,000
And so I think one interesting data point would be, you know, where are people coming from?

317
00:40:32,000 --> 00:40:38,000
Because we would get people that would say, oh, I come from North Dakota once a month

318
00:40:38,000 --> 00:40:43,000
just to stock up on my stuff and or Texans.

319
00:40:43,000 --> 00:40:49,000
I always love selling to Texans because they always had the highest receipts.

320
00:40:49,000 --> 00:40:57,000
They would spend easily over three hundred dollars for each tourist from Texas.

321
00:40:57,000 --> 00:41:02,000
So, yeah, we've hit on these good uses of these

322
00:41:02,000 --> 00:41:06,000
or just these ways that these data points can be used in unexpected ways.

323
00:41:06,000 --> 00:41:12,000
So, like you said, I think Colorado.

324
00:41:12,000 --> 00:41:18,000
Correct me if I'm wrong, but they may even break down revenue by resident or non-resident.

325
00:41:18,000 --> 00:41:25,000
You know, I think I guess this is probably a factor of the dispenser I worked at

326
00:41:25,000 --> 00:41:29,000
because they wanted to do the bare minimum to get by.

327
00:41:29,000 --> 00:41:38,000
But other dispensaries track your your ID and that's in a database somewhere that they keep, I'm sure.

328
00:41:38,000 --> 00:41:42,000
But the dispenser I worked at was more of the.

329
00:41:42,000 --> 00:41:45,000
All right, we're not in jail. OK, let's keep doing what we're doing.

330
00:41:45,000 --> 00:41:52,000
Yeah, I definitely get this sense. I get the sense after going through the Washington State data that

331
00:41:52,000 --> 00:41:59,000
the quality is really low and it's probably going to be that while for it's going to be like that for a while, I would think,

332
00:41:59,000 --> 00:42:05,000
until there's kind of more stringent enforcement of, you know, for the dispensaries.

333
00:42:05,000 --> 00:42:10,000
But I think Keegan, you mentioned this before, they're trying just to get these businesses.

334
00:42:10,000 --> 00:42:13,000
They're trying to get their tax money, right?

335
00:42:13,000 --> 00:42:17,000
And their tax money, the states are. Yeah, the states are, of course.

336
00:42:17,000 --> 00:42:23,000
So the more more of these companies are operating, the more tax revenue they're going to get.

337
00:42:23,000 --> 00:42:33,000
So I'm sure they don't want to push too hard to push some of these folks out of business, but to make it more time.

338
00:42:33,000 --> 00:42:37,000
Oregon used to track people's driver's license.

339
00:42:37,000 --> 00:42:46,000
And then Jeff Sessions became attorney general and they realized that like he could come in at any time and take all that information.

340
00:42:46,000 --> 00:42:52,000
So then they that the governor signed an order to purge all that information.

341
00:42:52,000 --> 00:43:01,000
They can't keep that information for more than a day or two, or maybe they don't even keep it anymore.

342
00:43:01,000 --> 00:43:09,000
So you're tracking it and then they stop tracking it. So, yeah, it's.

343
00:43:09,000 --> 00:43:18,000
You've hit on something. I don't know if it necessarily applies here, but there's a principle I heard is almost be careful what you measure,

344
00:43:18,000 --> 00:43:22,000
because that dictates a lot of your analysis.

345
00:43:22,000 --> 00:43:32,000
But in this case, it depends on who wants to be careful. But like in Colorado, like, for example.

346
00:43:32,000 --> 00:43:37,000
Exactly. So if you measured the amount of sales going to people in Wyoming,

347
00:43:37,000 --> 00:43:43,000
that may lead you to suspect that it's on this crossing.

348
00:43:43,000 --> 00:43:46,000
We had this conversation a couple of months ago.

349
00:43:46,000 --> 00:43:56,000
There was an article in the Oregonian about those towns along the Idaho border, and they had disproportionately high cannabis sales

350
00:43:56,000 --> 00:44:01,000
because all these people were coming over from Idaho and buying cannabis.

351
00:44:01,000 --> 00:44:08,000
And the parking lots of all the dispensaries were full of people with Idaho license plates.

352
00:44:08,000 --> 00:44:21,000
And yeah, there was this thing, New Mexico, when the legislators of New Mexico said, yeah, we want to we want to legalize it before Texas does so we can get all that revenue from Texas.

353
00:44:21,000 --> 00:44:28,000
So there's a great analysis there you could do like a heat map along the county, you know, along the borders.

354
00:44:28,000 --> 00:44:32,000
Yeah, you know where it's legal where it's not legal.

355
00:44:32,000 --> 00:44:38,000
So I tried to make that pitch to a town closer to the border of Wyoming.

356
00:44:38,000 --> 00:44:48,000
And I pitched it to them. I said, hey, about 20% of our customers are coming from Wyoming.

357
00:44:48,000 --> 00:44:55,000
You could capture that whole market because they were in between the town I was working at and Wyoming border.

358
00:44:55,000 --> 00:45:04,000
So they would have they would have saved all these people driving from Wyoming about 30 minutes, which is a huge convenience for them.

359
00:45:04,000 --> 00:45:19,000
And inevitably, the town shot it down because they were skeptical about weed in the first place and they didn't want it on their streets, which it was it already was there because they were driving down to my town.

360
00:45:19,000 --> 00:45:27,000
And then, I don't know, smoking it there. But, you know, it is what it is.

361
00:45:27,000 --> 00:45:31,000
You can't fight culture sometimes.

362
00:45:31,000 --> 00:45:38,000
I think you mentioned this earlier where you were in if you didn't then I can mention it now.

363
00:45:38,000 --> 00:45:51,000
At Cana Con, a regulator from Oklahoma spoke and she just generally said that they're not really there just to bust people for breaking the violations.

364
00:45:51,000 --> 00:45:56,000
They want to help businesses operate by the rules.

365
00:45:56,000 --> 00:46:11,000
And so I think that just sort of, you know, gets at the idea where there's all the if they just blanket ban it, then there's sort of shady things that happen.

366
00:46:11,000 --> 00:46:19,000
And so, you know, maybe stuff was coming in before Oklahoma legalized it from Colorado to Oklahoma.

367
00:46:19,000 --> 00:46:30,000
And so they're like, OK, you know, but just set some regulations and then let people operate by these rules and help them operate by their rules.

368
00:46:30,000 --> 00:46:41,000
And so that's just the trend I've sort of observed.

369
00:46:41,000 --> 00:46:55,000
It may kind of help people like, you know, not have to like take things across the borders because that was another thing that was brought up is that's actually, you know, that's pretty it is a federal fence.

370
00:46:55,000 --> 00:47:02,000
So it's definitely not something to take lightly.

371
00:47:02,000 --> 00:47:14,000
My one rebuttal to that is people were throwing up resistance to like, oh, these people are going over state lines and that's the federal offense, like you mentioned.

372
00:47:14,000 --> 00:47:24,000
But my rebuttal to that is like every year right before the Fourth of July, people go to Wyoming from Colorado to buy fireworks.

373
00:47:24,000 --> 00:47:37,000
Fireworks are illegal in Colorado, but they go up to Wyoming anyway and take them down to Colorado and shoot them off. And, you know, we have fireworks injuries left and right from kids just playing with them in their backyards.

374
00:47:37,000 --> 00:47:42,000
So that was my one rebuttal. But yeah, no, it is a good point because it is.

375
00:47:42,000 --> 00:47:48,000
I think it's a federal crime to bring drugs over state line versus fireworks over state line.

376
00:47:48,000 --> 00:47:57,000
Well, it was just something that was brought up once again at CannaCon. Somebody spoke about just the legality.

377
00:47:57,000 --> 00:48:08,000
So he basically said, OK, you know, if you're operating in the cannabis industry, you've sort of got a mark on your back.

378
00:48:08,000 --> 00:48:24,000
You know, you have to operate professionally and within the scope of the law. And the last thing you want to do is take a bunch of cannabis from one state where it's legal to another state where it's not.

379
00:48:24,000 --> 00:48:28,000
That's like a whole can of worms.

380
00:48:28,000 --> 00:48:45,000
But anywho, I don't know. It just makes me think that, like, you know, I think Oklahoma does touch Colorado. So the panhandle supported me things that they were just just tired of busting people all day long.

381
00:48:45,000 --> 00:48:51,000
And they decided, OK, we just need to get some regulations in place.

382
00:48:51,000 --> 00:48:58,000
I don't know why they actually did it. There's a town in the panhandle called Texahoma, I think.

383
00:48:58,000 --> 00:49:03,000
And they that's right where Colorado touches Oklahoma.

384
00:49:03,000 --> 00:49:12,000
And I think they just just like position some extra police force there just for that reason.

385
00:49:12,000 --> 00:49:29,000
But yeah, because I have a friend that started a business down Oklahoma, right, when it opened up because he was doing everything janky up in Colorado and decided to go legit when Oklahoma legalized.

386
00:49:29,000 --> 00:49:46,000
We kind of touched on this last week is when you when they do do that, you know, it brings this whole, you know, company, all this revenue out from the black market into like the open into the legitimate business now.

387
00:49:46,000 --> 00:49:56,000
So now they're you know, they're employing their employing attorneys are getting a bank. They're they're hiring marketers to employees.

388
00:49:56,000 --> 00:49:59,000
So on and so forth.

389
00:49:59,000 --> 00:50:02,000
So just.

390
00:50:02,000 --> 00:50:10,000
I just think it's bringing things into the sunlight so that way that people can actually operate legitimately by the rules.

391
00:50:10,000 --> 00:50:12,000
It's more transparent.

392
00:50:12,000 --> 00:50:15,000
In the long run, it creates a safer product.

393
00:50:15,000 --> 00:50:28,000
And it provides what we care about is better data. So that's a maybe we have a.

394
00:50:28,000 --> 00:50:37,000
Or in a sense, thank you.

395
00:50:37,000 --> 00:50:43,000
Well, it's been a good conversation today. So.

396
00:50:43,000 --> 00:50:47,000
I have prepared just.

397
00:50:47,000 --> 00:50:54,000
Just like I guess just with two minutes, Sean, maybe I can show you in Paul and Chris as well.

398
00:50:54,000 --> 00:51:08,000
Charles just I just did a quick forecast of Oklahoma sales. So this could be just.

399
00:51:08,000 --> 00:51:19,000
So this could show you how you could go about forecasting with your data because you're going to have a lot more data points than I do.

400
00:51:19,000 --> 00:51:22,000
So basically.

401
00:51:22,000 --> 00:51:29,000
Just real quick, and I'll publish this to to GitHub, but.

402
00:51:29,000 --> 00:51:38,000
You can actually get so Oklahoma is publishing just their monthly tax revenue.

403
00:51:38,000 --> 00:51:42,000
And so you can download that.

404
00:51:42,000 --> 00:51:45,000
And pretty good monthly numbers.

405
00:51:45,000 --> 00:51:51,000
Yes, so they're getting about five million dollars a month in tax revenue.

406
00:51:51,000 --> 00:52:10,000
And so you can actually back out the total sales rate. So if the tax is seven percent, then we can back out the total revenue.

407
00:52:10,000 --> 00:52:24,000
So you're all just good and just run this whole thing. So essentially, we'll just back out the revenue from the tax from the amount of tax.

408
00:52:24,000 --> 00:52:30,000
And then essentially forecast that forward.

409
00:52:30,000 --> 00:52:39,000
And I'll share the script with you. So we're essentially just using or bringing the models for forecasting.

410
00:52:39,000 --> 00:52:46,000
And I've got some links at the top for you to sort of read up on a reading of forecasting.

411
00:52:46,000 --> 00:52:53,000
But it's basically just using historic values to the forecast forward.

412
00:52:53,000 --> 00:53:08,000
So just to show you so far, they started publishing data in July of 2020 and going through.

413
00:53:08,000 --> 00:53:18,000
May of 2021.

414
00:53:18,000 --> 00:53:31,000
And so cannabis revenues on average about 78 to about 80 million in sales per month.

415
00:53:31,000 --> 00:53:43,000
And just to show you the forecasting model, we only have 11 observations.

416
00:53:43,000 --> 00:53:49,000
And we're basically just forecasting within a are wine.

417
00:53:49,000 --> 00:53:58,000
So that's just using the past month's observations essentially to forecast forward.

418
00:53:58,000 --> 00:54:10,000
So our forecasts aren't the best. And so if you plot this out, you know, I don't know how realistic those forecasts are.

419
00:54:10,000 --> 00:54:19,000
But, you know, it at least gives us, you know, a dollar amount.

420
00:54:19,000 --> 00:54:22,000
So just to wrap this up real quick.

421
00:54:22,000 --> 00:54:35,000
You know, essentially, I'm forecasting, you know, Oklahoma to have, you know, one billion in cannabis sales in 2021.

422
00:54:35,000 --> 00:54:42,000
And that would be a, you know, I think like a conservative estimate.

423
00:54:42,000 --> 00:54:52,000
So I would, you know, I would say more than that. But, you know, there could be errors, you know, maybe, maybe sales dip way back down in the fall.

424
00:54:52,000 --> 00:55:02,000
So we'll see. But that's sort of a principle I like to share of forecasting is an iterative process.

425
00:55:02,000 --> 00:55:08,000
So basically, we're going to save these forecasts.

426
00:55:08,000 --> 00:55:15,000
And in the coming months, you know, we can see what the actuals were.

427
00:55:15,000 --> 00:55:22,000
And at the end of the year, how far we were off.

428
00:55:22,000 --> 00:55:36,000
Yeah. One billion sounds reasonable. That's I don't know for the first year of sales, Colorado saw one billion, I think, on the fourth year.

429
00:55:36,000 --> 00:55:39,000
I don't know, correct me if I'm wrong there.

430
00:55:39,000 --> 00:55:42,000
But I think that's what I remember.

431
00:55:42,000 --> 00:55:52,000
And that's, that's why it's so interesting to look at these numbers is I think you're right. I think it took Colorado quite a while to get to one billion in sales.

432
00:55:52,000 --> 00:56:04,000
So there was still a lot of resistance and hesitancy initially, whereas Oklahoma is coming in at it with all these other states kind of paving the path already.

433
00:56:04,000 --> 00:56:14,000
I wonder if the kind of the late bloomers will actually move super fast because they'll have probably better access to investors will see less risk.

434
00:56:14,000 --> 00:56:19,000
Yeah. Yeah, because all the banks still don't want to touch it.

435
00:56:19,000 --> 00:56:27,000
They have the technology. So we've got to so, you know, in the production function in economics, you've got technology.

436
00:56:27,000 --> 00:56:41,000
And so what's happened is so like a can of corn is all the existing technology that's been produced in Colorado and Oregon, California, Michigan, or where have you.

437
00:56:41,000 --> 00:56:50,000
It's all the actually there is some people who are doing some amazing stuff with hydrocarbons in Illinois.

438
00:56:50,000 --> 00:57:07,000
They were there. So you have people bring their technologies to Oklahoma, and then they're able to just start up quickly, but it just starting up at a staggering staggering rate.

439
00:57:07,000 --> 00:57:23,000
That's that's something I noticed as well as people were getting were kind of cutting their teeth on the Colorado market. And then after so many years, you know, their friend back in their home state was saying, Hey, let's start a business.

440
00:57:23,000 --> 00:57:41,000
This state just legalize. Just take everything you know and bring it over here. And that's kind of what my friends did in Oklahoma they got really good at extraction, and then took all their experience and then use that to kind of create their

441
00:57:41,000 --> 00:57:44,000
business out of it.

442
00:57:44,000 --> 00:58:03,000
Exactly. So, just for example, there are like some like, like, trimming machines that were built in Colorado and maybe not Colorado I'm not sure where they were built but in one of those states, and then, yes, the processing.

443
00:58:03,000 --> 00:58:19,000
And this is where a lot of players are missing out because, you know, I'm sure there are like manufacturers out there who do build processing equipment and stuff like that, but they just didn't want to touch the cannabis space.

444
00:58:19,000 --> 00:58:33,000
They just left an opening for a startup to come in and basically say okay we're doing, you know, cannabis processing equipment.

445
00:58:33,000 --> 00:58:42,000
And now they've got this little successful business there so just interesting.

446
00:58:42,000 --> 00:59:00,000
It's been, I mean, I'm a little biased but I personally think it's been a fascinating and just see the industry to watch grow and kind of take its form, just because it does have such unique circumstances in which it has grown and formed, you know,

447
00:59:00,000 --> 00:59:14,000
dual state legalized federally illegal, kind of everything is kind of create a push and pull dynamic on how it's going to progress and move forward.

448
00:59:14,000 --> 00:59:31,000
So, for next week, I'm going to dust off my old Colorado analysis and see how I get that up to up to speed. And then, yeah, I'd love to hear more about your analysis.

449
00:59:31,000 --> 00:59:35,000
And then we can continue to bounce ideas off each other.

450
00:59:35,000 --> 00:59:45,000
And I think, I think Paul could be a good resource for you Sean because he is quite the retail analysis expert.

451
00:59:45,000 --> 00:59:50,000
I wouldn't say that I'm learning as I go.

452
00:59:50,000 --> 01:00:05,000
I was gonna say I'm more than happy to share the data that I have with you guys for you to play with on your own time. I have no attachment to it or no need to keep it secret.

453
01:00:05,000 --> 01:00:09,000
So, I'm more than happy to share that with you guys.

454
01:00:09,000 --> 01:00:21,000
It could always be interesting like, like I said, the data points I'm interested in are basically the day of the week effects. So, what days of the week have the biggest sales.

455
01:00:21,000 --> 01:00:27,000
And I'm kind of curious about morning versus afternoon.

456
01:00:27,000 --> 01:00:35,000
But I'm sure there's tons of good questions but those are just the two that come to mind.

457
01:00:35,000 --> 01:00:40,000
If you have any questions, you know, reach out to us.

458
01:00:40,000 --> 01:00:49,000
And then you know check out Keegan's GitHub, my GitHub, cannabis industry data analysis there.

459
01:00:49,000 --> 01:00:56,000
Yeah, so Sean I joined this group of probably what four weeks ago or so five weeks ago maybe.

460
01:00:56,000 --> 01:01:07,000
And I think Keegan have been super helpful with my master's program answering kinds of questions about data and stuff. So, yeah, thanks guys. I really appreciate that you give me.

461
01:01:07,000 --> 01:01:08,000
Sure.

462
01:01:08,000 --> 01:01:23,000
It's always fun. I have a lot of fun every week. So, we'll be in touch. Until then, have fun crunching numbers and we'll be here next week.

463
01:01:23,000 --> 01:01:36,000
Quick I don't know if you guys saw I put a link in the chat box chat window rather to GM's job posting board so if you see anything you're interested in just give me a shout out at my email account and we'll see what we can do.

464
01:01:36,000 --> 01:01:39,000
Yeah, I'm copying pasting that right now.

465
01:01:39,000 --> 01:01:40,000
All right.

466
01:01:40,000 --> 01:01:42,000
Great to meet you all.

467
01:01:42,000 --> 01:01:53,000
I'm having a fun engaging conversation. This is, I don't usually get to nerd out on numbers like this so this is kind of a nice fun group here.

468
01:01:53,000 --> 01:02:10,000
Definitely. That's, that's the that's the whole point you know, join a group of like minded individuals because it's fun so we'll, we'll crunch some numbers next week and like I said I'll dust off my own Colorado analysis because I think you'll

469
01:02:10,000 --> 01:02:16,000
enjoy that and I'll share it with you and then maybe you can pick up from what I have.

470
01:02:16,000 --> 01:02:19,000
Sweet. Yeah, I'm excited for that.

471
01:02:19,000 --> 01:02:24,000
Awesome. Well, until then, have a productive week.

472
01:02:24,000 --> 01:02:25,000
Take care guys.

473
01:02:25,000 --> 01:02:27,000
Good to see you. Good to meet you.

474
01:02:27,000 --> 01:02:51,000
Bye.