Meeting Title: Insomnia Customer Segmentation Analysis Check-in Date: 2025-11-06 Meeting participants: Robert Tseng, Amber Lin
WEBVTT
1 00:01:39.010 ⇒ 00:01:40.130 Amber Lin: Hi!
2 00:01:40.840 ⇒ 00:01:41.860 Robert Tseng: Hello!
3 00:01:42.050 ⇒ 00:01:46.730 Amber Lin: You have such a busy day. I saw your calendar, and there’s just no gaps!
4 00:01:47.560 ⇒ 00:01:50.150 Robert Tseng: Yeah, today is kind of rough.
5 00:01:50.660 ⇒ 00:01:55.939 Amber Lin: I have a short meeting, I want you to have some time to rest, especially if you’re gonna review my work later.
6 00:01:57.960 ⇒ 00:01:58.590 Amber Lin: Yeah.
7 00:01:59.410 ⇒ 00:02:05.730 Amber Lin: Last week, we talked about, like, the… you sent me the different…
8 00:02:06.240 ⇒ 00:02:07.580 Robert Tseng: Like, the…
9 00:02:07.580 ⇒ 00:02:15.870 Amber Lin: I think we were… we were talking about pricing, but then… and then I used it in the Insomnia stuff. I was like, oh, everything kind of…
10 00:02:16.040 ⇒ 00:02:33.530 Amber Lin: relates together, because I… when we started talking, like, 2 weeks ago, I had no clue of what’s… whatever was… we were talking about, and what was happening, and now Default’s looking at pricing, and then Insomnia’s looking at lifecycle and the different segments.
11 00:02:33.560 ⇒ 00:02:36.339 Amber Lin: And it really all comes down to, like.
12 00:02:37.150 ⇒ 00:02:46.709 Amber Lin: It’s customer segmentation, because default also wants to look at, oh, maybe can we do, like, vertical pricing, and that is a different type of segmentation.
13 00:02:46.900 ⇒ 00:02:47.570 Robert Tseng: Yeah.
14 00:02:47.570 ⇒ 00:02:48.390 Amber Lin: Yeah.
15 00:02:49.790 ⇒ 00:03:00.459 Amber Lin: Very, very cool, very interesting. I also watched the Eden Roadmap planning stuff, so I know there’s some analysis coming there, too. And that’s also about pricing. I was like, wow.
16 00:03:01.010 ⇒ 00:03:06.180 Amber Lin: people really need a lot of pricing. I guess that’s where all the money comes from.
17 00:03:06.180 ⇒ 00:03:06.990 Robert Tseng: Yeah.
18 00:03:07.570 ⇒ 00:03:10.619 Robert Tseng: We’re becoming a customer segmentation company.
19 00:03:10.620 ⇒ 00:03:13.750 Amber Lin: That’s true.
20 00:03:13.750 ⇒ 00:03:17.250 Robert Tseng: Maybe if we come up with a good solution, then that ends up…
21 00:03:17.370 ⇒ 00:03:31.030 Robert Tseng: something… or pivot into doing something like that. I mean, there could be a… there could be a playbook. Like, every one of our customers end up wanting to go that way because they were never able to do it before.
22 00:03:31.030 ⇒ 00:03:34.640 Amber Lin: Yeah. Because they don’t… they have nothing to base it off of.
23 00:03:34.870 ⇒ 00:03:36.220 Robert Tseng: Yeah, totally.
24 00:03:36.410 ⇒ 00:03:50.599 Amber Lin: Yeah, and I think that comes before any of the other analysises, because if you change price, you just get so much more money than, oh, this campaign did better, and this helped that, and it works together, too.
25 00:03:53.250 ⇒ 00:04:01.050 Robert Tseng: Yeah, I think there’s, I mean, the campaign segmentation is… I mean, that’s a no-brainer, because, like.
26 00:04:01.540 ⇒ 00:04:06.819 Robert Tseng: it helps… But conversion rates are just so low, so it’s like…
27 00:04:07.370 ⇒ 00:04:24.739 Robert Tseng: you still… doing segment… segmentation to acquire a new customer versus, like, yeah, pricing experiments that impact your existing customer base, yeah, I mean, it’s… you’re not gonna… the impact is always gonna be more if you do it for your existing customers, which is why…
28 00:04:24.740 ⇒ 00:04:25.300 Amber Lin: Yeah.
29 00:04:25.300 ⇒ 00:04:34.350 Robert Tseng: segmentation for Insomnia, if you’re able to, you know, do better on the own channel side, like, you’re not reacquiring these customers, like, they’re already
30 00:04:34.820 ⇒ 00:04:38.330 Robert Tseng: Yeah, you’re not really spending more dollar to acquire it. You’re trying to just.
31 00:04:38.330 ⇒ 00:04:38.690 Amber Lin: Yeah.
32 00:04:38.690 ⇒ 00:04:43.850 Robert Tseng: them with stuff that’s… are, you already have control over. So…
33 00:04:43.850 ⇒ 00:04:44.340 Amber Lin: Oh.
34 00:04:44.340 ⇒ 00:04:48.149 Robert Tseng: Yeah, like, I think, that’s… that’s kind of where…
35 00:04:49.240 ⇒ 00:05:03.540 Robert Tseng: we’re playing more. Like, I don’t really think we’re doing, like, paid ad segmentation. Like, I don’t want us to be building audiences, and, like, every, every, like, growth marketer is gonna end up coming into these situations and… and… and…
36 00:05:03.670 ⇒ 00:05:11.349 Robert Tseng: Pushing for their own… types of segmentation, so I don’t want to, like, step on their toes, yeah.
37 00:05:11.350 ⇒ 00:05:16.819 Amber Lin: So that’s… that’s, like, marketing… market segmentation, and we’re just working with
38 00:05:17.090 ⇒ 00:05:30.050 Amber Lin: existing data. And that makes sense, because all the data we touch is what they already have, so it doesn’t really make sense for us to get more data about customers who don’t exist yet, and then do segmentation based on that.
39 00:05:30.310 ⇒ 00:05:34.270 Robert Tseng: Yeah. I mean, maybe that… that will be a… some… like…
40 00:05:35.490 ⇒ 00:05:51.489 Robert Tseng: they’re… they’re always going to want to know, like, oh, like, what do we not know about the customers that we’re trying to acquire? And, you know, that’s why, like, bigger companies, they’ll, like, buy syndicated data from, like, Forester and whatever, to try to, like.
41 00:05:51.950 ⇒ 00:06:01.769 Robert Tseng: basically buy more attributes about their… that describe their… their… the new customers that they’re going after, but that’s…
42 00:06:02.480 ⇒ 00:06:07.570 Robert Tseng: I think that that’s pretty… I mean, I don’t think we have…
43 00:06:07.700 ⇒ 00:06:11.430 Robert Tseng: The… most of our clients are not really in a place to do that, like.
44 00:06:11.710 ⇒ 00:06:15.909 Robert Tseng: You’re not really spending hundreds of thousands of dollars buying, like.
45 00:06:16.180 ⇒ 00:06:22.779 Robert Tseng: Market research reports, unless you’re already doing, like, 8 or, like, kind of… you’re doing over $100 million in sales.
46 00:06:23.270 ⇒ 00:06:27.300 Amber Lin: Oh… Okay, that’s interesting.
47 00:06:33.340 ⇒ 00:06:41.170 Robert Tseng: I mean, I would like to be able to do that. We haven’t hired anybody that has that background, though. Like, I don’t even really have that.
48 00:06:41.450 ⇒ 00:06:41.940 Amber Lin: Hmm.
49 00:06:41.940 ⇒ 00:06:44.359 Robert Tseng: background, like, I’ve never… What exactly is…
50 00:06:44.360 ⇒ 00:06:45.539 Amber Lin: What’s the difference?
51 00:06:47.100 ⇒ 00:06:50.480 Amber Lin: I thought the skills would translate pretty well, but is.
52 00:06:50.480 ⇒ 00:06:53.860 Robert Tseng: Yeah, yeah, I mean, the skin, it’s… Well…
53 00:06:54.080 ⇒ 00:06:57.390 Robert Tseng: You have to be able to,
54 00:06:58.120 ⇒ 00:07:03.929 Robert Tseng: Well, syndicated data is just, like, a massive data set, and you have to, like, be able to know how to
55 00:07:04.310 ⇒ 00:07:14.409 Robert Tseng: Do the research to find, like, this… the report that you want to buy. Then you have to… yeah, and then know how to bring it into,
56 00:07:15.620 ⇒ 00:07:18.339 Robert Tseng: what you already know about the customer. So…
57 00:07:18.340 ⇒ 00:07:18.710 Amber Lin: Mmm.
58 00:07:18.710 ⇒ 00:07:22.840 Robert Tseng: I mean, I don’t think it’s that hard, but…
59 00:07:23.120 ⇒ 00:07:27.389 Robert Tseng: Yeah, it’s not really something we’ve… we’ve had to do yet.
60 00:07:27.820 ⇒ 00:07:28.700 Amber Lin: I see.
61 00:07:29.180 ⇒ 00:07:32.219 Amber Lin: We’re getting there, our customers are becoming bigger.
62 00:07:32.220 ⇒ 00:07:32.660 Robert Tseng: Yeah.
63 00:07:32.660 ⇒ 00:07:33.790 Amber Lin: That’s pretty vague.
64 00:07:35.280 ⇒ 00:07:48.269 Amber Lin: I’m very excited that I am able to do this analysis, and it’s very, very interesting. I find it much more interesting than project management, I have to be honest.
65 00:07:48.600 ⇒ 00:07:50.579 Robert Tseng: Oh, great, yeah.
66 00:07:51.800 ⇒ 00:07:57.769 Robert Tseng: I mean, I would prefer your time to be here, I feel like it’s… You know, it’s… it’s…
67 00:07:57.900 ⇒ 00:08:05.080 Robert Tseng: It’s… it’s higher value, like, billable-wise for the… for the So, sorry.
68 00:08:05.080 ⇒ 00:08:05.570 Amber Lin: Yeah.
69 00:08:05.570 ⇒ 00:08:06.490 Robert Tseng: I’m fine with that.
70 00:08:06.490 ⇒ 00:08:20.330 Amber Lin: And I also… and especially, like, I get stuck, but then every week I get to ask you, and I was like, oh, this is something I’ve never thought about before, and it was like, oh, then this enables me to do a lot of things, because when I started on a default.
71 00:08:20.480 ⇒ 00:08:34.250 Amber Lin: analysis. It was mostly Utem asked me to do revenue and, like, usage concentration, and I did that, and it meant nothing, because I was so confused, like, why the correlation just meant what they were currently doing, and I had no.
72 00:08:34.250 ⇒ 00:08:34.610 Robert Tseng: Yeah.
73 00:08:34.610 ⇒ 00:08:52.050 Amber Lin: until I figured out, oh, this is about pricing. And then I was like, oh, then this number actually means something. And it’s very fun, because a lot of it, the technical stuff can be done with AI. It’s really about how to interpret and to…
74 00:08:52.400 ⇒ 00:08:55.070 Amber Lin: design the approach, so…
75 00:08:55.070 ⇒ 00:08:55.640 Robert Tseng: Yep.
76 00:08:55.640 ⇒ 00:09:01.410 Amber Lin: I didn’t even need to do too much SQL. I… I typed prompts… prompts into cursor.
77 00:09:01.670 ⇒ 00:09:05.919 Robert Tseng: Yeah, yeah, I think that’s… that’s kind of all you need to really do now.
78 00:09:05.920 ⇒ 00:09:09.459 Amber Lin: Yeah, which is awesome. This is so different than when you started.
79 00:09:09.660 ⇒ 00:09:10.060 Robert Tseng: Yeah.
80 00:09:10.060 ⇒ 00:09:11.280 Amber Lin: I bet.
81 00:09:14.160 ⇒ 00:09:25.080 Amber Lin: Okay, I don’t have new questions for this week, it mostly was just doing the insomnia analysis, and I…
82 00:09:25.350 ⇒ 00:09:32.600 Amber Lin: did the first part, and I found that they have 34% champions, which is what you guessed. Yeah.
83 00:09:32.810 ⇒ 00:09:40.200 Amber Lin: And then I was… let me scroll down…
84 00:09:40.480 ⇒ 00:09:44.429 Amber Lin: And then I was trying to compare how different…
85 00:09:44.930 ⇒ 00:09:52.850 Amber Lin: these segments really are, because the scale is a relative scale, and I wanted to understand, like, does it mean that they just have
86 00:09:53.220 ⇒ 00:09:55.950 Amber Lin: one order different? Does it.
87 00:09:55.950 ⇒ 00:09:56.270 Robert Tseng: Yeah.
88 00:09:56.270 ⇒ 00:09:56.670 Amber Lin: they have.
89 00:09:56.670 ⇒ 00:09:58.520 Robert Tseng: What’s the standard deviation between…
90 00:09:58.520 ⇒ 00:10:01.959 Amber Lin: Yeah, yeah, so I wanted to do that, and I was trying to…
91 00:10:02.130 ⇒ 00:10:20.019 Amber Lin: somewhat reproduce what they had, what they did for the segmentation, mostly trying to find the fields. But I need to… I need to do more of that. I did it really late last night, and I don’t think I interpreted what was happening. So I’ll do that.
92 00:10:20.550 ⇒ 00:10:25.010 Amber Lin: And then… let’s see…
93 00:10:25.320 ⇒ 00:10:34.749 Amber Lin: And then I have the data on their engagement behavior, so I can analyze the ad based on their segments.
94 00:10:34.860 ⇒ 00:10:37.290 Amber Lin: Demolade helped me with that, and…
95 00:10:37.450 ⇒ 00:10:45.769 Amber Lin: Lastly, do you know how they calculate LTV currently? They don’t have churn rates, so I… I don’t know how to…
96 00:10:45.920 ⇒ 00:10:50.079 Amber Lin: compare, like, their segmentation, how it relates to their current LTV.
97 00:10:51.040 ⇒ 00:10:57.779 Robert Tseng: Ltv is just, like, the number of… like, it’s just a sum of all the orders that they’ve purchased to date.
98 00:10:59.090 ⇒ 00:11:02.310 Amber Lin: Oh. Oh, I s- I see, I see.
99 00:11:02.310 ⇒ 00:11:02.930 Robert Tseng: Yeah.
100 00:11:03.400 ⇒ 00:11:05.410 Amber Lin: Oh, then I have that.
101 00:11:05.630 ⇒ 00:11:06.440 Robert Tseng: Yeah.
102 00:11:11.780 ⇒ 00:11:19.260 Amber Lin: That’s so much easier. I somehow saw… I somehow thought it was about, like, predicting churn rates or whatever, but I guess that’s projected LTV.
103 00:11:19.260 ⇒ 00:11:20.899 Robert Tseng: Yeah, that’s for projected LPV.
104 00:11:21.800 ⇒ 00:11:22.370 Amber Lin: Okay.
105 00:11:22.370 ⇒ 00:11:29.660 Robert Tseng: Which is the next step. Once you have a good understanding of LTV, you’re able to break them out, then… then you want to be able to, like, layer…
106 00:11:29.660 ⇒ 00:11:43.880 Robert Tseng: So, if someone’s in a lower LTV, category, but their projected LTV is in, like, the champions category, then, like, that tells you something that’s, like, there’s more potential to be realized for that customer base.
107 00:11:44.410 ⇒ 00:11:45.609 Robert Tseng: You know, yeah, so…
108 00:11:45.610 ⇒ 00:11:59.329 Amber Lin: I hear. Okay, okay, that’s cool. And when I… when I planned the analysis, I also want to see how the segments have changed over time. We currently don’t have time-stamped,
109 00:11:59.560 ⇒ 00:12:10.190 Amber Lin: RF, like, your segmentation tags, it’s not timestamped, but I think I can have a workaround from that, because when we downloaded the data, it was October…
110 00:12:10.470 ⇒ 00:12:19.379 Amber Lin: 15, and it’s been, like, more than half a month now, so I might just ask Sam or whoever to download the data again. I’ll just see what changed.
111 00:12:19.790 ⇒ 00:12:20.140 Robert Tseng: Yeah.
112 00:12:20.140 ⇒ 00:12:20.639 Amber Lin: That might be…
113 00:12:20.640 ⇒ 00:12:25.789 Robert Tseng: And that could be a good proxy for you on, like, how often these things change, yeah.
114 00:12:25.790 ⇒ 00:12:26.990 Amber Lin: Yeah, that would be cool.
115 00:12:27.190 ⇒ 00:12:41.630 Amber Lin: Okay, that… I think that’s a pretty… like, I cover most stuff and diagnose… what I can do with the current data. Is there anything else I should think of, of, like, what the current…
116 00:12:42.010 ⇒ 00:12:45.240 Amber Lin: Segmentation is not doing.
117 00:12:46.100 ⇒ 00:12:48.499 Amber Lin: Like, this is the analysis I plan to do.
118 00:12:48.850 ⇒ 00:12:50.420 Robert Tseng: Have you talked to Birdie yet?
119 00:12:50.530 ⇒ 00:12:52.040 Amber Lin: No.
120 00:12:52.200 ⇒ 00:12:55.450 Amber Lin: Because I am not in the channel, I’m waiting to be introduced.
121 00:12:55.840 ⇒ 00:12:59.290 Robert Tseng: I thought I asked a few times to do that, but .
122 00:12:59.290 ⇒ 00:13:00.720 Amber Lin: Utam sent the message.
123 00:13:00.940 ⇒ 00:13:01.899 Robert Tseng: Oh, I see.
124 00:13:01.900 ⇒ 00:13:03.120 Amber Lin: She has responded yet.
125 00:13:03.300 ⇒ 00:13:04.390 Robert Tseng: I see, okay.
126 00:13:05.420 ⇒ 00:13:08.840 Robert Tseng: Yeah, so I mean, like, knowing…
127 00:13:09.180 ⇒ 00:13:13.790 Robert Tseng: How often, like, how often are customers moving between stages? Important, for sure.
128 00:13:14.430 ⇒ 00:13:18.550 Robert Tseng: Yeah, and then,
129 00:13:21.950 ⇒ 00:13:24.500 Robert Tseng: I… think…
130 00:13:27.330 ⇒ 00:13:36.449 Robert Tseng: I mean, the limitations are pretty straightforward, like, if this is… these segments are purely… you already know how they’re calculated, then it’s like, okay, well, what other… what additional…
131 00:13:37.290 ⇒ 00:13:45.030 Robert Tseng: things would you want to bring in? I think we talked about loyalty last time, and it’s something we have access to, so… Yeah, absolutely.
132 00:13:45.030 ⇒ 00:13:50.270 Amber Lin: can see the loyalty data, in the dataset that I’m exploring. There’s… Okay.
133 00:13:50.390 ⇒ 00:13:53.280 Amber Lin: There’s some level of punch data there already.
134 00:13:53.450 ⇒ 00:13:58.189 Robert Tseng: Yeah, because within the loyalty program, we know what offers we’re sending to them.
135 00:13:59.730 ⇒ 00:14:10.540 Robert Tseng: You know, I think with the loyalty program, yes, everything is, like, discount or price-related, but I… I mean, I don’t know what types of… yeah, I don’t know what the loyalty program contains, but…
136 00:14:10.730 ⇒ 00:14:13.430 Robert Tseng: I, you know…
137 00:14:14.380 ⇒ 00:14:21.189 Robert Tseng: I mean, I’m just gonna speak from loyalty programs that I… the loyalty program I built out for… for Ruggable,
138 00:14:22.010 ⇒ 00:14:35.520 Robert Tseng: Yeah, I mean, there were… there was… we called… we called it, like, a Rugged Insiders program, and so they would pretty much get access to certain paid surveys that we would use, and they would… they would end up kind of being,
139 00:14:37.180 ⇒ 00:14:41.849 Robert Tseng: part of our design process when we were launching new collections, and so…
140 00:14:42.290 ⇒ 00:14:50.770 Robert Tseng: people that were really engaged with the surveys, like, we, like, we had a way of categorizing them as well. We… and,
141 00:14:51.330 ⇒ 00:14:55.110 Robert Tseng: Yeah, I… actually, they were really good predictors at, like.
142 00:14:55.840 ⇒ 00:15:00.250 Robert Tseng: When we were launching… when we were designing collections to figure out,
143 00:15:00.360 ⇒ 00:15:09.399 Robert Tseng: which designs would perform the best. The Rug Insiders had a better… but they predicted, what was gonna sell better than, like, some of the market research data we bought.
144 00:15:09.430 ⇒ 00:15:20.740 Robert Tseng: So, I mean, to me, that’s like a way of capturing, kind of, like, intended behavior data, because it’s like, if they’re engaging a lot with the surveys, we also get all the answers to their surveys.
145 00:15:20.930 ⇒ 00:15:36.709 Robert Tseng: Yes, we’re having to pay them out, but, like, they’re giving us really good feedback so that when we were building new collections, we would run our first designs by the Rug Insiders and have them rank them, and that was part of the.
146 00:15:36.710 ⇒ 00:15:37.410 Amber Lin: Oh…
147 00:15:37.630 ⇒ 00:15:47.210 Robert Tseng: like, the calculation that we made to recommend to the design team which ones to actually pursue. So, that became a cool feedback loop.
148 00:15:48.640 ⇒ 00:15:59.459 Robert Tseng: And, yeah, I mean, that was off of, like, loyalty segmentation. So, I… I don’t really think their loyalty program is nearly as robust as what we… I built out before.
149 00:15:59.790 ⇒ 00:16:00.730 Robert Tseng: Probably just, like.
150 00:16:00.850 ⇒ 00:16:01.800 Amber Lin: Redeem…
151 00:16:01.800 ⇒ 00:16:02.220 Robert Tseng: after…
152 00:16:02.220 ⇒ 00:16:04.140 Amber Lin: this many points.
153 00:16:04.140 ⇒ 00:16:12.100 Robert Tseng: Yeah, like, yeah, or like, here’s your birthday, here, like, when you buy, make enough purchases, you get, you know, X discount, or whatever.
154 00:16:12.650 ⇒ 00:16:23.929 Robert Tseng: But, yeah, I still think it’s… it’s interesting to look at, like, who’s responsive to those things. Are people actually working up to… to these different, awards? Like…
155 00:16:24.420 ⇒ 00:16:31.280 Robert Tseng: Yeah, maybe there’s something about the… Top segment of…
156 00:16:31.780 ⇒ 00:16:39.220 Robert Tseng: people who have qualified for certain gifts, the people who have, you know, earned the most in the loyalty program, like, if they’re.
157 00:16:39.640 ⇒ 00:16:54.870 Robert Tseng: can we match them to a particular… like, do they look the same as the champions? Or, like, you know, basically, what, like, what are the nuances for, like, being able to segment off of the loyalty program versus just looking at purely transactional data?
158 00:16:55.410 ⇒ 00:17:12.929 Amber Lin: Yeah, I was thinking about that, too. I don’t know how much punch data we have. I don’t think we have specific offers of what this person has received before. It’s mostly aggregate numbers of, oh, this many points, this many, sessions, and we have… I think we have that.
159 00:17:13.000 ⇒ 00:17:19.220 Amber Lin: But I… I will take a closer look specifically at loyalty. I think it gives a lot of interesting things.
160 00:17:20.540 ⇒ 00:17:25.909 Robert Tseng: Okay, yeah, I mean, if there’s no way to tie the points to a customer.
161 00:17:26.530 ⇒ 00:17:31.090 Robert Tseng: then you probably will hit a dead end, but I’m assuming you should be able to.
162 00:17:31.390 ⇒ 00:17:31.980 Amber Lin: Yeah.
163 00:17:32.320 ⇒ 00:17:40.520 Amber Lin: And then I did a bit of planning for how to introduce a lifecycle-based segmentation.
164 00:17:40.850 ⇒ 00:17:47.359 Amber Lin: not that different, it’s still… Quite a bit of it is based on purchase.
165 00:17:47.360 ⇒ 00:17:49.950 Robert Tseng: Yeah, your first cookie, your second cookie, I mean, yeah.
166 00:17:49.950 ⇒ 00:17:50.300 Amber Lin: Yeah.
167 00:17:50.300 ⇒ 00:17:51.190 Robert Tseng: that looks like. So.
168 00:17:51.190 ⇒ 00:17:59.160 Amber Lin: stuff like that. And then, I also want to propose at least find the… Like, the rates between the…
169 00:17:59.670 ⇒ 00:18:06.870 Amber Lin: different… different tiers. I think there’s a pretty standard, but we don’t know that right now, and it’s very hard to…
170 00:18:07.010 ⇒ 00:18:25.809 Amber Lin: like, compare how things… how campaigns worked, apart from just revenue. Like, we don’t know how to make people more loyal, how to prevent them from becoming, like, from churning, or from being inactive, like, because we don’t have any benchmarks to even see what impacts of our actions are.
171 00:18:28.160 ⇒ 00:18:32.839 Amber Lin: Do you think we should… sorry, like, like, I just think this is pretty…
172 00:18:33.660 ⇒ 00:18:37.880 Amber Lin: standard practice? But I don’t know, I’ve never done this before.
173 00:18:39.450 ⇒ 00:18:41.920 Robert Tseng: What do you mean by we can’t see the impact of it?
174 00:18:42.130 ⇒ 00:19:02.059 Amber Lin: Oh, I meant that right now, because we just don’t have life cycle, stages, so if we can’t tell if we have more loyal customers, or we are starting… we’re starting to see more people churn, we also don’t really see where people are
175 00:19:02.310 ⇒ 00:19:08.019 Amber Lin: when they start to turn, or what causes them to turn. So, say, if we see…
176 00:19:08.080 ⇒ 00:19:22.709 Amber Lin: less people converting from their second to third cookie, then maybe we can say, oh, this means that, and then we can have experiments to try and improve that rate. But we don’t have that rate.
177 00:19:22.880 ⇒ 00:19:24.370 Amber Lin: Right now, even.
178 00:19:24.850 ⇒ 00:19:29.660 Robert Tseng: Yeah, I mean, I would say churn is not that important,
179 00:19:29.870 ⇒ 00:19:35.660 Robert Tseng: Yeah, this is not, like, a subscription business where they have to… they’re paying a fee every month, like, it…
180 00:19:35.660 ⇒ 00:19:36.150 Amber Lin: That’s true.
181 00:19:36.150 ⇒ 00:19:40.790 Robert Tseng: I think we… I mean, time since last purchase is good enough.
182 00:19:41.140 ⇒ 00:19:48.290 Robert Tseng: So, I mean, I just sent you a picture. I was, like, standing in an insomnia, just, like, staring at their menu for…
183 00:19:48.960 ⇒ 00:20:04.460 Robert Tseng: For some time, and trying to, like, triangulate, well, what would a life cycle look like, right? If I was a first time at Insomnia, I don’t know if the cookies are good, maybe I’m gonna just go and buy one cookie, right? So that’s why I think first cookie or first purchase
184 00:20:04.570 ⇒ 00:20:11.000 Robert Tseng: Not… yeah, first purchase, first cookie, first classic or deluxe cookie, I think is important.
185 00:20:11.060 ⇒ 00:20:22.420 Robert Tseng: And then let’s say I like the cookie and I go back again. I’m like, I want to get, like, a box. You know, they’re… they’re really promoting their 12-pack, like, that’s, like, their… that’s, like, an important skew to them, so…
186 00:20:22.420 ⇒ 00:20:34.510 Robert Tseng: But then it becomes, like, first 12-pack, first 6-pack, or whatever it is. First deluxe, like, 6-pack. And it doesn’t always have to all be, like, my product.
187 00:20:34.510 ⇒ 00:20:40.729 Robert Tseng: type, but I think the product type and the sequence in the product types, like, if I… if I…
188 00:20:40.730 ⇒ 00:20:44.069 Amber Lin: I looked at, like, if I looked at, like, our champions.
189 00:20:44.070 ⇒ 00:20:57.200 Robert Tseng: segment, and I know that everyone in the champion segment has made 10 purchases, I would look at, like, well, what are they purchasing first? Are they actually following this progression? Or are they just, like, obsessed with cookies, and they just go straight for the
190 00:20:57.700 ⇒ 00:21:14.810 Robert Tseng: for the 12-pack, they’re not even working their way up to it. I don’t really know. Is there anybody that’s actually going from one pack, or one single cookies, to… to, like, 12-pack cookies? Like, maybe, like, that ends up being something that needs to happen. So, I guess, like.
191 00:21:14.810 ⇒ 00:21:15.570 Amber Lin: Whoa.
192 00:21:15.570 ⇒ 00:21:18.340 Robert Tseng: Another analogy I would make from Ruggable is, like.
193 00:21:18.340 ⇒ 00:21:19.560 Amber Lin: Okay, like…
194 00:21:19.560 ⇒ 00:21:24.119 Robert Tseng: Rugs are really difficult to purchase because, like, sizing mattered a lot, so…
195 00:21:24.130 ⇒ 00:21:42.060 Robert Tseng: That’s why, you know, being a D2C e-com company, it was hard to, like… you have to, like, measure your space out before you buy the bug. We know there’s a lot of friction to purchase it. Most people don’t know the size rug that they want. So there was, like, kind of a, oh, well, do we just position ourselves as, like.
196 00:21:42.120 ⇒ 00:21:59.479 Robert Tseng: you decide on whatever size you want, and we will custom make it for you, or do we give you, like, set sizes? And, like, the market research showed that actually, yeah, people will want to want, like, one standard rug. And so we just created, like, 3 to 5, like, standard rug sizes.
197 00:21:59.600 ⇒ 00:22:12.240 Robert Tseng: people would make their first rug purchase, and if they really liked their product, then they would come back and they would want… and then we could start to sell them bundles. So, we wanted everybody to be… and, like, our average, like, customer,
198 00:22:12.460 ⇒ 00:22:20.100 Robert Tseng: I think has… well, this is not just ruggable customer, but average rug buyer has 5 rugs in their home.
199 00:22:20.340 ⇒ 00:22:33.010 Robert Tseng: So, we know that if we get them to buy their first rug, then it’s like, okay, well, how do we try to go and enter the other… get them to buy rugs in the other spaces? So, that was, like, kind of always a North Star for
200 00:22:33.010 ⇒ 00:22:41.590 Robert Tseng: the product team, when they’re thinking about the customer, for first-time purchasers, like, how do we get them to 5 rugs? And,
201 00:22:41.730 ⇒ 00:23:01.020 Robert Tseng: yeah, there were different pathways to go. Like, if they bought a living room rug, then we know, okay, then we gotta really target them with bath mats. And so, bath mats usually come in pairs, there’s one next to the bathtub, there’s one next to the… to the toilet. And yeah, like, the data kind of supported that. We saw that for users, for customers that did
202 00:23:01.020 ⇒ 00:23:03.119 Robert Tseng: Buy a rug in a different room.
203 00:23:03.200 ⇒ 00:23:17.910 Robert Tseng: When they did end up purchasing bathroom… bath mats, they would buy two at a time. And so, yeah, we just, like, that’s kind of how you… how we pushed the marketing team to be like, this is how you’re going to be able to cross-sell bath mats into
204 00:23:17.910 ⇒ 00:23:26.739 Robert Tseng: first-time customers that purchase, like, whatever, living room… living room rugs. So, like, there’s that kind of, like, narrative that, like.
205 00:23:26.820 ⇒ 00:23:32.230 Robert Tseng: I mean, the insomnia experience is just so disconnected, like, I don’t really know, like.
206 00:23:32.380 ⇒ 00:23:38.199 Robert Tseng: what… how do you really, like, influence customers to go from one purchase to another? Right.
207 00:23:38.650 ⇒ 00:23:47.649 Robert Tseng: Obviously, you don’t want them to stay as, like, single cookie purchaser. The $3 cookie is, like, not gonna do anything, so…
208 00:23:47.650 ⇒ 00:23:48.140 Amber Lin: Yeah.
209 00:23:48.140 ⇒ 00:23:55.169 Robert Tseng: I mean, I don’t know what that mix looks like, but that’s where I feel like you don’t really need that much complicated… if you have
210 00:23:55.170 ⇒ 00:24:11.089 Robert Tseng: customers, the orders that they purchase, and you have these, like, segments, you can do a lot of that analysis just, like, off of that. Or, like, orders, product type, like, when they purchased it, or whatever. Like, yeah, you’d be able to come up with a lot of this kind of, narrative.
211 00:24:11.750 ⇒ 00:24:18.559 Amber Lin: I’ve never thought of that before. See, this is why I want to talk to you each week. I don’t even know that that was a…
212 00:24:18.590 ⇒ 00:24:34.590 Amber Lin: concept of thinking. So is this… it sounds like it’s a combination of market research of, well, people who buy cookies, if I can do specifically cookies at night, like, what does it… what do they usually
213 00:24:34.590 ⇒ 00:24:52.540 Amber Lin: want, or what are they usually like, so we can work towards that. And then also the overall journey of the customer. Like, when I think of customer journey, I always just think of what I’ve seen you do on Mixpanel before, so that’s, it’s like, oh, that’s a… but that’s a digital product.
214 00:24:52.540 ⇒ 00:24:53.580 Robert Tseng: Yeah.
215 00:24:54.400 ⇒ 00:25:00.070 Amber Lin: like, people… yes, there’s, like, an app for Insomnia, but they’re still…
216 00:25:00.240 ⇒ 00:25:16.940 Amber Lin: getting a cookie, and it’s very related to a local store, so I think I need to do a bit more research and understand that a bit better, because I didn’t have that part of thinking when I was structuring or doing my analysis, and that’s… that was really, really helpful and insightful.
217 00:25:17.460 ⇒ 00:25:22.120 Robert Tseng: Okay, yeah, great. I mean, I think… fortunately, like, we don’t need to…
218 00:25:22.440 ⇒ 00:25:39.210 Robert Tseng: I mean, they have… they should have so much historical data, like, yeah, you can do a lot of customer research just off in, like, their internal data, so… and they also have a lot of… a lot of products. I… I don’t really know if they all kind of fit together, but, yeah, I think, like, this is… this is the interesting…
219 00:25:39.210 ⇒ 00:25:45.859 Robert Tseng: kind of… to me, this is true, like, customer insights. Like, when you look at job descriptions that are, like.
220 00:25:45.860 ⇒ 00:25:54.280 Robert Tseng: Customer Insights Manager, or, like, Lead Customer Insights, or whatever, like, this… that’s what they are doing. Like, they’re not… MixedPanel is, like, such a…
221 00:25:54.420 ⇒ 00:26:01.760 Robert Tseng: low-level kind of customer insight, in my opinion. Like, it’s, I mean, yeah, digital product companies use it, but, like…
222 00:26:01.760 ⇒ 00:26:05.529 Amber Lin: Can you classify, like, low level and high level? What’s the difference?
223 00:26:06.050 ⇒ 00:26:10.739 Robert Tseng: Well, like, for MixedPanel, like, there’s not that much you could do. It’s just, like,
224 00:26:10.940 ⇒ 00:26:14.730 Robert Tseng: Yes, it’s all… it’s dependent on all the events that you’re tracking, and then.
225 00:26:14.730 ⇒ 00:26:15.130 Amber Lin: And you’re just…
226 00:26:15.130 ⇒ 00:26:23.620 Robert Tseng: It’s doing, like, various types of, like, funnel reporting, where you’re trying to, like, understand you know.
227 00:26:23.800 ⇒ 00:26:36.530 Robert Tseng: usage of this feature, like, and then, like, probably some sort of, like, revenue event, like, if they’re by… if it’s a PLG motion, and they’re going from a free to paid user, it’s, like.
228 00:26:36.530 ⇒ 00:26:44.130 Robert Tseng: Workflow 1 to paid conversion, workflow two to paid conversion. I was trying to understand, which is, like, what I’m doing on README, is like, hey.
229 00:26:44.370 ⇒ 00:27:08.099 Robert Tseng: your core features, these two core features, one is clearly a quick… a bigger driver of revenue than the other, and actually, when you’re… when they… when users are using your AI features, they boost both of the conversion rates for both of these types of projects. And so, it’s not that the AI features themselves are what people are paying for, but, like, people are using the AI features to, like, I guess.
230 00:27:08.100 ⇒ 00:27:14.249 Robert Tseng: get more value out of your core features, and that does have a revenue impact. Like, that’s basically a summary
231 00:27:14.250 ⇒ 00:27:18.029 Robert Tseng: Of the analysis that I shared with them last week. So…
232 00:27:18.030 ⇒ 00:27:26.279 Robert Tseng: like, you’re kind of doing that type of matching. I mean, yeah, but then for, you know, a physical product company with a lot of SKUs.
233 00:27:27.540 ⇒ 00:27:43.399 Robert Tseng: it’s not as, like, the workflow… there’s no workflow, it’s just, like, they go to the store and they buy something, I guess, at least for food and beverage. It’s, like, really hard to anticipate. For rug… for rugs, it’s a bit more complicated, because, like I said, there’s, like, there are different…
234 00:27:43.680 ⇒ 00:28:00.789 Robert Tseng: things that customers have to think about before they make their purchase. Like, they have to size out their situation, and whatever, and so that helps with lifecycle messaging. When someone gets on the email list, we’re telling them, this is all you need to know about, like, Rugs 101. Different types of rugs, this is how you size it.
235 00:28:00.790 ⇒ 00:28:01.300 Amber Lin: This is…
236 00:28:01.300 ⇒ 00:28:11.719 Robert Tseng: want different shapes look good in different rooms, so you’re, like, kind of slowly nurturing, kind of, the customer to, like, get to a point where they’re like, okay, I’m confident and ready to buy a rug.
237 00:28:11.830 ⇒ 00:28:12.400 Robert Tseng: But, like.
238 00:28:12.400 ⇒ 00:28:15.150 Amber Lin: That’s the same for Eden, too, right? Because they…
239 00:28:15.150 ⇒ 00:28:29.359 Robert Tseng: It’s just like, yeah, here’s the… here’s the results, like, people are losing all this weight, takes this much time, the side effects are not that bad, whatever. You’re just, like, trying to, you know, tell… get a customer to a point where they’re like, alright, I’m ready to buy.
240 00:28:29.480 ⇒ 00:28:32.170 Robert Tseng: But for cookies, it’s an impulsive purchase, like, it is.
241 00:28:32.170 ⇒ 00:28:32.730 Amber Lin: That’s very much…
242 00:28:32.730 ⇒ 00:28:39.230 Robert Tseng: like, oh, hello, you want a chocolate chip cookie? Like, it’s 10 PM, insomnia’s the place to go. Like, that’s, like.
243 00:28:39.230 ⇒ 00:28:42.559 Amber Lin: They’re solely impulse-based.
244 00:28:42.560 ⇒ 00:28:44.090 Robert Tseng: Yeah, it’s like a solely impulse-based.
245 00:28:45.300 ⇒ 00:29:03.909 Robert Tseng: I don’t know if people are going to be planning a cookie purchase for a week, you know? So, like, I think the dynamics look different. I don’t really know, like, what we could do to expedite that. I mean, I… or to catch them at the right time. I think that’s an interesting kind of, like, question to figure out.
246 00:29:04.340 ⇒ 00:29:10.029 Robert Tseng: But, yeah, I mean, I live next to a bunch of cookie companies, or, like, cookie shops, and
247 00:29:10.050 ⇒ 00:29:27.089 Robert Tseng: I’m, like, walking into these cookie shops and just standing there, people watching and trying to understand, like, what are people doing to buy the cookies. But, yeah, I mean, you’re gonna… like, that’s… that’s how I’ve been, like, trying to think about how to help in South, you know?
248 00:29:27.090 ⇒ 00:29:37.929 Amber Lin: Yeah, I’ve been asking the people around me who have bought, like, this cookie before, and it, like, right now, everybody has a different reason, because it is so impulse-based.
249 00:29:37.930 ⇒ 00:29:51.250 Amber Lin: Yeah. That was really helpful. I’ll go look more into that. I bet that will give me a bit more inspiration on what analysis to do, and I’ll send… I’ll send Looms to you later today. I have time now, because all my meetings are over.
250 00:29:51.550 ⇒ 00:29:52.270 Robert Tseng: Okay.
251 00:29:52.740 ⇒ 00:29:53.100 Amber Lin: Yeah.
252 00:29:53.100 ⇒ 00:29:57.579 Robert Tseng: I mean, yeah, I think you’re getting… yeah, you’re really just, like, trying to tell…
253 00:29:58.140 ⇒ 00:30:16.199 Robert Tseng: like, the story… you’re trying to explain what are customers, like, doing? How are they, like, really interacting with this experience that we’ve given them? Like, are they actually doing the… when you’re planning out your lifecycle stages, I mean, we have to start somewhere, so you need to come up with a framework, and I think it makes sense to go bottom up. It’s like…
254 00:30:16.200 ⇒ 00:30:28.060 Robert Tseng: Okay, they go for the small cookie, then they go up to the bigger box, or whatever, and they upgrade to a premium product, and, like, that’s maybe intuitively how you set it up, but then maybe you’ll observe that, like, nobody actually does that, and.
255 00:30:28.520 ⇒ 00:30:39.539 Robert Tseng: you know, the classic cookie purchasers, they actually… they’ll just buy… they’ll just… they always stay there. Those are just, like, single cookie buyers, they don’t actually… they don’t actually upgrade. And, like.
256 00:30:39.540 ⇒ 00:30:39.860 Amber Lin: Mmm.
257 00:30:39.860 ⇒ 00:30:58.260 Robert Tseng: Maybe you just accept, well, forget it. Like, that’s not the right people to go after. But somebody who comes in and is like, like, cookie sandwich, maybe they, like, they fall in love with the cookies more, and they end up, like, going to buy the bigger ones. Like, I don’t really know. So, like, I think that’s kind of, like.
258 00:30:58.560 ⇒ 00:31:10.500 Robert Tseng: You kind of have to keep going back and forth, like, you’re… you set some framework, you test it out, you try to see what the data is actually telling you, and then you adjust, like, what’s, like, a reasonable lifecycle, kind of, like, stage.
259 00:31:11.590 ⇒ 00:31:24.879 Amber Lin: Cool. I think the biggest takeaway today is that I realized I was doing the lifecycle based on this… based on the SaaS product company. It was, like, awareness, and then conversion, and whatever. It’s like, they just buy the cookie.
260 00:31:25.530 ⇒ 00:31:26.080 Robert Tseng: Yeah, yeah.
261 00:31:29.120 ⇒ 00:31:35.210 Amber Lin: Cool, okay, this was really, really helpful. I am now motivated to explore more. Thank you so much.
262 00:31:35.210 ⇒ 00:31:36.790 Robert Tseng: Great, yeah, alright.
263 00:31:36.790 ⇒ 00:31:37.769 Amber Lin: Yeah, have a good one.
264 00:31:37.770 ⇒ 00:31:38.820 Robert Tseng: Talk to you later. Bye.
265 00:31:38.820 ⇒ 00:31:39.900 Amber Lin: Great, bye.