Meeting Title: Communication Sync - Readme Date: 2025-10-13 Meeting participants: Henry Zhao, Robert Tseng, Uttam Kumaran
WEBVTT
1 00:00:34.960 ⇒ 00:00:36.130 Henry Zhao: Hey, Robert.
2 00:00:36.130 ⇒ 00:00:37.120 Robert Tseng: Hey, Henry.
3 00:00:59.400 ⇒ 00:01:00.240 Uttam Kumaran: Hey guys.
4 00:01:01.050 ⇒ 00:01:02.630 Henry Zhao: Alright, we’re all here, how’s it going?
5 00:01:04.260 ⇒ 00:01:05.909 Uttam Kumaran: Oh, a lovely day.
6 00:01:06.030 ⇒ 00:01:07.239 Uttam Kumaran: Lonely day.
7 00:01:08.010 ⇒ 00:01:08.950 Henry Zhao: Good.
8 00:01:08.950 ⇒ 00:01:10.900 Uttam Kumaran: I can’t believe it’s 3!
9 00:01:11.150 ⇒ 00:01:16.440 Uttam Kumaran: I haven’t blinked, like, 8 hours. Yeah.
10 00:01:17.180 ⇒ 00:01:18.229 Henry Zhao: They were like… Oh, yeah.
11 00:01:21.610 ⇒ 00:01:23.520 Uttam Kumaran: Yeah, they won’t buy fast.
12 00:01:25.580 ⇒ 00:01:32.870 Henry Zhao: Alright, so Robert, I think we can start talking about README. I think that’s the important thing that we want to talk about, right? Just to align, kind of, where we’re at on that.
13 00:01:33.820 ⇒ 00:01:34.500 Robert Tseng: Yep.
14 00:01:35.870 ⇒ 00:01:42.690 Henry Zhao: Do you currently need anything from me? Right now, I’m just gonna go back over kind of the call we had.
15 00:01:42.790 ⇒ 00:01:46.960 Henry Zhao: Were you able to get the access to MongoDB, and should I have access to that?
16 00:01:48.260 ⇒ 00:01:57.739 Robert Tseng: Yeah, I think the Mongo access is in 1Pass. I’m going in there now to just, like, see if they’ve updated my access to stuff.
17 00:01:58.160 ⇒ 00:02:03.590 Robert Tseng: I can share my screen, we can look at it, but… Let’s see…
18 00:02:07.700 ⇒ 00:02:10.279 Robert Tseng: Do you have access to the README OnePass?
19 00:02:11.200 ⇒ 00:02:12.579 Henry Zhao: Let me check
20 00:02:19.610 ⇒ 00:02:20.440 Henry Zhao: Yes.
21 00:02:21.260 ⇒ 00:02:22.560 Robert Tseng: Okay, cool.
22 00:02:24.080 ⇒ 00:02:31.290 Robert Tseng: So… Well, here I am in it right now.
23 00:02:31.970 ⇒ 00:02:37.009 Robert Tseng: Great, I’m assuming it’s this README project here.
24 00:02:38.370 ⇒ 00:02:43.020 Robert Tseng: Look at Data Explorer… Still not authorized.
25 00:02:43.950 ⇒ 00:02:46.160 Henry Zhao: What do we need to do here in MongoDB?
26 00:02:46.840 ⇒ 00:02:52.399 Robert Tseng: So, let’s see if we can…
27 00:02:57.450 ⇒ 00:02:59.220 Robert Tseng: The heck is going on?
28 00:03:05.990 ⇒ 00:03:17.689 Robert Tseng: Self-serve, so I shipped a few things to them that are… I mean, these are all kind of work in progress, but the first thing, when you jumped on a call with them, I was walking them through this, which was just kind of like…
29 00:03:17.820 ⇒ 00:03:22.079 Robert Tseng: A notebook to kind of break down their core conversion funnel.
30 00:03:22.210 ⇒ 00:03:31.210 Robert Tseng: Mongo Access, I was to verify that these numbers actually make sense, so…
31 00:03:31.960 ⇒ 00:03:43.240 Robert Tseng: Especially, like, user signups, just wanting to make sure, are there actually this many users in Mongo? Because if they’ve been created, or if they were there, they should have been created in Mongo.
32 00:03:44.850 ⇒ 00:03:52.240 Robert Tseng: Did I lose money already? Okay. Yeah, so that still seems to not be something we can verify. And then…
33 00:03:52.480 ⇒ 00:03:56.349 Robert Tseng: I think Mongo also has their Stripe
34 00:03:56.530 ⇒ 00:03:59.519 Robert Tseng: Transactions, or are they, like, kind of…
35 00:03:59.820 ⇒ 00:04:14.199 Robert Tseng: they’re able to keep, like, a payment log there, so that would also verify the subscriptions as well. So, that’s the purpose of Mongo. But even without it, I think it’s fine, it’s just that I won’t be able to say this without high confidence.
36 00:04:15.660 ⇒ 00:04:16.410 Robert Tseng: Yeah, I don’t know.
37 00:04:16.410 ⇒ 00:04:21.369 Henry Zhao: I’m trying to kind of figure out… yeah, I’m trying to kind of figure out what MongoDB is, and how that…
38 00:04:21.579 ⇒ 00:04:22.720 Henry Zhao: Plays a role in…
39 00:04:22.720 ⇒ 00:04:39.339 Uttam Kumaran: MongoDB is just gonna be their product database, right? So, MongoDB is going to be, like, the users, the types of objects that those users are creating, right? So README is a documentation platform. So you’re going to see users, you’re going to see documents, you’re gonna see all… it’s just their product database, basically.
40 00:04:39.340 ⇒ 00:04:39.990 Henry Zhao: Okay.
41 00:04:40.790 ⇒ 00:04:49.299 Uttam Kumaran: So this would be a source of truth for users, right? So similarly, on the default side, right, they don’t have a… they didn’t have an analytics warehouse, so they’re just like, go get our users.
42 00:04:49.410 ⇒ 00:04:52.960 Uttam Kumaran: From our users table, that’s just in the product database.
43 00:04:53.190 ⇒ 00:04:53.740 Robert Tseng: Yep.
44 00:04:53.970 ⇒ 00:05:12.679 Robert Tseng: So, same deal here, they don’t have an analytics database, they don’t have data warehousing or anything, so they just have the… they just have MongoDB. And yeah, so, like, you know, there was that doc that Alicia… Alicia had shared, and she had a few questions. I kind of took her questions, I broke it down into this analysis, I kind of…
45 00:05:13.240 ⇒ 00:05:36.370 Robert Tseng: it is kind of open-ended, because she had a bunch of questions, and I had to basically, like, storytell through it. You know, you can either go top-down or bottom-up, you know, and I chose to go top-down. So I just, like, core conversion funnel, kind of define what the segments are, free versus paid. You know, I asked her, like, on the call, you know, we’re walking through which ones you want to exclude. She decides to exclude enterprise.
46 00:05:36.370 ⇒ 00:05:40.600 Robert Tseng: And then she wants to know, okay, well, across the segments, like, which…
47 00:05:40.600 ⇒ 00:05:46.980 Robert Tseng: you know, let’s break down each step of the funnel and try to understand, like, what’s driving users each step, right? So…
48 00:05:47.080 ⇒ 00:05:57.510 Robert Tseng: That was kind of how this… how this analysis continues to get extended. I think what I didn’t have good visibility on on the first attempt was, like, on the payment side, like.
49 00:05:57.660 ⇒ 00:06:02.970 Robert Tseng: for their PLG users when they’re starting in some sort of
50 00:06:03.130 ⇒ 00:06:20.560 Robert Tseng: free plan. It’s not really free, they have… they just have, like, a free trial for, like, 14 days, and they have to upgrade. And so, kind of, like, thinking through how to, like, actually build that out in Amplitude. They do have limited tracking, but I think you can just be creative with, like, kind of how
51 00:06:20.760 ⇒ 00:06:25.000 Robert Tseng: you answer some of these, so I think that’s… that’s kind of where I think.
52 00:06:25.240 ⇒ 00:06:31.620 Robert Tseng: amplitude knowledge is necessary so that I can actually create, like, proxies and stuff.
53 00:06:31.750 ⇒ 00:06:45.230 Robert Tseng: And then in the last call that you were in last week, she was talking about launching AI features, and wanting to see usage. So I kind of created a separate notebook to just explore, like, what they’re doing.
54 00:06:45.230 ⇒ 00:06:57.529 Robert Tseng: Clearly, some of these features have already been live before their actual launch, which was around this time, August 6th. So, you know, this is a 30-day view, but if we
55 00:06:57.530 ⇒ 00:07:04.869 Robert Tseng: Drill into this, like… more, let’s just say, I don’t know, like, last 14 days…
56 00:07:06.390 ⇒ 00:07:19.340 Robert Tseng: this is what I expect to be walking her through more tomorrow, and we’re just gonna talk about features that are being used, like, in, you know, I kind of showed you last time, or some time ago on a loom, probably.
57 00:07:19.400 ⇒ 00:07:27.789 Robert Tseng: using the Amplitude Chrome extension, when you’re in their app, using our login, which is also in 1Pass. I think there is, like,
58 00:07:28.550 ⇒ 00:07:29.325 Robert Tseng: Do-do-do…
59 00:07:30.990 ⇒ 00:07:47.749 Robert Tseng: yeah, this quote-unquote God Mode account, which… you can create your own account, but if you just use mine, or you can ask for God Mode access, you just get access to all the features. So, yeah, I kind of went in there, played around with all the features, and kind of found all of the different events that they launched from there.
60 00:07:47.750 ⇒ 00:07:51.539 Robert Tseng: from their AI products. So that’s kind of where this data came from.
61 00:07:57.860 ⇒ 00:08:04.680 Robert Tseng: Yeah. So, with the access that she gave me, I think there’s a couple more things that I didn’t
62 00:08:04.800 ⇒ 00:08:09.289 Robert Tseng: finish that I wanted to extend,
63 00:08:10.360 ⇒ 00:08:14.060 Robert Tseng: I don’t remember off the top of my head what I wrote to her, so…
64 00:08:23.280 ⇒ 00:08:26.709 Robert Tseng: So, that can’t be done. Oh, right, so…
65 00:08:27.560 ⇒ 00:08:36.739 Robert Tseng: Yeah, I mean, I’ve done a little bit of digging now into understanding, like, free versus paid users, so this is kind of the extension of this one. I think some…
66 00:08:36.890 ⇒ 00:08:42.909 Robert Tseng: things that I discovered. Well, yeah, there isn’t, like, a clear, like, trial…
67 00:08:43.429 ⇒ 00:08:49.920 Robert Tseng: events that gets fired when your trial is up. So, I had to create these, like, segments that were, like.
68 00:08:50.150 ⇒ 00:08:57.769 Robert Tseng: Activity within the first 14 days, and then activity… if no active… you know, and if it was activity beyond 14 days.
69 00:08:57.870 ⇒ 00:09:00.449 Robert Tseng: And that’s kind of, like, how I…
70 00:09:00.960 ⇒ 00:09:14.420 Robert Tseng: kind of gave myself some sort of proxy for, like, before trial, after trial. She’s saying that after 14 days, users will get forced to the paywall, so they have to downgrade, or they have to upgrade.
71 00:09:14.960 ⇒ 00:09:19.899 Robert Tseng: So, I think that kind of gives enough separation to understand, like.
72 00:09:20.010 ⇒ 00:09:27.320 Robert Tseng: the free versus paid, but yeah, I mean, I think I wanted to create a couple more views around that to show
73 00:09:27.370 ⇒ 00:09:42.189 Robert Tseng: Does behavior change after their trial, or… or not? And, like, yeah, anyway, just… so I think that’s, like, a bit open-ended to me still, didn’t finish that. And then, yeah, because of their tier and amplitude.
74 00:09:42.350 ⇒ 00:09:47.449 Robert Tseng: we don’t get to create custom cohorts. I think that was something that I was calling out.
75 00:09:47.660 ⇒ 00:09:50.100 Robert Tseng: Yeah, like.
76 00:09:50.400 ⇒ 00:10:07.709 Robert Tseng: So, that’s… everything that you do has to be within chart, which is kind of annoying, because how I prefer to create cohorts is, like, going into here, and, you know, they’re already maxed out, and I can’t delete any of these, because even though none of these are being used, like, yeah, I just…
77 00:10:08.190 ⇒ 00:10:21.199 Robert Tseng: you can usually set multiple conditions on multiple events and, like, build something a bit more complex. Can’t really do that right now. So, that’s kind of something that we’re limited by, and…
78 00:10:21.330 ⇒ 00:10:27.110 Robert Tseng: Yeah, so I think that’s… That’s all I can really address for her now.
79 00:10:27.230 ⇒ 00:10:32.420 Robert Tseng: I think what I’m looking to get out of tomorrow’s call is, like.
80 00:10:32.770 ⇒ 00:10:35.999 Robert Tseng: Okay, well, I know that they’re trying to…
81 00:10:36.450 ⇒ 00:10:54.340 Robert Tseng: put some dashboard in front of their head of product, which I basically just took their… the notebook that I started to put together, and I just, like, dumped it into a dashboard. So it’s, like, it’s a bit cleaner, and whatever, and I would like to get this to her this week, and be like.
82 00:10:54.340 ⇒ 00:11:01.060 Robert Tseng: okay, this is the best we have. Give this to your head of product, you know, you already have the insights, whatever. So I think that’s the readout we’re working towards.
83 00:11:02.710 ⇒ 00:11:16.850 Robert Tseng: And then, yeah, I think beyond that, it’s just continuing to poke at them, letting us get access to more stuff, because I think with that, we’ll be able to give more advice on, like, what else they should be focused on.
84 00:11:17.020 ⇒ 00:11:17.640 Henry Zhao: Yeah.
85 00:11:17.640 ⇒ 00:11:18.699 Robert Tseng: Yeah, I don’t really.
86 00:11:18.700 ⇒ 00:11:19.180 Henry Zhao: I think…
87 00:11:19.180 ⇒ 00:11:20.339 Robert Tseng: Yeah, go ahead.
88 00:11:20.340 ⇒ 00:11:33.040 Henry Zhao: Yeah, I think the feedback I have is just, for these new clients, I think it would be helpful for me to get a little additional context before I start in terms of what we’re working on and what are the client’s priorities. I think
89 00:11:33.040 ⇒ 00:11:50.779 Henry Zhao: it’s a little bit hard to know where to go, or to understand some of the calls when I don’t have that context, you know what I mean? So, I think that’s why for Eden, and for this first week, like, I just felt a little bit lost onboarding, because I just didn’t know, like, where’s the North Star? Like, where are we trying to head to? What is the problem that they’re trying to solve?
90 00:11:51.140 ⇒ 00:11:58.910 Henry Zhao: So… like, I just see a bunch of different charts, but it’s like, what is that going to be used for, and what are the things that can actually be changed because of that, you know?
91 00:11:59.760 ⇒ 00:12:02.000 Robert Tseng: Sure.
92 00:12:02.000 ⇒ 00:12:11.340 Henry Zhao: So, like, default, it was very clear, because they were moving to PLG, they wanted to see how people are using the product, like, that was very clear. Eden became clear after I kind of met a few times with Cutter.
93 00:12:11.620 ⇒ 00:12:17.720 Henry Zhao: So yeah, if there’s any docs that you can share, like you did with me for some other clients, that’s always really helpful.
94 00:12:18.350 ⇒ 00:12:20.400 Robert Tseng: Yeah, I mean, have you looked at this at all?
95 00:12:20.400 ⇒ 00:12:21.000 Henry Zhao: Yeah.
96 00:12:21.190 ⇒ 00:12:27.459 Robert Tseng: Okay, well, this was the best I could put together. I feel like I just kind of summarized all the work that we did to start.
97 00:12:27.460 ⇒ 00:12:28.100 Henry Zhao: Hmm.
98 00:12:28.300 ⇒ 00:12:32.380 Robert Tseng: Kind of advising them on how they set up even any of their tracking plan.
99 00:12:32.720 ⇒ 00:12:48.889 Robert Tseng: And after that, then we started to collect data, some of the initial reporting work on some of the questions that they went through. I mean, I know you said you watched a bunch of this stuff, but like, yeah, I mean, like, this is… this is it, you know, like, I… it’s not like we’ve poured that many hours into it, like, I think it’s kind of…
100 00:12:49.050 ⇒ 00:12:53.910 Robert Tseng: like, I… I don’t know how… how much better I could have… could have done this, like, I… yeah.
101 00:12:53.910 ⇒ 00:13:07.130 Henry Zhao: No, but how do you know what story to tell in your Amplitudes dash if you… like, when you first onboard, right? Like, did you spend some time with Alicia, or how did you kind of get pointed in the right direction of even knowing, like, what story to tell?
102 00:13:08.710 ⇒ 00:13:14.130 Henry Zhao: Yeah, so, I mean, actually, no, this is my… I’ve talked to Alicia maybe one more time than you’ve seen her, so…
103 00:13:14.130 ⇒ 00:13:30.849 Robert Tseng: There was a previous stakeholder, Phoebe, yeah, she just gave me docs. She just gave me, like, a similar question, and I went around, poked around the products, and, yeah, I mean, I guess I have some benchmarks in my mind of, like, what I expect a product onboarding to look like. Okay.
104 00:13:31.370 ⇒ 00:13:32.190 Robert Tseng: you know.
105 00:13:32.600 ⇒ 00:13:52.379 Robert Tseng: I was able to chat with them on features that they were launching, and, like, because I kind of had my own point of view on, you know, how much of it is driven by within the product, using things like tooltips or coach marks, versus, like, having onboarding guides that are distributed over email. Like, but anyway, like, the focus is still, like, I guess.
106 00:13:53.510 ⇒ 00:14:14.980 Robert Tseng: at least… I think a general rule of thumb is for product analytics engagements, on a PLG motion, the focus is always going to be on onboarding. It’s just going to be, like, how do you get more users that are signing up to actually use the features, and which features are the most important? So being able to tie that to, like, paid sign-ups in some way.
107 00:14:14.980 ⇒ 00:14:28.549 Robert Tseng: So, like, I don’t know, I guess that’s the general kind of goal that I always have in mind when I enter these types of engagements. And even if that’s not the question they’re asking, I know that’s kind of how they want to frame it. So, that’s kind of how I started to outline my notebook.
108 00:14:28.550 ⇒ 00:14:29.690 Robert Tseng: And stuff.
109 00:14:30.630 ⇒ 00:14:38.339 Robert Tseng: But yeah, I mean, I guess, like, I understand you need to spend time with the clients, so I feel like you should just ask questions to Alicia, and just, like.
110 00:14:38.340 ⇒ 00:14:38.920 Henry Zhao: Yeah.
111 00:14:38.920 ⇒ 00:14:48.070 Robert Tseng: Yeah, I mean, I… you’re… I know you do that, you’re good at that, you just… you’ll keep asking questions until you figure it out, so I think it’s really just coming with that same, you know, perspective.
112 00:14:48.290 ⇒ 00:14:50.709 Robert Tseng: When you’re starting a new client like this one.
113 00:14:50.710 ⇒ 00:14:51.849 Henry Zhao: Yeah.
114 00:14:52.460 ⇒ 00:15:01.619 Henry Zhao: Yeah. Yeah, I think last week when I went into the meeting, my expectation was to kind of just… you had shared this before, so I was just gonna hear, like, the additional context you gave.
115 00:15:01.840 ⇒ 00:15:05.400 Henry Zhao: Around kind of what you pulled and what you were presenting to Alicia.
116 00:15:05.400 ⇒ 00:15:06.690 Robert Tseng: Yeah.
117 00:15:07.090 ⇒ 00:15:08.620 Henry Zhao: Yeah, so…
118 00:15:09.500 ⇒ 00:15:12.609 Robert Tseng: Yeah, I mean, all good. I don’t… like I said, I don’t think…
119 00:15:12.770 ⇒ 00:15:17.409 Robert Tseng: I know you’re really detail-oriented, and you want to, like, make sure you fully understand things, I think…
120 00:15:17.930 ⇒ 00:15:24.340 Robert Tseng: maybe I maybe operate under more ambiguity than most people do. I think that’s actually true. I feel like every… Yeah.
121 00:15:24.460 ⇒ 00:15:28.450 Robert Tseng: person that has come across, like, found that to be true.
122 00:15:28.450 ⇒ 00:15:36.920 Henry Zhao: Like I said, I think you onboard to new clients really quickly, where I’m… my experience is more like working at a full-time place, where I just onboard to one place, you know what I mean?
123 00:15:36.920 ⇒ 00:15:37.420 Robert Tseng: Yeah.
124 00:15:37.420 ⇒ 00:15:40.390 Henry Zhao: Probably need some training on… like, just need some practice on that, too.
125 00:15:40.390 ⇒ 00:15:45.810 Robert Tseng: Yeah, no, all good. I mean, you’re just making assumptions, and you’re calling out your assumptions as you go, and then…
126 00:15:45.860 ⇒ 00:16:05.610 Robert Tseng: anything that you’re uncertain about, like, there were plenty of things I was uncertain about, which is why when I was walking her through the notebook that I had put together, I was like, specifically, you know, I didn’t read this through, I was telling her, okay, is this the right window to look at, September 2025? She’s like, oh, actually, no, we don’t look at it by month, and actually different
127 00:16:05.610 ⇒ 00:16:12.939 Robert Tseng: different plans, like enterprise plans, I wouldn’t expect them to convert in a month. Like, that context she never gave me until I got on the call.
128 00:16:13.400 ⇒ 00:16:18.950 Robert Tseng: Okay, great. Like, that’s… that’s good context for me. So, I think that’s just kind of what I do. I, like, kind of…
129 00:16:18.960 ⇒ 00:16:30.850 Robert Tseng: go after what I, you know, generally can set up, and I know what assumptions I’m making, and so I can call them out and just adjust them as I get real feedback from them in our check-ins. So…
130 00:16:30.850 ⇒ 00:16:40.429 Robert Tseng: It’s… it’s… I mean, hopefully you see it as less pressure to come with everything buttoned up, like, I… I think we’re expected to go through a few iterations before we nail it.
131 00:16:40.540 ⇒ 00:16:47.619 Robert Tseng: And yeah, so I think that’s… that’s… that’s maybe something that could help, kind of,
132 00:16:48.830 ⇒ 00:16:52.070 Robert Tseng: Help you approach it with less,
133 00:16:52.370 ⇒ 00:16:59.709 Robert Tseng: I don’t know, lower your bar a bit, or like, I don’t know, like, it’s okay if you’re figuring it out with the client. At least you’re.
134 00:17:00.180 ⇒ 00:17:04.180 Robert Tseng: through it in a way that they’ve never seen before, so they’re not gonna know either. Yeah.
135 00:17:04.180 ⇒ 00:17:04.770 Henry Zhao: Yeah.
136 00:17:05.270 ⇒ 00:17:05.839 Robert Tseng: Yeah.
137 00:17:05.849 ⇒ 00:17:13.639 Henry Zhao: Yeah, I think my worry is just sometimes I feel like the charts I make are very, like, superficial, or… I said this last week when you presented, right, that they’re very obvious.
138 00:17:16.339 ⇒ 00:17:16.659 Robert Tseng: Yeah.
139 00:17:16.660 ⇒ 00:17:17.520 Henry Zhao: I mean, like…
140 00:17:18.240 ⇒ 00:17:37.010 Robert Tseng: No, and I think there… that will be… that will be true. Like, I think, you know, averages are always misleading. Like, I think I’m very, very, like, against showing averages, but they help establish some sort of baseline if they don’t have any, like, kind of idea. Like, I know that when I show this chart.
141 00:17:37.190 ⇒ 00:17:41.240 Robert Tseng: I need to break it down more, so I need to know what the segments they care about are.
142 00:17:41.650 ⇒ 00:17:42.160 Henry Zhao: Yeah.
143 00:17:42.160 ⇒ 00:17:55.069 Robert Tseng: And, like, that’s… that’s something that won’t really come up until I actually show it to them, and they’re gonna… they may not know what segments they want to look at, so being able to come in with some perspective of how I would cut up the data, I think, is also part of the consultation.
144 00:17:55.070 ⇒ 00:18:05.380 Uttam Kumaran: Yeah, so segmentation for me is also, like, the most important. Like, the moment you put up this average, you’re gonna get asked, okay, what’s the breakdown by tier? And so, like, you need to go…
145 00:18:05.560 ⇒ 00:18:25.189 Uttam Kumaran: that, like, one more step, and this is, like, it… for me, like, again, like, I’ve done a bunch of analysts, I think of everything, like, 80-20. So I’m always trying to find the 80-20 story, and find the driver story of, like, hey, if a number is high, what is contributing to that? If a number is low, what is also contributing to that? And that’s…
146 00:18:25.190 ⇒ 00:18:34.809 Uttam Kumaran: That’s, like, my generic, like… that’s how I approach everything. Like, there’s always gonna be some type of 80-20 story, and I go search for that.
147 00:18:34.810 ⇒ 00:18:51.930 Uttam Kumaran: And usually that will lead you down a bunch of things. If you’re thinking about churn, there’s always, like, a leaky bucket story. Okay, like, how are we losing people? Where is the maturity of the loss coming from? You know, like, that could be revenue loss, that could be user loss. So when you’re talking about
148 00:18:52.010 ⇒ 00:19:08.690 Uttam Kumaran: funnels, there’s always, like, okay, out of the 80… let’s say they… let’s say our data shows that, like, 80% churn, like, 80%, we get 20% make it through the funnel, where is the majority of the loss coming from? And that’s a great thing for you to just go drive down and say, like.
149 00:19:08.730 ⇒ 00:19:23.279 Uttam Kumaran: we have a fucked up, like, modal somewhere, or we have some messed up part of the onboarding flow that loads slowly, and we need to fix that, right? And, like, that’s… that’s usually it. From my mind, and, like, what I’ve seen a lot of, you know?
150 00:19:23.940 ⇒ 00:19:24.560 Henry Zhao: Yeah.
151 00:19:25.390 ⇒ 00:19:48.369 Robert Tseng: Yeah, and then, like, I think you’ll… you’ll know, like, you’ll notice things, like, I didn’t have the context for what happened here until I showed it, walked her through it, right? I was like, I don’t know what dropped here, like, something broke, or in your product, and she was like, oh, I didn’t actually realize we shipped some features there. I guess part of the product was broken, and so that explains why engagement, like, randomly dipped during this week, but then it kind of went back up.
152 00:19:48.370 ⇒ 00:19:57.110 Robert Tseng: I wouldn’t really call it stable, but, like, you know, whatever. So, like, I think being able to call out those inflection points is important, too.
153 00:19:57.110 ⇒ 00:20:00.339 Robert Tseng: But yeah, these are all level 1 insights at this point, you know.
154 00:20:00.340 ⇒ 00:20:00.780 Henry Zhao: Okay.
155 00:20:00.780 ⇒ 00:20:20.619 Robert Tseng: own inflection point, whatever. Like, we’re not really doing driver analysis yet. I think, like, eventually, that’s kind of what this needs to get towards. Like, once we have a better sense of the segmentation, and, like, we actually can create custom cohorts that are kind of chaining together behaviors, you know, when user uses, like.
156 00:20:20.650 ⇒ 00:20:38.700 Robert Tseng: XYZ AI features, they’re converting at, like, 50%, or whatever. Like, being able to, like, pull out a nugget like that, I think that’s kind of, like, what this… what the amplitude work ends up becoming. But they’re not gonna be able to know… they’re gonna look at this, and it just looks like spaghetti to them. They don’t know how to, like…
157 00:20:38.700 ⇒ 00:20:39.090 Henry Zhao: Yeah, exactly.
158 00:20:39.090 ⇒ 00:20:55.569 Robert Tseng: mix and match these things. But when I look at this chart, I can already pick up a few things that I would investigate. You know, like, obviously, to me, this is, like, this is a starting point. If these are all AI features, everyone’s hitting this step first, and so maybe there’s this actual, like, step
159 00:20:55.570 ⇒ 00:21:03.499 Robert Tseng: maybe there’s a flow to the different AI features that users are going… are flowing through as they’re exploring the product.
160 00:21:03.500 ⇒ 00:21:14.129 Robert Tseng: So that would be a… that would be a chart I would add to try to visualize that. They have something in here called, like, a journeys chart or whatever. So, or then I would maybe be looking
161 00:21:14.190 ⇒ 00:21:16.730 Robert Tseng: yeah, I mean, I would see, like.
162 00:21:17.260 ⇒ 00:21:22.810 Robert Tseng: yeah, I mean, I guess some of these features kind of… Especially, you know.
163 00:21:22.810 ⇒ 00:21:47.740 Robert Tseng: wanting to tie this back to some of their core features, which maybe I know a bit better than you, because I’ve played around their product better. But yeah, I guess these are all, like, additional paid premium features that are supposed to enhance existing features, so I would want to visualize something relating how these features are maybe driving activity on their core features as well. So, like, I don’t know, there’s, like, a couple ideas that I have just from looking at this
164 00:21:47.740 ⇒ 00:21:49.120 Robert Tseng: chart of, like.
165 00:21:49.120 ⇒ 00:22:01.839 Uttam Kumaran: Like, again, a good thing you should show is, like, one, you should show the paid users, and all you need to do is create a flag, have used an AI feature, and then you can look at, out of the… our most paying users, have they… how much percentage of that
166 00:22:02.180 ⇒ 00:22:12.290 Uttam Kumaran: like, have used an AI feature. What you’re probably gonna see is probably what you’re seeing across every AI tool, is they shipped a bunch of stuff and nobody’s using it. And so, okay, you have a product, you have, like, a
167 00:22:12.490 ⇒ 00:22:25.690 Uttam Kumaran: you have, like, a… people don’t know that these are… either people don’t know, or they don’t like it. It’s, like, useless. And, like, it’s okay, you chase down what’s right. Because if it’s worth it, then people will use it, right? Yeah. If it’s not worth it…
168 00:22:25.690 ⇒ 00:22:36.620 Robert Tseng: That’s kind of what we’re saying here. Like, I mean, fortunately, everything, everyone here is a paid user in order to use these AI features, but they have this launch, there’s a spike, whatever, and then it kind of just flattens back out, no one’s using.
169 00:22:36.620 ⇒ 00:22:42.539 Uttam Kumaran: Yeah, but it’s 16, 15, that’s abysmal, right? Like, it’s like… it’s like, nobody’s using this. They probably spent…
170 00:22:42.670 ⇒ 00:22:45.830 Uttam Kumaran: Hundreds of thousands of dollars developing each of these, right?
171 00:22:45.950 ⇒ 00:22:47.130 Uttam Kumaran: So…
172 00:22:47.260 ⇒ 00:22:53.749 Uttam Kumaran: you have to go back and be like, okay, guys, what happened here? Did we… did we just, like… are we not pushing it? Or is it, like.
173 00:22:53.860 ⇒ 00:22:55.160 Uttam Kumaran: Hard to use.
174 00:22:55.320 ⇒ 00:23:06.079 Uttam Kumaran: And then you identify, okay, we’ve done everything, I’m pushing it, so everybody’s aware of these features. Like, we have pop-ups, we’ve done emails, whatever. Okay, then let’s find out, okay, who is using it? What’s unique about them?
175 00:23:06.170 ⇒ 00:23:21.589 Uttam Kumaran: You know, so, like, you know, you find out, okay, like, we see consistently one… one… for example, this is what I used to do in product, is I used to go, like, okay, this one company’s using it, okay, I’m just gonna call them. And I’m like, wait, why the heck do you guys like this? What’s the deal? Okay, perfect, now you have the store.
176 00:23:21.710 ⇒ 00:23:23.549 Uttam Kumaran: That’s… that’s,
177 00:23:24.120 ⇒ 00:23:34.049 Uttam Kumaran: that’s kind of it. Like, I don’t… I think sometimes we underestimate, like, they’re not even able to even do these types of basic analysis, Henry. Like, they’re not even looking for much.
178 00:23:34.370 ⇒ 00:23:37.839 Henry Zhao: They’re just looking for someone to go, like, dig and explore.
179 00:23:38.570 ⇒ 00:23:38.890 Uttam Kumaran: Okay.
180 00:23:39.330 ⇒ 00:23:44.470 Uttam Kumaran: You know, and ask, like, the earnest question of, like, who’s using it? Why are they using it?
181 00:23:44.490 ⇒ 00:23:47.340 Henry Zhao: Yeah. But this is helpful guidance.
182 00:23:48.580 ⇒ 00:24:03.530 Robert Tseng: Okay, great. Yeah, I mean, the last thing I’ll say is, like, I don’t know, I think the notebook style helps. Like, obviously you know that everything is kind of in progress. I create my own follow-up questions. They’re not asking them to me, but, like, this is me if I were kind of… I timebox this. I’m like, I’m only spending an hour or two hours on this.
183 00:24:03.530 ⇒ 00:24:16.780 Robert Tseng: do whatever I can, kind of synthesize my call-outs, my assumptions, whatever. If I have a couple nuggets I can pull out and share in a call, that’s great. And then I’m gonna just write a few follow-up questions to let them know if I had more time, this is what I would look into.
184 00:24:16.780 ⇒ 00:24:29.120 Robert Tseng: And so I bring that to the call, and they’re like, oh, wait, that’s a good question, I want you to look into that. Then that kind of cues up the next step. Okay. Yeah, I think that’s kind of, like, kind of how you can keep the conversation going. Obviously, the pace on this is pretty slow right now.
185 00:24:29.120 ⇒ 00:24:32.219 Robert Tseng: But yeah, I think, there’s, you know.
186 00:24:32.250 ⇒ 00:24:36.460 Robert Tseng: I don’t expect you, you know, from a time split perspective.
187 00:24:36.510 ⇒ 00:24:43.770 Robert Tseng: you should be spending, you know, assuming you have, whatever, like, 10 hours a week on README, I think, like.
188 00:24:44.040 ⇒ 00:24:50.349 Robert Tseng: I don’t know, like, 25% of your time should actually be in amplitude, like, exploring stuff.
189 00:24:50.720 ⇒ 00:25:04.110 Robert Tseng: Yep, like, 50% of your time should just be thinking about, like, how you would, like, architect their… their, like, their data strategy, really. And then, like, the other 25% is, like, packaging things into, like.
190 00:25:04.190 ⇒ 00:25:15.690 Robert Tseng: getting ready to, like, walk someone through this, yeah, like, I spend more time thinking through the messaging than I actually do on, like, these charts. These charts take, like, minutes to build, right? So, like…
191 00:25:15.690 ⇒ 00:25:16.190 Henry Zhao: Yeah, yeah.
192 00:25:16.190 ⇒ 00:25:28.229 Robert Tseng: Yeah, it’s a lot more important how we drive the conversation, and like, you know, what steps… what doors get opened past this. So, that’s kind of how I think about.
193 00:25:28.230 ⇒ 00:25:28.740 Henry Zhao: Okay.
194 00:25:28.740 ⇒ 00:25:29.940 Robert Tseng: Like, allocations-wise.
195 00:25:29.940 ⇒ 00:25:41.290 Henry Zhao: So for this week, for now, I’m gonna probably just watch over some of those videos again, dig around in the product, kind of look at the work you did, and just try to get a better understanding. I think I still probably need to spend some time there.
196 00:25:41.290 ⇒ 00:25:50.839 Robert Tseng: Yeah, all good. I’m gonna, like, block off one more hour, I’m gonna do some stuff before I prepare for tomorrow’s call, and then, I mean, you’ll see. I’m gonna try to drive her towards, like, a couple…
197 00:25:51.210 ⇒ 00:25:55.740 Robert Tseng: a couple things that I, I want to steer this, this, this…
198 00:25:55.810 ⇒ 00:26:02.920 Robert Tseng: kind of, engagement towards, and just kind of see… see where they buy it. Like, obviously, AI feature, kind of, like.
199 00:26:02.940 ⇒ 00:26:18.170 Robert Tseng: tracking and kind of development, I think, is a big part on their mind, because they just… they spent the past quarter, like, building all these things, but clearly it’s not being used, so trying to talk through, like, how do we activate this more, and, like, kind of being a thought partner to her on, like.
200 00:26:18.170 ⇒ 00:26:29.669 Robert Tseng: being able to drive more of adoption around that feature set is important. And then I also want to get more involved on the pricing side. They rolled out a bunch of pricing things, but, seems like that’s something that they’re not really, like.
201 00:26:29.680 ⇒ 00:26:35.609 Robert Tseng: kind of, looking into too closely, so I think that would be where I would be pushing her tomorrow.
202 00:26:36.080 ⇒ 00:26:41.829 Henry Zhao: Okay, and then I think… Sorry, and I just think whenever there’s a new client, probably the first thing I want to just ask you, Robert, is like…
203 00:26:41.830 ⇒ 00:27:01.009 Henry Zhao: what is this client’s deal with us? Like, is this a vague problem where we’re just, like, saying we’re going to fix your data, or is this a specific, we want ABC, and these are A and B and C? And how much time should we be spending on each, just to kind of have that North Star, and it will, I think, guide me more on where I should be spending my time and how to tackle the problem.
204 00:27:01.150 ⇒ 00:27:04.070 Robert Tseng: Okay, yeah, that’s fair. I mean, I would say…
205 00:27:04.360 ⇒ 00:27:22.830 Robert Tseng: you know, for early-stage clients, which I would still consider README, there’s always one narrowly scoped thing that we’re doing. Maybe they want one dashboard, maybe they want to get one tool working, whatever, some technical solution, but I would say it’s mostly strategy. It’s mostly just, like, gearing up for the next thing that…
206 00:27:22.830 ⇒ 00:27:30.590 Robert Tseng: we can work on, right? Like, I always share the example of Eden came in, had me come in and just audit their mix panel.
207 00:27:30.680 ⇒ 00:27:47.600 Robert Tseng: We haven’t even touched Mixpanel in, I don’t know, 10 months or whatever. But, like, yeah, I, like, I build out one dashboard for their web analytics team on… in Mixpanel, and then I opened the conversation to a bunch of other things and took it in a completely different direction. So, I think that’s kind of, like, the head that you should have when you’re going in.
208 00:27:47.600 ⇒ 00:27:58.010 Henry Zhao: Okay. Yeah. So that’s what we’re doing. So whenever we have a client, we should kind of do what they signed us up to do, but then also start looking for that, those things that we can upsell, or those other things where we can provide value.
209 00:27:58.010 ⇒ 00:28:01.660 Robert Tseng: They often don’t know what they need, like… Absolutely, they have no idea what they need.
210 00:28:01.660 ⇒ 00:28:02.399 Uttam Kumaran: Yeah, they have no.
211 00:28:02.400 ⇒ 00:28:03.139 Henry Zhao: Very helpful.
212 00:28:03.320 ⇒ 00:28:17.279 Uttam Kumaran: So their only understanding of great data is, like, give me a dashboard. We come and we… we do what we know is best, which is, like, you have to go question this data, tell these stories. And so another thing, this is where, like, I…
213 00:28:17.280 ⇒ 00:28:26.499 Uttam Kumaran: this is not gonna be the case in, like, data engineering and some of the other engineering work you’re doing. It’s like, we gotta, like, talk. Like, until you have the internal monologue.
214 00:28:26.500 ⇒ 00:28:45.850 Uttam Kumaran: as you can see, Robert and I both ask a question and answer it, and then ourselves, ask it, answer it, right? Like, until you generate that, you have to send… so I would just, like, say, hey, I’m going down this rabbit hole about, like, I’ve noticed this on the day of going down this rabbit hole, just, like, live blog the things that you’re seeing. That’s all… that honestly is what
215 00:28:45.850 ⇒ 00:28:49.550 Uttam Kumaran: you should be building the notebook off of. But this is gonna… you’re never gonna arrive
216 00:28:49.560 ⇒ 00:28:59.870 Uttam Kumaran: at, like, the cleanest answer, but you’re trying to get enough of a story and enough pieces of evidence to tell a story. So I would send… just send, like, live blog your thoughts into Slack.
217 00:28:59.910 ⇒ 00:29:17.199 Uttam Kumaran: But, like, until you… and then when you hit a snag, just say, hey, I’m thinking about this hypothesis, I hit a snag, what else could I look at? One of us could answer. But, like, I wouldn’t… I would just do that. Like, we… until… until you have that live… until you have that monologue in your brain that can answer the things.
218 00:29:17.260 ⇒ 00:29:19.180 Uttam Kumaran: We have to bounce ideas.
219 00:29:19.930 ⇒ 00:29:34.730 Uttam Kumaran: You know, so I would just send those to Slack, say, I noticed that, like, certain amount of… I noticed some 80-20 thing here, or I noticed some obscure numbers, or I noticed the number’s really low, I’m just gonna live blog how I get down to it, you know? It’ll help you articulate your thoughts.
220 00:29:35.780 ⇒ 00:29:36.809 Henry Zhao: Okay, sounds good.
221 00:29:38.970 ⇒ 00:29:39.710 Robert Tseng: Oh!
222 00:29:41.850 ⇒ 00:29:42.430 Uttam Kumaran: Okay.
223 00:29:42.430 ⇒ 00:29:42.960 Henry Zhao: Bye.
224 00:29:42.960 ⇒ 00:29:43.949 Robert Tseng: Anything else?
225 00:29:43.970 ⇒ 00:29:58.690 Henry Zhao: So, when should we maybe talk, just thinking about the future, like, so, when January comes around, or whenever I’m able to bring on new clients, or whenever I want to go to a conference or something like that to help out with more of the sales side of it.
226 00:29:58.900 ⇒ 00:30:10.019 Henry Zhao: Do we want to have a call, the three of us, to just kind of go over the business model? How do we bring on a new client? Because I think that’ll also help me with my existing clients, just to kind of understand the business model.
227 00:30:10.020 ⇒ 00:30:24.289 Henry Zhao: understand how you guys have done upsells in the past, understand what we’re bringing… what clients are bringing on us for, and then how you guys kind of added to that beyond the original scope, etc. So, just get a lot more context there, and have a formal discussion about it.
228 00:30:25.040 ⇒ 00:30:32.909 Uttam Kumaran: I mean, I would rather just… we do that in our… one of our Monday sales meetings, and we’ll just carve out one of these Mondays to talk about that. I mean, we talked a little bit about it
229 00:30:33.140 ⇒ 00:30:40.580 Uttam Kumaran: today in delivery, right? How we upsell. So, I would rather do it in a forum where a bunch of other people can learn, so…
230 00:30:40.580 ⇒ 00:30:41.140 Henry Zhao: Okay.
231 00:30:41.140 ⇒ 00:30:43.610 Uttam Kumaran: We can carve out time next week to talk about
232 00:30:44.020 ⇒ 00:30:50.900 Uttam Kumaran: whatever, but I guess, like, I would want to know some of the questions in advance, because most of this we’ve written down, it’s in Notion already.
233 00:30:51.320 ⇒ 00:30:53.870 Henry Zhao: Oh, okay, can you share those notions if I… in case I don’t have them?
234 00:30:54.760 ⇒ 00:30:58.920 Uttam Kumaran: You have it. It’s just… go just type in sales into Notion, and…
235 00:30:59.140 ⇒ 00:31:02.089 Uttam Kumaran: There’s about 300 years worth of
236 00:31:02.820 ⇒ 00:31:07.369 Uttam Kumaran: stuff we wrote about how we sell around here, so… I would read all of that first.
237 00:31:07.370 ⇒ 00:31:19.089 Robert Tseng: a couple meetings that were helpful, especially when we were onboarding some of the earlier, like, go-to-market people, because we walked through all the docs there. But yeah, I think I kind of teased that opportunity. I mean…
238 00:31:19.090 ⇒ 00:31:35.970 Robert Tseng: I know that Eden… Eden was asking me to go to New Mexico, and I’m like, I’m not going to New Mexico, I’m sorry, it’s just too far, and I’m not going this month. So, I mean, if that’s… if that’s something you want to do, I think that could be, like… I think that’s an expansion opportunity, like… Yeah.
239 00:31:36.160 ⇒ 00:31:47.359 Robert Tseng: I think if you… you went there, you’d understand their business better. I mean, I’m sure everything… I mean, yeah, we’ll… we’ll kind of work out what that looks like. And, yeah, while you’re there, like, I can also…
240 00:31:47.660 ⇒ 00:31:51.619 Robert Tseng: Kind of talk you through… we can… we can talk more closely about how you would
241 00:31:51.740 ⇒ 00:32:08.990 Robert Tseng: kind of what… what I’m hoping you to discover while you’re… while you’re there. So, like, I think… and I also told you about the conference, we can… we can kick that down the line a bit more. We’re not going to end up going to HealthUSA this… this… this week. So, but yeah, so, like, I think there… there will be opportunities where you’re kind of…
242 00:32:09.040 ⇒ 00:32:17.809 Robert Tseng: if you want to take them, we’ll just send you, and then we’ll have to just coach you while you’re there. Sometimes you’ll be there together, sometimes you’ll be there by yourself.
243 00:32:18.170 ⇒ 00:32:25.229 Henry Zhao: Yeah, you gave me some coaching at, in Chicago, and I will read those notions, I think that’ll be really good, and then at our one-on-one, we can talk about Albuquerque.
244 00:32:25.520 ⇒ 00:32:31.319 Henry Zhao: Okay, cool. And just get some guidance on what we need to… what I should be doing in Albuquerque, and what I should be bringing to the business.
245 00:32:31.470 ⇒ 00:32:32.490 Robert Tseng: Okay, cool.
246 00:32:32.490 ⇒ 00:32:33.440 Henry Zhao: Yeah? Okay.
247 00:32:34.360 ⇒ 00:32:35.310 Robert Tseng: Sounds good.
248 00:32:35.310 ⇒ 00:32:36.210 Uttam Kumaran: Alright.
249 00:32:36.210 ⇒ 00:32:38.680 Henry Zhao: Perfect, thank you guys for the help, this has been great.
250 00:32:39.050 ⇒ 00:32:46.649 Uttam Kumaran: Great, yeah. Of course, yeah, the only thing is livestream your thoughts. Analysis is a much different flow than engineering work.
251 00:32:46.800 ⇒ 00:32:59.649 Uttam Kumaran: You have to, like… it’s gonna be so helpful. You’re gonna… you can easily get this type of stuff from us, but it just has to go in Slack, because I can’t… I just can’t meet. But if you send it to in Slack, in the analysis channel, or in the data channel, I’ll give you my…
252 00:32:59.650 ⇒ 00:33:03.760 Henry Zhao: Okay. Yeah, I was about to ask which channel, I think the data team channel, right, probably.
253 00:33:04.200 ⇒ 00:33:09.579 Robert Tseng: Software does a good job of this. He, like, records looms, and I always give really detailed feedback. So, I mean, I wish people.
254 00:33:09.580 ⇒ 00:33:10.130 Henry Zhao: Yeah, you do.
255 00:33:10.330 ⇒ 00:33:11.090 Robert Tseng: Yeah.
256 00:33:11.420 ⇒ 00:33:12.670 Henry Zhao: Yeah, I’ll start doing…
257 00:33:12.670 ⇒ 00:33:22.799 Uttam Kumaran: It’s just something… it’s just something that everybody has to get used to, because sometimes we’re like, well, I’m gonna do it for 3 days, and then I’ll get somewhere, and I’m like, dude, if you would have asked me 10 minutes into this thing, I could have given you the answer.
258 00:33:22.800 ⇒ 00:33:24.029 Henry Zhao: Yeah, no, I’ll do that much.
259 00:33:24.030 ⇒ 00:33:25.929 Uttam Kumaran: I think everybody’s gonna… yeah, okay, okay.
260 00:33:25.930 ⇒ 00:33:26.520 Henry Zhao: Cool.
261 00:33:27.550 ⇒ 00:33:29.530 Henry Zhao: Alright, thanks guys, have a nice rest of your day.
262 00:33:29.530 ⇒ 00:33:30.900 Uttam Kumaran: to you guys. Yeah, you too.
263 00:33:30.900 ⇒ 00:33:31.530 Henry Zhao: Yep. Bye.