Meeting Title: Analytics Pairing Date: 2025-08-19 Meeting participants: Robert Tseng, Annie Yu
WEBVTT
1 00:00:41.810 ⇒ 00:00:42.930 Annie Yu: Hello, Robert.
2 00:00:44.330 ⇒ 00:00:45.350 Robert Tseng: Oh, hey, Annie!
3 00:00:46.290 ⇒ 00:00:50.240 Robert Tseng: … Alright, one sec, I’m gonna finish this message.
4 00:00:59.450 ⇒ 00:01:09.590 Robert Tseng: Okay, yeah, … I am. Thanks for your, …
5 00:01:10.010 ⇒ 00:01:12.410 Robert Tseng: response on the message, I know.
6 00:01:12.520 ⇒ 00:01:14.239 Robert Tseng: Not always the most comfortable.
7 00:01:14.420 ⇒ 00:01:17.379 Robert Tseng: thing to say, and I don’t know, I… I prefer to…
8 00:01:17.720 ⇒ 00:01:25.140 Robert Tseng: talk rather than, use messages. So, if you have any other questions around that, just let me know, but I guess, …
9 00:01:26.300 ⇒ 00:01:32.290 Robert Tseng: Yeah, obviously, I just, you know, there are… W-we’re…
10 00:01:32.740 ⇒ 00:01:38.619 Robert Tseng: Something that we’ve been talking about more is, like, … wanting to…
11 00:01:41.070 ⇒ 00:01:45.270 Robert Tseng: Invest more in the people that we have, …
12 00:01:45.680 ⇒ 00:02:03.640 Robert Tseng: Okay, I mean, I don’t know, maybe you don’t think about it, but obviously we had to make some decisions, and we let some people go recently, so… … yeah, I think that obviously has not been good for morale on team, and I don’t… ugh, just a very difficult decision to make for us as well, so…
13 00:02:03.880 ⇒ 00:02:12.080 Robert Tseng: I mean, with this… I mean, with your situation, I’m not trying to scare you and make you seem like there’s a bunch of risk or whatever, but I’m just…
14 00:02:12.400 ⇒ 00:02:17.750 Robert Tseng: Trying to get ahead of that, and, you know, just learning… from…
15 00:02:17.890 ⇒ 00:02:20.360 Robert Tseng: the previous situations, like, I feel like…
16 00:02:20.530 ⇒ 00:02:27.409 Robert Tseng: We didn’t… we waited too long before we could give people a chance, and …
17 00:02:28.560 ⇒ 00:02:34.220 Robert Tseng: Yeah, I think, that’s… You know, even if it’s an uncomfortable thing to…
18 00:02:35.020 ⇒ 00:02:41.059 Robert Tseng: to… to do, like, I… I… yeah, we just don’t have all these, …
19 00:02:41.260 ⇒ 00:02:50.260 Robert Tseng: safety nets in place, that maybe, you know, some… a bigger organization would have. You know, I think…
20 00:02:50.820 ⇒ 00:02:54.549 Robert Tseng: I… Have been in roles where
21 00:02:54.860 ⇒ 00:03:02.379 Robert Tseng: yeah, you get, like, 6 months to a year to figure it out, and … I don’t know, I just feel like we don’t really have…
22 00:03:02.570 ⇒ 00:03:06.550 Robert Tseng: Six months to a year to figure things out, so… …
23 00:03:06.740 ⇒ 00:03:10.519 Robert Tseng: Yeah, anyway, I’m not… I hope I’m not making anything
24 00:03:10.870 ⇒ 00:03:15.259 Robert Tseng: worse by saying whatever I’m saying, but, yeah, I…
25 00:03:16.830 ⇒ 00:03:23.180 Robert Tseng: I just… I just… okay, anyway, I just wanted to at least give you some…
26 00:03:23.510 ⇒ 00:03:36.470 Robert Tseng: reassurance that, like, I… I think this is me trying to, like, learn from those previous, situations, and, like, just kind of get ahead of things. Like, obviously, like, you were doing great work on… on Eden, and…
27 00:03:36.720 ⇒ 00:03:48.730 Robert Tseng: like, I mean, we’re nowhere near, like… like, I… again, we’re not… it’s not like we’ve been evaluating, you know, other… other options or whatever, but I think just from, like, a…
28 00:03:49.010 ⇒ 00:03:53.169 Robert Tseng: Long-term planning perspective, and just wanting to see,
29 00:03:53.510 ⇒ 00:04:04.309 Robert Tseng: Like, a clearer path to you being able to expand into… expand your skill set in a way that’s able to serve, like, our current clients’ needs, rather than us having to go and
30 00:04:04.470 ⇒ 00:04:11.040 Robert Tseng: hire another product analyst right now, like, that would be my preference. So, I think that’s why I’m…
31 00:04:11.190 ⇒ 00:04:25.019 Robert Tseng: trying to spend some more time with you, and it won’t always be, like, this me blah blah blahing at you. Like, I think we should just do some work together, and you can just, like, learn from me as we’re going through things. …
32 00:04:25.390 ⇒ 00:04:27.660 Robert Tseng: Yeah. Do you have any kind of…
33 00:04:29.490 ⇒ 00:04:32.550 Robert Tseng: Thoughts or questions about that so far?
34 00:04:32.930 ⇒ 00:04:35.349 Annie Yu: I would say, I think…
35 00:04:35.460 ⇒ 00:04:41.260 Annie Yu: at least having a few pairing sessions would be really great, because I, I think…
36 00:04:41.940 ⇒ 00:04:51.090 Annie Yu: I thought I had an understanding of the business context, but then I think it’s the, … actual…
37 00:04:51.420 ⇒ 00:04:59.030 Annie Yu: Like, what’s doable, what’s not, and, like, the hypothesis that you set up. So, all those kind of things.
38 00:04:59.350 ⇒ 00:05:10.860 Annie Yu: where it would be great to learn from you, so I think it would be great, to have these sessions. Okay. I do want to ask one question about…
39 00:05:11.050 ⇒ 00:05:14.810 Annie Yu: The… so the coaching session.
40 00:05:15.720 ⇒ 00:05:16.260 Annie Yu: I….
41 00:05:16.260 ⇒ 00:05:18.480 Robert Tseng: Is that with Brian? Yeah, you can stop that.
42 00:05:18.480 ⇒ 00:05:32.500 Annie Yu: Yeah, okay, okay. I’m not sure. I think I’ve learned a lot, and I think it’s helpful, but it’s just not directly tied to my day-to-day tasks, like, let alone, like, product analytics, so I was not sure.
43 00:05:32.630 ⇒ 00:05:36.320 Annie Yu: … Yeah.
44 00:05:36.790 ⇒ 00:05:40.240 Robert Tseng: Okay, yeah, I mean, you can, you can pause on that.
45 00:05:40.600 ⇒ 00:05:47.759 Robert Tseng: Unless you find it helpful, like, that’s still a resource for you. Like, yeah, I think that’s… it’s optional for you.
46 00:05:49.060 ⇒ 00:05:56.809 Annie Yu: No, yeah, no, I think it’s helpful, definitely, but I, yeah, I… I just want to make sure…
47 00:05:57.050 ⇒ 00:06:04.159 Annie Yu: the intent… I think the intent for that session was about analytics engineering, because that’s what
48 00:06:04.320 ⇒ 00:06:09.120 Annie Yu: Awesha and Utam have talked about in the threads, …
49 00:06:09.560 ⇒ 00:06:16.509 Annie Yu: So, unless I’m leveraged… I’m utilizing him in the wrong way, but I don’t think so.
50 00:06:17.520 ⇒ 00:06:18.220 Robert Tseng: Okay.
51 00:06:21.580 ⇒ 00:06:31.849 Robert Tseng: Yeah, well, I mean, I spoke with Utam, and he said that he just wanted you to bring whatever you’re working on to the session, and that you would just be able to work through it with, with Brian, but …
52 00:06:32.360 ⇒ 00:06:34.330 Robert Tseng: Seems like that’s not the case, yeah.
53 00:06:34.540 ⇒ 00:06:40.789 Annie Yu: I mean, yeah, I could… I mean, I could go through the code, but there’s not a lot of…
54 00:06:41.580 ⇒ 00:06:47.210 Annie Yu: My tickets that has, like, programming stuff, …
55 00:06:47.850 ⇒ 00:06:50.600 Annie Yu: So, yeah. And I don’t think he’s…
56 00:06:51.190 ⇒ 00:06:55.410 Annie Yu: Knowledgeable around, like, the analysis side of things.
57 00:06:55.830 ⇒ 00:06:59.780 Robert Tseng: Yeah. Okay, I mean, that’s fair. I think, like I said, I…
58 00:06:59.950 ⇒ 00:07:12.959 Robert Tseng: I don’t know, I feel like our solution to this has always been, like, oh, go find other friends that can go and help people, and we’ve been outsourcing our coaching, I guess, but now I’m kind of of the belief that
59 00:07:12.970 ⇒ 00:07:27.880 Robert Tseng: we should just spend some more of our own time to, like, work with people on stuff that… we know our business better than anybody, so, like, better to just spend our own time, if we can. And so, I think for me, in the next two weeks, I’m…
60 00:07:28.120 ⇒ 00:07:41.459 Robert Tseng: I’m, gonna be out of town in 2 weeks, so I think for us, we could have… it’s really… this is gonna be, like, 4 sessions, I think, before I step out for a week. So hopefully, within these 2 weeks, we’ll be able to
61 00:07:41.580 ⇒ 00:07:51.440 Robert Tseng: See… yeah, we’ll be able to change the narrative here, and you’ll feel more… more confident to go and jump into some of these other projects.
62 00:07:52.190 ⇒ 00:07:53.440 Annie Yu: Yeah, okay.
63 00:07:53.440 ⇒ 00:07:53.930 Robert Tseng: Okay.
64 00:07:53.930 ⇒ 00:08:00.630 Annie Yu: That, does that mean after each session, there would be, like, a… Clear, kind of.
65 00:08:00.760 ⇒ 00:08:02.669 Annie Yu: Task for me to, kind of.
66 00:08:02.890 ⇒ 00:08:07.309 Annie Yu: Get my hands on, or would that be more, like, during the session?
67 00:08:07.750 ⇒ 00:08:20.970 Robert Tseng: Yeah, I mean, I think during the one, I’ll just kind of go through whatever’s outstanding. Like, I don’t… like, I… like, honestly, I haven’t really had too much time to come into this, like, super prepared. I think the goal is to give you something actionable after every session.
68 00:08:21.020 ⇒ 00:08:40.899 Robert Tseng: Like, maybe you’ll have a question that we’ll be able to… I’ll be able to react to whatever you’re getting hung up on and, like, kind of create something that… or at least direct you on what could be a good, next thing that you could… you could, work on. But, yeah, I… I know… I don’t have, like, a curriculum for you or anything. Like, I… I think this is really just…
69 00:08:41.130 ⇒ 00:08:45.350 Robert Tseng: We’ll just spend some time together and try to figure it out.
70 00:08:45.950 ⇒ 00:08:46.680 Annie Yu: Okay.
71 00:08:46.850 ⇒ 00:08:47.420 Robert Tseng: Okay.
72 00:08:48.160 ⇒ 00:08:54.570 Robert Tseng: Cool, so I’ll share my screen. I have a couple things pulled up here, … Let’s see…
73 00:08:55.780 ⇒ 00:09:05.700 Robert Tseng: I’m just gonna use a spark plug example. I, I… whatever Amber… I don’t know what message Amber sent you, but you could ignore that for now. We’ll just assume that we’re just kind of talking about this one.
74 00:09:05.830 ⇒ 00:09:10.340 Robert Tseng: So, … Let’s see…
75 00:09:14.930 ⇒ 00:09:19.490 Robert Tseng: Okay, and this is reported, so everything’s all good.
76 00:09:20.350 ⇒ 00:09:32.580 Robert Tseng: Cool. So, on this question, so kind of going back, kind of recapping a little bit, I mean, we’ll just cut through some of the fluff here. I think the ask here was just that,
77 00:09:34.280 ⇒ 00:09:42.800 Robert Tseng: you know, we’re trying to better understand user engagement, right? I don’t know if you’ve gotten a chance to go into the products, but if you go into the product, …
78 00:09:43.630 ⇒ 00:09:45.840 Robert Tseng: Spark plug…
79 00:09:50.550 ⇒ 00:09:58.730 Robert Tseng: I… don’t recall… Let’s see… Hmm.
80 00:10:05.470 ⇒ 00:10:06.789 Robert Tseng: There we go.
81 00:10:08.380 ⇒ 00:10:09.930 Robert Tseng: And…
82 00:10:13.080 ⇒ 00:10:16.769 Robert Tseng: Cool. So, yeah, this is, like, like.
83 00:10:18.060 ⇒ 00:10:31.249 Robert Tseng: I don’t know how much you understood for this. If you… if you already knew this, just let me know. Okay, so this is the demo environment. This is from the retailer perspective. And so, Sparkplug is a three-sided marketplace. You have your brands. Brands are, like.
84 00:10:31.380 ⇒ 00:10:34.220 Robert Tseng: … like, the
85 00:10:34.840 ⇒ 00:10:48.430 Robert Tseng: the companies that own the products, right? And, the retailers, you’ll notice that most of the wheat retailers are wheat shops. Like, this is… most… their entire business is really built for wheat shops, so….
86 00:10:48.430 ⇒ 00:10:50.780 Annie Yu: For what? You said for….
87 00:10:50.780 ⇒ 00:10:51.960 Robert Tseng: Marijuana.
88 00:10:52.200 ⇒ 00:10:53.379 Annie Yu: Oh, okay.
89 00:10:53.380 ⇒ 00:10:54.370 Robert Tseng: Yeah, weed.
90 00:10:54.790 ⇒ 00:10:55.400 Annie Yu: Okay.
91 00:10:55.870 ⇒ 00:11:15.239 Robert Tseng: Yeah, so… so you have your weed shops that are the retailers, and then you have, like, your weed products that are the brands, and then you have employees. Employees are kind of the users of the… of the app, or the main users of the app. So that’s kind of what is really being tracked here in this panel.
92 00:11:15.810 ⇒ 00:11:19.850 Robert Tseng: And, … Yeah, I guess, like.
93 00:11:20.210 ⇒ 00:11:28.259 Robert Tseng: what I had investigated. I think the open-ended question was, okay, Back in December.
94 00:11:29.120 ⇒ 00:11:34.179 Robert Tseng: why did users drop off? Like, why was there a dip, right? And so.
95 00:11:34.230 ⇒ 00:11:53.159 Robert Tseng: we started pretty high level. I think there wasn’t… I wasn’t really giving much direction there, it was just kind of go in… go into the product and figure out, like, why… why there was a dip. And so I think that’s a really common starting point. I think with, product analytics, it may feel like there’s so many different options, because
96 00:11:53.160 ⇒ 00:12:05.429 Robert Tseng: You can, you know, you’re basically trying to, like, pick a needle in the haystack along the customer journey, and then, identify, like, a causal relationship, or…
97 00:12:05.450 ⇒ 00:12:19.729 Robert Tseng: some sort of, like, explanatory relationship with a particular, you know, trend that you’re seeing. And that is quite difficult if you just view it from just this, like, wide open view.
98 00:12:20.150 ⇒ 00:12:26.940 Robert Tseng: … I think my… I have a couple perspectives on this. One is that, a lot of…
99 00:12:26.960 ⇒ 00:12:41.220 Robert Tseng: the… a lot of long-term behavior that you care about, or the long-term outcomes that you care about. It’s obviously lifetime value, like, high engagement, you know, all of your best users, or quote-unquote best users.
100 00:12:41.220 ⇒ 00:12:52.180 Robert Tseng: there’s a lot of early signal that you can get from, like, the first stages of engagement, which is why I think onboarding is so important, especially for software companies and just app companies.
101 00:12:52.190 ⇒ 00:12:58.859 Robert Tseng: So, I think that, to me, already narrows down my focus. Okay, I get this open-ended question on why is engagement down?
102 00:12:58.860 ⇒ 00:13:17.730 Robert Tseng: Well, I’m probably gonna find my answer, well, I’ll look at all activity generally, but then I’m gonna go and figure out, what, like, the early… early signals are from the… from, from, from a user’s engagement. So, yeah, if you have any questions, you can just cut me off anytime, but otherwise, I’ll just kind of…
103 00:13:17.910 ⇒ 00:13:19.420 Robert Tseng: Kind of frame it that way.
104 00:13:19.580 ⇒ 00:13:30.260 Annie Yu: Yeah, so wait, so now we are seeing a drop-off in December, and we are gonna look for that early signal, so does that mean we’re looking for things before that?
105 00:13:30.870 ⇒ 00:13:44.049 Robert Tseng: Yeah, well, so I would… I would… I would suspect I would probably go and look at customer behavior by cohort, like, in, you know, earlier, probably October, November, looked pretty stable, and then, you know, I got this request, like, back in, like, March or something, or…
106 00:13:44.100 ⇒ 00:14:03.219 Robert Tseng: And so, I would look at January, February, March, or something, and just kind of get a sense of that. But, you know, it turns out, like, I already answered this question. December, like, it’s not… it’s not really from the user… it’s… tracking broke down. That was the… that was the conclusion I gave them. I… I looked into it a bit more, I saw, like, the volume of… of,
107 00:14:03.490 ⇒ 00:14:06.189 Robert Tseng: Like, if we go into this pre…
108 00:14:06.370 ⇒ 00:14:20.859 Robert Tseng: engagement funnel analysis, pre- and post-vite update. There’s this particular update that they had sent. Like, I had to ask some questions, like, was there anything that was pushed in December or whatever? Turns out, hey, okay, they made this, like, big instrumentation, like, change. And, …
109 00:14:21.320 ⇒ 00:14:40.840 Robert Tseng: yeah, like, you know, I kind of investigated, okay, well, there’s a 20% drop-off, you know, we have this pre-pre-vite period, post-vite period, some hypotheses that I had, and things that I wanted to go investigate. So I started looking at, kind of, like, what is that ideal, what’s…
110 00:14:41.030 ⇒ 00:14:56.040 Robert Tseng: what’s the ideal, like, customer activation point? Well, it’s like, they log into the app, and they complete a spark. I think that’s just kind of intuitively what I had figured out. And, I, I obviously ran….
111 00:14:56.350 ⇒ 00:14:58.040 Annie Yu: That’s for an employee.
112 00:14:58.040 ⇒ 00:15:11.310 Robert Tseng: Yes, this is for an employee. Everything we care about is only about employee engagement. Because if I think about it, there aren’t that many retailers. I don’t exactly know how many users there are, but I could always just kind of do… I could just do a quick, quick search, …
113 00:15:11.620 ⇒ 00:15:14.290 Robert Tseng: FIGO and, …
114 00:15:14.650 ⇒ 00:15:23.509 Robert Tseng: If I just looked at, any activity, … well, I already kind of created this custom event, which is another mechanic of
115 00:15:23.670 ⇒ 00:15:27.949 Robert Tseng: Of, mixed panel, but we can just kind of skip that. Maybe I’ll just go to events.
116 00:15:28.270 ⇒ 00:15:37.219 Robert Tseng: They have a bunch of events here already. I would just go, all events. Okay, how do I break this down? I want to break this down by role.
117 00:15:37.420 ⇒ 00:15:41.580 Robert Tseng: … Yeah, so… None.
118 00:15:42.300 ⇒ 00:16:00.689 Robert Tseng: not set, retailer, brand, super admin. So, none, turns out, is employees. You wouldn’t know that unless you kind of just assumed, and, like, you asked some questions. Yeah, turns out this role just means it’s employee. So it’s like, yeah, obviously, you know, …
119 00:16:00.900 ⇒ 00:16:05.569 Robert Tseng: something like 90% of the traffic is from employees. And so.
120 00:16:05.650 ⇒ 00:16:21.540 Robert Tseng: I don’t really care too much. I don’t understand the dynamics between the different types of users, but I’m just gonna go after the biggest one. So I’m just gonna uncheck all of them and kind of take a mental note that I’m only going to be looking at employee engagement. So, that’s kind of why I narrowed it already. Before I even started doing anything.
121 00:16:21.540 ⇒ 00:16:29.610 Robert Tseng: one of the different types of users on the platform. Turns out employees drive 90% of the event volume. That’s the one I’m gonna go look at first.
122 00:16:30.080 ⇒ 00:16:35.029 Robert Tseng: So I can go, and I’m looking at employee user engagements, I look at, kind of, yeah.
123 00:16:35.160 ⇒ 00:16:50.110 Robert Tseng: So, from an employee perspective, what does the user journey look like? Well, they can log in, they can complete a Spark. There’s a mobile app to this as well. I don’t really think they can really get to that from here, but… great. Employee view.
124 00:16:50.940 ⇒ 00:16:58.079 Robert Tseng: Yeah, this… you know, I didn’t download the app, but I just kind of got a sense. This is kind of how the Spark works for an employee.
125 00:16:58.310 ⇒ 00:16:59.420 Robert Tseng: …
126 00:16:59.870 ⇒ 00:17:09.249 Robert Tseng: Yeah, they get a text, so I do get texts, and they… I have to go and log in, and so I pull up the app, there’s a couple things I do, maybe there’s a tile I click on.
127 00:17:09.430 ⇒ 00:17:13.859 Robert Tseng: And then some course or something that’s just, like.
128 00:17:16.250 ⇒ 00:17:19.099 Robert Tseng: Yeah, that the… that the, that the retailer added.
129 00:17:19.750 ⇒ 00:17:20.530 Robert Tseng: …
130 00:17:23.250 ⇒ 00:17:36.089 Robert Tseng: So they come in, they interact with whatever the SPARC is, which is just like a mini course, or it’s a mini interactive module. They complete it, and then they get some sort of cash reward, that they can claim.
131 00:17:36.500 ⇒ 00:17:47.750 Robert Tseng: Later on. So, later on, they can claim the reward, yeah, and that becomes, you know, a cash-out thing where they can go and cash out this balance. So, it’s like…
132 00:17:47.880 ⇒ 00:17:54.800 Robert Tseng: Okay, they go on, they, they interact with… it’s some, like, habit… it’s basically, to me, like a habit-tracking app.
133 00:17:54.990 ⇒ 00:18:12.870 Robert Tseng: that’s somewhat guided. You complete it, you get a reward, and then you basically have, like, a debit balance that’s kind of like Venmo, and then you can cash it out. So, that’s kind of, like, how I’ve mapped out the user journey in my head. So, I care about the first spark completion.
134 00:18:12.970 ⇒ 00:18:24.689 Robert Tseng: And I looked at some different things, like the conversion percentage, time to conversion, you know, if there’s any… anything weird about that, you know, and, you know, all these different things.
135 00:18:25.360 ⇒ 00:18:34.789 Robert Tseng: Anyway, like, it gave me to… it brought me to my conclusion that this was, like, probably not user error, like, nothing in here looks that… looks that weird. So, …
136 00:18:35.150 ⇒ 00:18:42.280 Robert Tseng: Yeah, and if anything, I got to see some differences between users before and after the change.
137 00:18:42.280 ⇒ 00:18:47.130 Annie Yu: Robert, can we pause for a bit? So, over the, …
138 00:18:47.720 ⇒ 00:18:54.849 Annie Yu: The first part, those line charts, so each line represents Kind of a stage.
139 00:18:58.390 ⇒ 00:19:00.060 Robert Tseng: Yeah, so let’s kind of click into that.
140 00:19:00.060 ⇒ 00:19:01.360 Annie Yu: Yeah.
141 00:19:02.030 ⇒ 00:19:11.280 Robert Tseng: Yeah, so in here, you have insights, I guess this is really a funnel report. So, Insights is kind of the main, like, like, reporting metric thing.
142 00:19:11.320 ⇒ 00:19:27.470 Robert Tseng: And you can… within it, you can create different types of reports here. So, simple as just, like, plotting some sort of metric over time, or funnel is just, you know, sequence from step A to step B. Retention is… I think you know retention pretty well, so I won’t say that.
143 00:19:27.550 ⇒ 00:19:42.390 Robert Tseng: Yeah, so I’m only really looking at one, one measure. And, like, if I were to go back and clean this up, I could have just, like, layered multiple things in. So maybe I do… okay, let’s look at, page view, or, like, Spark completed.
144 00:19:42.520 ⇒ 00:19:46.719 Robert Tseng: to… or Spark… Or what is this? …
145 00:19:49.800 ⇒ 00:19:58.760 Robert Tseng: sparked details. … I don’t know if that’s exactly what I want.
146 00:20:00.020 ⇒ 00:20:06.570 Robert Tseng: Great Spark, da-da-da-da-da-da. Employee… …
147 00:20:10.440 ⇒ 00:20:19.730 Robert Tseng: Yeah, let’s just say, … Employee Active Spark to, … hash out.
148 00:20:26.040 ⇒ 00:20:27.500 Robert Tseng: …
149 00:20:30.530 ⇒ 00:20:33.689 Robert Tseng: Wait, this is actually supposed to be a funnel. So, ….
150 00:20:33.690 ⇒ 00:20:43.270 Annie Yu: How do you decide to see the button versus page view? Because when they click on a button, it goes to the next page view, right?
151 00:20:44.080 ⇒ 00:20:53.650 Robert Tseng: Yeah, I mean, I think there’s probably some… like, I… I would, … within the app.
152 00:20:54.260 ⇒ 00:21:02.420 Robert Tseng: I guess, if I didn’t have the app, then I could just go and I could pick, like, a user that’s kind of going through the sequence, so… …
153 00:21:03.670 ⇒ 00:21:08.230 Robert Tseng: I don’t know, like, … Let’s see…
154 00:21:22.360 ⇒ 00:21:30.729 Robert Tseng: like, ideally, you would have the app, open, and you can track it, which I don’t have it in this view, like, I would have to…
155 00:21:31.000 ⇒ 00:21:32.190 Robert Tseng: ….
156 00:21:34.480 ⇒ 00:21:40.450 Annie Yu: Also, it would show, like, a page, and then tell you… We track this page view.
157 00:21:40.450 ⇒ 00:21:56.240 Robert Tseng: Yeah… like… If I went into this… And… I… … Mix panel.
158 00:21:58.580 ⇒ 00:22:06.360 Robert Tseng: Expollo Extension… hone… …
159 00:22:06.960 ⇒ 00:22:11.969 Robert Tseng: I might have to get, like, a debugger tool, … I mean, I guess you could just…
160 00:22:12.200 ⇒ 00:22:13.629 Robert Tseng: I could just look at the…
161 00:22:13.740 ⇒ 00:22:16.210 Robert Tseng: I could just inspect and look at the console.
162 00:22:16.460 ⇒ 00:22:17.790 Robert Tseng: See what’s firing.
163 00:22:18.800 ⇒ 00:22:19.740 Robert Tseng: …
164 00:22:20.710 ⇒ 00:22:29.460 Robert Tseng: Oh, your one amplitude, like, it’s… you could easily see, as you’re interacting with an application, you will see all the different things that are being tracked at every step.
165 00:22:29.590 ⇒ 00:22:34.749 Robert Tseng: There’s probably some debugger tool that I could go and get from XPanel.
166 00:22:35.120 ⇒ 00:22:39.640 Robert Tseng: I didn’t build my initial reports on this device, so I don’t have it all set up.
167 00:22:40.250 ⇒ 00:22:46.109 Robert Tseng: But another way I could get around this is, if I go here, and I go to events, and…
168 00:22:46.950 ⇒ 00:22:49.270 Robert Tseng: I am…
169 00:22:50.150 ⇒ 00:22:57.199 Robert Tseng: I basically need to find myself, which I think is a little bit difficult to do. Maybe there is, like, a…
170 00:22:57.440 ⇒ 00:23:02.140 Robert Tseng: Robert… Oh, there we go, there’s me.
171 00:23:02.310 ⇒ 00:23:02.670 Annie Yu: Yeah.
172 00:23:04.860 ⇒ 00:23:14.110 Robert Tseng: Okay, cool. So, yeah, I think, I get to see a little sequence here. So, as I’m clicking around, dashboard, retailer, da-da-da-da, okay.
173 00:23:14.650 ⇒ 00:23:15.720 Robert Tseng: …
174 00:23:16.640 ⇒ 00:23:21.459 Robert Tseng: Yeah, on the Spark side, may not be able to see that, like, I might have to get, like, the…
175 00:23:21.620 ⇒ 00:23:26.350 Robert Tseng: … Spark plug….
176 00:23:26.990 ⇒ 00:23:29.669 Annie Yu: So this is not the Sparks table.
177 00:23:31.110 ⇒ 00:23:36.710 Robert Tseng: This is… Not a spark, yeah.
178 00:23:37.140 ⇒ 00:23:48.060 Robert Tseng: … I don’t even… yeah, I don’t think I have the Spark app, so… Spark, demo…
179 00:23:48.170 ⇒ 00:23:50.140 Robert Tseng: Now I’m in the demo app.
180 00:23:55.200 ⇒ 00:24:04.989 Robert Tseng: … Can I… How do I show my device on the screen?
181 00:24:07.590 ⇒ 00:24:10.010 Robert Tseng: Advice… …
182 00:24:20.680 ⇒ 00:24:22.240 Robert Tseng: Okay…
183 00:24:32.040 ⇒ 00:24:35.100 Robert Tseng: Okay, there we go. So…
184 00:24:35.540 ⇒ 00:24:42.260 Robert Tseng: I got the text, I’m on here now, and I’m gonna go and… click on this thing, I’m…
185 00:24:42.590 ⇒ 00:24:44.400 Robert Tseng: I don’t know.
186 00:24:50.000 ⇒ 00:24:53.750 Annie Yu: So you’re technically, an employee?
187 00:24:54.470 ⇒ 00:24:55.040 Annie Yu: your account.
188 00:24:55.040 ⇒ 00:24:55.560 Robert Tseng: Hi.
189 00:24:55.560 ⇒ 00:24:56.940 Annie Yu: Ascendant.
190 00:24:56.940 ⇒ 00:25:00.439 Robert Tseng: Yes, I’m technically an employee, correct.
191 00:25:01.660 ⇒ 00:25:06.650 Robert Tseng: So, active sports table. Alright, well, finally, fired. So…
192 00:25:06.830 ⇒ 00:25:11.979 Robert Tseng: I guess that’s what this event is. Like, it is kind of just, like, I have to… I’m figuring out, like, what…
193 00:25:12.230 ⇒ 00:25:22.539 Robert Tseng: what is being tracked, because I didn’t… I didn’t set up the tracking myself, so, yeah, I kind of have to familiarize myself with the product and figure out how everything is… is… is being tracked.
194 00:25:22.540 ⇒ 00:25:23.750 Annie Yu: ….
195 00:25:25.630 ⇒ 00:25:33.050 Robert Tseng: Yeah, it’s not a great situation, but, like, it’s doable. Like, we can make enough assumptions to figure out, like, how this works, so…
196 00:25:33.680 ⇒ 00:25:40.619 Robert Tseng: … There are no… I guess I had already completed a SPARC before.
197 00:25:40.990 ⇒ 00:25:44.240 Robert Tseng: Is there another one I could use? …
198 00:25:50.730 ⇒ 00:25:52.110 Robert Tseng: Yes, yes.
199 00:26:15.190 ⇒ 00:26:20.990 Robert Tseng: Something feels off. Maybe this… maybe it got disconnected to a different person.
200 00:26:22.220 ⇒ 00:26:23.340 Robert Tseng: Correct.
201 00:26:29.270 ⇒ 00:26:31.150 Robert Tseng: Who is this person?
202 00:26:32.370 ⇒ 00:26:34.230 Robert Tseng: Britt.
203 00:26:35.630 ⇒ 00:26:37.609 Robert Tseng: Nope, doesn’t exist.
204 00:26:37.610 ⇒ 00:26:38.730 Annie Yu: Oh, great.
205 00:26:39.290 ⇒ 00:26:40.190 Robert Tseng: ….
206 00:26:44.930 ⇒ 00:26:47.759 Annie Yu: Does that mean we found a problem?
207 00:26:48.090 ⇒ 00:26:53.879 Robert Tseng: Yeah, I… it’s weird that things… things don’t connect. I…
208 00:27:01.810 ⇒ 00:27:06.859 Robert Tseng: Okay, I’m kind of running into a dead end on the employee engagement side. I think, …
209 00:27:07.830 ⇒ 00:27:18.140 Robert Tseng: yeah, I can’t trace it back to this person specifically. I don’t remember how I did it. It’s been, like, over a month, or… it’s been, like, almost 2 months, so, like, I have no context. I would… it would probably take me, like.
210 00:27:18.610 ⇒ 00:27:22.080 Robert Tseng: 30 minutes just to figure that out, so… …
211 00:27:23.140 ⇒ 00:27:31.130 Robert Tseng: I guess what I can say is… yeah, if we just stayed… if we just stayed high level, and we tried to, like, drill down into it.
212 00:27:31.710 ⇒ 00:27:48.640 Robert Tseng: I guess that’s why I have to do, like, a top-down approach to this analysis. I have to start really, … I have to start very high, and then I may not know all the nuances and the sequence of the events, but, like, at least I can structure something that makes sense, and I can put it in front of the client.
213 00:27:48.640 ⇒ 00:27:54.510 Robert Tseng: If I’m wrong, they can let me know, but they can’t really disagree with this. This is just, like, activity by customer.
214 00:27:54.510 ⇒ 00:27:58.320 Robert Tseng: It’s not really, like, a full funnel or anything.
215 00:27:58.470 ⇒ 00:28:10.660 Robert Tseng: And so I’m just monitoring certain trends, like, I… I kind of, like, figure it out, okay, this is their… this is their highest retailer, this is where jars, this is where all the tra… a lot of the traffic is coming from.
216 00:28:10.770 ⇒ 00:28:16.540 Robert Tseng: I know what the completion rates are in terms of, like, the wallet usage and everything, and… …
217 00:28:16.890 ⇒ 00:28:21.280 Robert Tseng: Anyway, so it was a kind of a smokescreen. I think there… there’s that piece to it.
218 00:28:21.850 ⇒ 00:28:22.750 Robert Tseng: …
219 00:28:23.120 ⇒ 00:28:36.899 Robert Tseng: And what I… what I had built when I sent it to you, I believe, is this employee engagement, or user engagement overview. I kind of, like, borrowed some of the same things and got down to,
220 00:28:37.820 ⇒ 00:28:40.860 Robert Tseng: I was… figuring out…
221 00:28:42.350 ⇒ 00:28:59.970 Robert Tseng: what can I actually segment by? Like, what if I… what if I segment it by brand, by product, by… again, I’m still staying high level, I’m just trying to create different cohorts so that I can view the same trends, but, like, I’m breaking it down into different
222 00:28:59.970 ⇒ 00:29:04.249 Robert Tseng: groups that I can see if there is something interesting that sticks out.
223 00:29:04.390 ⇒ 00:29:10.550 Robert Tseng: … I think my conclusion when I did that was just that,
224 00:29:15.770 ⇒ 00:29:18.379 Robert Tseng: Whatever I had here. I…
225 00:29:19.220 ⇒ 00:29:33.449 Robert Tseng: engagement was not flat across the network. There are… it looks different across different retailers. There are certain retailers where engagement is way higher, and so maybe we look at, like, okay, for this retailer, why does engagement dip? It looks like there is, like, a…
226 00:29:34.400 ⇒ 00:29:40.849 Robert Tseng: bi-weekly seasonality to engagements. And that’s, like, something that I just need to call out.
227 00:29:41.010 ⇒ 00:29:42.720 Robert Tseng: Which makes sense, it’s like…
228 00:29:42.920 ⇒ 00:30:01.739 Robert Tseng: The… an employee… my hypothesis would be, okay, well, an employee is only going to be using this app and completing Sparks twice a month, because they get paid on a bi-weekly basis. And so, that’s probably why they’re, like, kind of going, you know, they’re doing this bi-weekly thing.
229 00:30:01.790 ⇒ 00:30:10.629 Robert Tseng: Or why they’re showing this bi-weekly behavior. Like, nobody told me that, but, like, I just inferred that, and, like, I called it out somewhere here. So I’m kind of…
230 00:30:11.070 ⇒ 00:30:28.870 Robert Tseng: tried to draw some inferences about the engagement and the seasonality of it. I talked about retailer-specific behavior curves and, like, how certain retailers are staying, and so if I were to keep running with a retailer, I would pick, like, one… I would just…
231 00:30:28.870 ⇒ 00:30:42.610 Robert Tseng: do, like, a retail analysis. I would only be looking at JARS Cannabis, that would help me break this down. Maybe I look at some of the, like, and I would look at some other lower, lower performing retailers as well. So, …
232 00:30:42.910 ⇒ 00:30:49.289 Robert Tseng: maybe it’s too noisy to go for anybody that’s that small, but maybe I say, okay, well.
233 00:30:49.390 ⇒ 00:31:06.510 Robert Tseng: Anybody that’s above 5,000 events, per… per day, I’m gonna look… or per week, I’m gonna look at them as, like, one category. I’m gonna set another category at 3,000, and then maybe, like, under 1,000 or something. And those are 3 buckets that I can go and look at and do more of a retail-specific analysis.
234 00:31:06.510 ⇒ 00:31:09.240 Annie Yu: Maybe you can set up with Mixpanel.
235 00:31:09.710 ⇒ 00:31:10.910 Annie Yu: Is that it?
236 00:31:11.000 ⇒ 00:31:12.380 Robert Tseng: Sorry?
237 00:31:12.380 ⇒ 00:31:16.110 Annie Yu: You said, like, based on the average, we segment them
238 00:31:16.470 ⇒ 00:31:20.819 Annie Yu: Is that something you can set up in Mixpanel, or you just….
239 00:31:20.820 ⇒ 00:31:27.660 Robert Tseng: Yeah, that’s something you can set up a mixed panel. So let’s say I, save this as new…
240 00:31:29.430 ⇒ 00:31:41.800 Robert Tseng: and… Yeah, so what I’m gonna do is, like, okay, don’t input active sparks, … Let’s break down by…
241 00:31:42.280 ⇒ 00:31:43.920 Robert Tseng: …
242 00:31:49.790 ⇒ 00:32:01.170 Robert Tseng: not groups… Let’s break down by… … So the name was…
243 00:32:05.600 ⇒ 00:32:10.540 Robert Tseng: I guess… Total events was, …
244 00:32:20.180 ⇒ 00:32:21.480 Robert Tseng: I set this up.
245 00:32:27.060 ⇒ 00:32:29.889 Annie Yu: And I have one more question here, so…
246 00:32:30.360 ⇒ 00:32:37.950 Annie Yu: Like, your example hypothesis, like, employees use it only twice a month, because they get pra- twice.
247 00:32:38.060 ⇒ 00:32:45.590 Annie Yu: They get paid twice a month. Is that… is that, like, a final takeaway, or that’s something we would expect to…
248 00:32:45.940 ⇒ 00:32:47.390 Annie Yu: Verify.
249 00:32:49.200 ⇒ 00:32:56.890 Robert Tseng: Yeah, I mean, I think that’s something I can call out, is, like, I think this is what’s happening, and then if that… if we’re wrong, they can tell us, like, oh…
250 00:32:57.050 ⇒ 00:32:58.810 Robert Tseng: like, they’re….
251 00:32:59.200 ⇒ 00:33:04.149 Annie Yu: there’s another reason for why it’s bi-weekly, but, like, I think that the point is just to, like, kind of….
252 00:33:04.150 ⇒ 00:33:05.150 Robert Tseng: at least…
253 00:33:06.170 ⇒ 00:33:17.109 Robert Tseng: they may not be looking at the data this way. They may not be… they’re probably not looking at seasonality, they’re not looking at any of these things, because it’s… this is probably way too much for your… for the, like.
254 00:33:17.220 ⇒ 00:33:28.300 Robert Tseng: their head of product is not… is not digging this deep into things. But anyway, I’m trying to show you the mechanic of, like, how to kind of set these groups by… by, …
255 00:33:28.910 ⇒ 00:33:35.289 Robert Tseng: our quantity, I think we have to be a bit creative, like, I don’t fully know off the top of my head, I’d probably use GPT to help me a little bit.
256 00:33:35.420 ⇒ 00:33:37.110 Robert Tseng: I think….
257 00:33:39.050 ⇒ 00:33:40.190 Annie Yu: So that’s doable.
258 00:33:40.190 ⇒ 00:33:42.640 Robert Tseng: It’s… it’s doable, yeah. …
259 00:33:45.800 ⇒ 00:33:50.169 Robert Tseng: I think an easier example would just…
260 00:33:54.530 ⇒ 00:34:04.420 Robert Tseng: I think you can do computed traits. We can see… user… not user… …
261 00:34:09.320 ⇒ 00:34:14.240 Robert Tseng: What is this? Weekly event volume, …
262 00:34:16.320 ⇒ 00:34:19.139 Robert Tseng: How do I break this down more?
263 00:34:29.340 ⇒ 00:34:43.720 Robert Tseng: Yeah, I… I’m not entirely sure off the top of my head, so I think what I would probably do instead… I mean, I just want to be able to pick… pick something tangible to go and look at on this call. Like, I think we’ve done a lot of talking, so I’m trying to, like, actually
264 00:34:43.929 ⇒ 00:34:49.909 Robert Tseng: pick something as, like, a next step. So… ….
265 00:34:51.139 ⇒ 00:34:56.239 Annie Yu: And there’s a filter to just drill down to one group.
266 00:34:57.460 ⇒ 00:35:05.940 Robert Tseng: Yeah, so filtering is just… is what you’d expect. This is just, like, a WHERE clause in a SQL. And then breakdown is just, like, a group by, pretty much.
267 00:35:07.980 ⇒ 00:35:08.670 Annie Yu: Okay.
268 00:35:09.160 ⇒ 00:35:12.069 Robert Tseng: So, let’s say, retail…
269 00:35:23.290 ⇒ 00:35:31.060 Robert Tseng: I mean, what helps me is, like, I just gotta go and out… I have to outline out what I’m… exactly what I’m gonna do, and then I can find my… I can try to figure out my way there. So…
270 00:35:31.200 ⇒ 00:35:34.509 Robert Tseng: I think it is hard to just jump into something and then, like.
271 00:35:35.180 ⇒ 00:35:38.619 Robert Tseng: Here, and then get to that, so…
272 00:35:39.080 ⇒ 00:35:44.000 Robert Tseng: What am I looking for specifically? Do-doo…
273 00:35:47.430 ⇒ 00:35:50.550 Robert Tseng: So, I’m trying to evaluate
274 00:36:07.540 ⇒ 00:36:10.560 Robert Tseng: And what are my known… known constraints?
275 00:36:12.620 ⇒ 00:36:19.060 Robert Tseng: I want to… does… …
276 00:36:22.560 ⇒ 00:36:33.060 Robert Tseng: How does user or, you know, employee engagement differ… are across… retailers.
277 00:36:33.180 ⇒ 00:36:38.040 Robert Tseng: … Look at highest engaged retailer.
278 00:36:38.220 ⇒ 00:36:39.820 Robert Tseng: Yes, Lois.
279 00:36:40.260 ⇒ 00:36:41.100 Robert Tseng: Great.
280 00:36:41.440 ⇒ 00:36:44.960 Robert Tseng: There must be, … Not quick.
281 00:36:49.210 ⇒ 00:36:51.489 Robert Tseng: I don’t like the formatting there.
282 00:36:55.380 ⇒ 00:37:10.909 Robert Tseng: like, I would expect, like, … maybe… I disengaged for… more engaged retailers, Have more products.
283 00:37:15.310 ⇒ 00:37:28.750 Robert Tseng: mortgage, retailers, Have more, or, are… morgue, or… Creating more sparks consistently.
284 00:37:33.210 ⇒ 00:37:37.679 Robert Tseng: offer Better rewards…
285 00:37:38.120 ⇒ 00:37:48.729 Robert Tseng: like, what are the different levers that a retailer can pull on to influence employee behavior? I think those are probably three that I would look at.
286 00:37:49.110 ⇒ 00:37:53.110 Annie Yu: … Yeah.
287 00:37:55.050 ⇒ 00:38:03.409 Annie Yu: The second and third, Makes sense to me. And for the first one, Is that something we…
288 00:38:03.810 ⇒ 00:38:07.030 Annie Yu: Have the ability to know.
289 00:38:08.230 ⇒ 00:38:11.559 Robert Tseng: I’m not sure. I think… I don’t… I guess I… I don’t know…
290 00:38:11.710 ⇒ 00:38:15.220 Robert Tseng: A lot of these… the answers to these questions, like, I…
291 00:38:16.130 ⇒ 00:38:27.730 Robert Tseng: I would… I would try to if I… if I… if I… if I try to look and I don’t see it, then I’m just gonna call it out, like, I don’t think I’m able to go and buy these products or whatever. So I think that’s kind of where no constraint, where it’s like.
292 00:38:27.940 ⇒ 00:38:29.640 Robert Tseng: Maybe we don’t…
293 00:38:30.570 ⇒ 00:38:39.939 Robert Tseng: we don’t know, like, we don’t… we don’t have a way of accurately segmenting, like, by product. I’m not sure about that.
294 00:38:40.140 ⇒ 00:38:41.430 Annie Yu: ….
295 00:38:45.330 ⇒ 00:38:46.090 Robert Tseng: Yeah.
296 00:38:47.820 ⇒ 00:38:48.890 Robert Tseng: So…
297 00:38:49.690 ⇒ 00:39:07.420 Robert Tseng: I mean, I wouldn’t actually draft this out in MixPanel, because I think sometimes this is pretty buggy, but, like, yeah, I would literally just, like, have a doc, and I would just write it out, and then I would just… and then I’m gonna go in and look for stuff. So let’s just keep it constrained, we’ll go look for one. So, employee engagement differ across retailers. Okay, let’s just go and look at that. So…
298 00:39:07.780 ⇒ 00:39:13.819 Robert Tseng: … So… employee….
299 00:39:13.820 ⇒ 00:39:17.840 Annie Yu: Monthly, active users.
300 00:39:19.140 ⇒ 00:39:24.790 Robert Tseng: Yeah, this is just monthly active users, but I’m… I can… I can go and edit it, so I’m just gonna break it out by…
301 00:39:25.100 ⇒ 00:39:29.360 Robert Tseng: Whoa… …
302 00:39:32.750 ⇒ 00:39:39.129 Robert Tseng: Or by retailer name. How do I get the retailer name again? I don’t really remember.
303 00:39:39.750 ⇒ 00:39:47.649 Robert Tseng: … I had retailer admin, this cohort I created, …
304 00:39:50.580 ⇒ 00:39:57.140 Robert Tseng: I had to do some weird things. I had to go to… retail…
305 00:40:04.920 ⇒ 00:40:07.459 Robert Tseng: What did I do here?
306 00:40:08.880 ⇒ 00:40:12.490 Robert Tseng: Oh, look, I have this set up fine.
307 00:40:13.990 ⇒ 00:40:14.890 Robert Tseng: …
308 00:40:23.090 ⇒ 00:40:23.960 Robert Tseng: Okay.
309 00:40:24.420 ⇒ 00:40:27.950 Robert Tseng: Just gotta have to flip this around. It’s crazy messy.
310 00:40:30.510 ⇒ 00:40:40.090 Robert Tseng: Okay, so I had it broken down by retailer. So let’s just say… … Filter by name.
311 00:40:45.390 ⇒ 00:40:46.390 Robert Tseng: is…
312 00:40:52.630 ⇒ 00:40:57.799 Robert Tseng: So this is not typically how it should be set up, because, like, this… somebody basically, like.
313 00:40:58.000 ⇒ 00:41:03.669 Robert Tseng: put… use… it’s, like, in JSON format, so that’s why this is kind of messed up, so…
314 00:41:03.890 ⇒ 00:41:10.610 Robert Tseng: I understand that there are some constraints here, but, like, I don’t know, I guess maybe you wouldn’t know that until you clicked into it. So…
315 00:41:10.970 ⇒ 00:41:26.369 Robert Tseng: ideally, I would just be able to filter by retail main, right? Like, that would be intuitive. But that’s just not how it’s set up, and it’s kind of frustrating that we don’t have a… like, we don’t… I’m not able to do that so easily, so I’m going to have to just go and do something else.
316 00:41:26.630 ⇒ 00:41:27.410 Robert Tseng: ….
317 00:41:28.870 ⇒ 00:41:30.870 Annie Yu: Oh. Contains.
318 00:41:30.870 ⇒ 00:41:31.820 Robert Tseng: Jars.
319 00:41:31.820 ⇒ 00:41:33.989 Annie Yu: And uncheck those lists, right?
320 00:41:35.080 ⇒ 00:41:37.790 Robert Tseng: Yeah… But.
321 00:41:37.790 ⇒ 00:41:40.000 Annie Yu: Below the… below that chart?
322 00:41:42.770 ⇒ 00:41:49.149 Robert Tseng: I mean, I could, but it’s just, like, I have to go in and uncheck all of them, and ugh.
323 00:41:51.200 ⇒ 00:41:54.010 Robert Tseng: Yeah, sure, I could just do this. …
324 00:42:01.190 ⇒ 00:42:06.609 Robert Tseng: like, I want to group this as one, which is, like, I don’t really….
325 00:42:06.610 ⇒ 00:42:07.540 Annie Yu: arms.
326 00:42:09.620 ⇒ 00:42:12.569 Robert Tseng: Yeah, cause to me, this is one… this is actually one…
327 00:42:12.790 ⇒ 00:42:16.269 Robert Tseng: retailer. They have multiple locations, but…
328 00:42:17.740 ⇒ 00:42:26.340 Robert Tseng: I mean, I could just look at them separately. I could just look at Michigan separately, whatever. But, like, how do I create this as a filter? Like, it’s kind of, like, what I’m…
329 00:42:26.590 ⇒ 00:42:31.230 Robert Tseng: what I’m trying to get to. So… ….
330 00:43:23.320 ⇒ 00:43:25.419 Annie Yu: And how do you know only these theories?
331 00:43:25.630 ⇒ 00:43:28.710 Annie Yu: Our, history are retailers.
332 00:43:29.080 ⇒ 00:43:32.960 Annie Yu: Because you… You’ve worked on it, or….
333 00:43:33.680 ⇒ 00:43:38.199 Robert Tseng: No, no, they’re… There are more… there are more retailers. …
334 00:43:39.510 ⇒ 00:43:42.980 Robert Tseng: Okay, there we go. I got that as one….
335 00:43:46.010 ⇒ 00:43:49.170 Robert Tseng: And… I’m… I’m just trying to…
336 00:43:49.700 ⇒ 00:43:59.660 Robert Tseng: I’m just trying to pick… pick… I’m, like, I’m literally just cherry-picking. Like, I… I don’t… I don’t have a reason for, like, why I picked JARS, other than JARS is the… is the largest one, from what I saw in the data.
337 00:43:59.940 ⇒ 00:44:00.330 Annie Yu: Yeah.
338 00:44:00.330 ⇒ 00:44:09.390 Robert Tseng: … Okay, so… Great. That looks better.
339 00:44:09.940 ⇒ 00:44:16.290 Robert Tseng: So, I mean, to me, this is like, I wanna… Convert this into a….
340 00:44:16.560 ⇒ 00:44:20.530 Annie Yu: How did you make them all in one color? What’d you just do?
341 00:44:20.530 ⇒ 00:44:25.299 Robert Tseng: So I moved it from… instead of breakdown to filter, so I had to go group.
342 00:44:25.530 ⇒ 00:44:28.789 Annie Yu: Yeah. To name to the retailer name.
343 00:44:28.790 ⇒ 00:44:35.120 Robert Tseng: Right, so now this is, like, one retail group, and then I want to do, like, a comp against another retail group.
344 00:44:35.490 ⇒ 00:44:41.300 Robert Tseng: Which… I will just say is…
345 00:44:50.250 ⇒ 00:44:52.270 Robert Tseng: Let’s just compare it to…
346 00:44:55.500 ⇒ 00:44:57.499 Robert Tseng: Loom, I guess.
347 00:44:58.790 ⇒ 00:45:01.400 Robert Tseng: Bloom and Walgreens. So…
348 00:45:06.240 ⇒ 00:45:14.389 Robert Tseng: No, I’m just gonna keep it simple. I’ll just do room, and I’ll do… Well, gains…
349 00:45:16.480 ⇒ 00:45:19.700 Robert Tseng: And I’ll break it down by 3.
350 00:45:25.860 ⇒ 00:45:29.970 Robert Tseng: …
351 00:45:32.170 ⇒ 00:45:40.519 Robert Tseng: I’ll do a… for… I don’t want to do stacked, I want to just do columns.
352 00:45:42.050 ⇒ 00:45:43.060 Robert Tseng: Great.
353 00:45:43.650 ⇒ 00:45:46.870 Robert Tseng: That makes… looks a bit better for me.
354 00:45:53.610 ⇒ 00:45:58.110 Robert Tseng: Okay, so I’m just gonna… I’ll rename this later.
355 00:45:58.520 ⇒ 00:46:04.829 Robert Tseng: … Okay, that gives me something to work with.
356 00:46:05.840 ⇒ 00:46:09.269 Robert Tseng: So these are the three, kind of, groups that I wanted to look at.
357 00:46:10.110 ⇒ 00:46:14.629 Robert Tseng: … Do they have more products? Let’s see…
358 00:46:19.750 ⇒ 00:46:24.030 Robert Tseng: So, let’s go and figure out, how do I get them by product?
359 00:46:24.140 ⇒ 00:46:25.370 Robert Tseng: So…
360 00:46:29.310 ⇒ 00:46:31.629 Robert Tseng: How do I find products?
361 00:46:33.500 ⇒ 00:46:34.800 Robert Tseng: …
362 00:46:38.520 ⇒ 00:46:40.259 Robert Tseng: If I go back to this…
363 00:46:40.390 ⇒ 00:46:43.289 Robert Tseng: Is there a way for me to figure out products?
364 00:46:44.050 ⇒ 00:46:47.870 Robert Tseng: Dashboard, sales… ….
365 00:46:55.560 ⇒ 00:47:00.020 Annie Yu: Like, I would not know they have product data in here.
366 00:47:00.800 ⇒ 00:47:06.319 Annie Yu: So that’s something… you just, like, try and see if you can find anything like that.
367 00:47:06.630 ⇒ 00:47:09.570 Robert Tseng: Yeah, I would probably just… I’m just… I’m just…
368 00:47:09.740 ⇒ 00:47:15.740 Robert Tseng: I think that there should be product data. Now I see, okay, they have partners, they have a few brands in here.
369 00:47:15.990 ⇒ 00:47:20.070 Robert Tseng: So, I would expect that they have supporters.
370 00:47:20.890 ⇒ 00:47:26.959 Robert Tseng: … But I would just want to make sure… is that actually true?
371 00:47:33.680 ⇒ 00:47:42.200 Robert Tseng: Yeah, I’m gonna assume that these are… We can find… vendors there. …
372 00:47:43.990 ⇒ 00:47:45.739 Robert Tseng: That’ll probably be my way in.
373 00:47:49.820 ⇒ 00:47:51.310 Robert Tseng: …
374 00:47:58.830 ⇒ 00:47:59.600 Robert Tseng: Okay.
375 00:48:00.990 ⇒ 00:48:05.379 Robert Tseng: Somebody is in here, retail engagement officer, that was me.
376 00:48:06.100 ⇒ 00:48:13.670 Robert Tseng: Inventory, retail dashboard… oh man, this is looking really ugly.
377 00:48:14.910 ⇒ 00:48:17.500 Robert Tseng: How do I… this?
378 00:48:17.800 ⇒ 00:48:21.239 Robert Tseng: Uniques of tab inventory.
379 00:48:22.470 ⇒ 00:48:24.419 Robert Tseng: Don’t know what this means.
380 00:48:32.250 ⇒ 00:48:33.210 Robert Tseng: Yeah.
381 00:48:33.530 ⇒ 00:48:40.329 Robert Tseng: Okay, maybe I can’t answer the product question, or I just don’t know how to, off the top of my head. So I’m gonna move on from that for now, and….
382 00:48:40.380 ⇒ 00:48:41.130 Annie Yu: If they do.
383 00:48:41.490 ⇒ 00:48:46.839 Annie Yu: Everything is in this, … everything is considered an event.
384 00:48:48.800 ⇒ 00:48:58.629 Robert Tseng: Yeah, everything is an event, so if it were to be a product, I would expect it to be on a retail user, as a user property, possibly.
385 00:48:58.860 ⇒ 00:49:08.319 Robert Tseng: So… I guess I could go into users, and I could add some filters to say role.
386 00:49:08.930 ⇒ 00:49:12.610 Robert Tseng: Is a retailer admin.
387 00:49:12.700 ⇒ 00:49:13.790 Annie Yu: He said….
388 00:49:14.370 ⇒ 00:49:16.740 Robert Tseng: Maybe I’ll just pick this one person…
389 00:49:16.940 ⇒ 00:49:19.189 Robert Tseng: What do we know about this person?
390 00:49:19.310 ⇒ 00:49:27.460 Robert Tseng: … groups… Retailers, park tables…
391 00:49:31.350 ⇒ 00:49:34.729 Robert Tseng: properties… dark details…
392 00:49:38.370 ⇒ 00:49:42.230 Robert Tseng: Yeah, I don’t really think there is product data.
393 00:49:45.950 ⇒ 00:49:52.280 Robert Tseng: I think if we… let’s look at, … Retailer Dashboard…
394 00:49:58.100 ⇒ 00:50:01.679 Robert Tseng: Yeah, this is honestly pretty unusable. …
395 00:50:09.950 ⇒ 00:50:20.979 Robert Tseng: Yeah, so I guess I would make a note to myself, it’s like, okay, we can’t even filter by product data here, and like, the way that product data should be set up, and I would probably write this down somewhere.
396 00:50:21.540 ⇒ 00:50:27.380 Robert Tseng: I guess if I can pull this up, and I go to Spark Plug…
397 00:50:27.930 ⇒ 00:50:30.090 Robert Tseng: I did have document already.
398 00:50:43.200 ⇒ 00:50:44.220 Robert Tseng: Okay.
399 00:50:45.130 ⇒ 00:51:02.520 Robert Tseng: no product data in MixedPanel, … would expect… Product data… the, … to be a… accessible.
400 00:51:05.130 ⇒ 00:51:06.529 Robert Tseng: That’s a proper review.
401 00:51:08.450 ⇒ 00:51:11.799 Robert Tseng: for a, I don’t know, retailer user.
402 00:51:14.110 ⇒ 00:51:20.790 Robert Tseng: … Yeah, so anyway, I would just kind of go… I would do that, and …
403 00:51:22.000 ⇒ 00:51:27.319 Robert Tseng: kind of move on to the next question. Next question for me would just be…
404 00:51:28.060 ⇒ 00:51:34.279 Robert Tseng: But, like, figuring out… are they creating more sparks consistently? So maybe I’m looking at…
405 00:51:34.590 ⇒ 00:51:47.889 Robert Tseng: Number of sparks created, so… What is the events when a spark is created? … Spark, read a spark.
406 00:51:49.610 ⇒ 00:51:53.190 Robert Tseng: Commission, test, tests, ….
407 00:52:00.810 ⇒ 00:52:05.089 Annie Yu: So your account can also create a spark?
408 00:52:05.420 ⇒ 00:52:12.349 Robert Tseng: Yeah, I mean, this is just a demo account, so I’m just, like, kind of throwing whatever into it. Like, I don’t fully understand the mechanics here, but…
409 00:52:12.950 ⇒ 00:52:19.090 Robert Tseng: … It’s enough to… Finish this part.
410 00:52:23.120 ⇒ 00:52:25.019 Robert Tseng: I don’t know, 15.
411 00:52:26.920 ⇒ 00:52:27.700 Robert Tseng: Bye.
412 00:52:28.340 ⇒ 00:52:30.450 Robert Tseng: Oh, it’s $5, I see.
413 00:52:31.590 ⇒ 00:52:32.380 Robert Tseng: …
414 00:52:46.300 ⇒ 00:52:47.090 Robert Tseng: ….
415 00:52:52.180 ⇒ 00:52:53.770 Annie Yu: Is it not recorded?
416 00:52:55.110 ⇒ 00:52:57.430 Robert Tseng: May not be coming through…
417 00:53:14.920 ⇒ 00:53:22.739 Robert Tseng: Yeah, I mean, I guess I… I would really need that MixedPanel debugger tool, otherwise I’m not gonna know what to do.
418 00:53:45.570 ⇒ 00:53:53.970 Robert Tseng: Maybe I should go into Fresh Paints, This is their CDP… …
419 00:54:07.150 ⇒ 00:54:09.150 Robert Tseng: Really? …
420 00:55:01.280 ⇒ 00:55:08.720 Robert Tseng: what do I know about my… Device ID….
421 00:55:14.740 ⇒ 00:55:19.109 Annie Yu: So what’s the relationship between fresh paint and mixed panel?
422 00:55:19.800 ⇒ 00:55:28.260 Robert Tseng: FreshPaint is where they do their event instrumentation, and they push events into Mixpanel. … Yeah.
423 00:55:31.130 ⇒ 00:55:40.050 Robert Tseng: So, if it’s gonna show up in here, it would probably show up in Fresh Paint, even if it didn’t get to MixedPanel. So…
424 00:55:40.640 ⇒ 00:55:47.510 Robert Tseng: I’m trying to identify myself, … I don’t.
425 00:55:47.650 ⇒ 00:55:48.970 Robert Tseng: No…
426 00:55:52.410 ⇒ 00:55:55.569 Robert Tseng: Please, just give me something easy to use.
427 00:56:28.910 ⇒ 00:56:33.770 Robert Tseng: Yeah, I mean… This’ll take me another…
428 00:56:34.220 ⇒ 00:56:37.719 Robert Tseng: However long to go and figure out, like, who this…
429 00:56:37.970 ⇒ 00:56:41.150 Robert Tseng: How do I even identify myself? Oh my goodness.
430 00:56:41.360 ⇒ 00:56:44.390 Annie Yu: Only to figure out what event means.
431 00:56:44.390 ⇒ 00:56:52.699 Robert Tseng: just to figure out what event I’m actually logging. Like, it’s just, like, there’s so much overhead in, like, the set… and just setting it up, …
432 00:56:54.230 ⇒ 00:57:00.940 Robert Tseng: I understand this is not easy to jump in and out of. I… I feel like I just have this…
433 00:57:01.340 ⇒ 00:57:10.879 Robert Tseng: I’ve… I’ve completely forgotten everything about… about this client over the past, like, 2 months, so… I could hardly recreate what I… what I did before. And I’m…
434 00:57:11.010 ⇒ 00:57:20.479 Robert Tseng: I feel like this was kind of frustrating for me. I feel like I’ve been trying to show you how to do stuff in this environment for the past hour, and it’s just not really…
435 00:57:21.590 ⇒ 00:57:25.149 Robert Tseng: It’s not really working, so, …
436 00:57:28.480 ⇒ 00:57:35.589 Robert Tseng: But, like, to me, like, all of this time, I don’t care, because I can just… we can just bill it to the client. But, like, we…
437 00:57:35.810 ⇒ 00:57:38.890 Robert Tseng: I… Just need to make it…
438 00:57:39.210 ⇒ 00:57:46.820 Robert Tseng: so that it’s repeatable, and there’s some structure in your exploration of it. They’re literally just paying us, like.
439 00:57:47.510 ⇒ 00:57:52.280 Robert Tseng: they’re just literally paying us to just explore things. So, like, I don’t…
440 00:57:52.770 ⇒ 00:57:58.079 Robert Tseng: I don’t mind that it takes a long time and everything, like, I…
441 00:57:59.500 ⇒ 00:58:05.189 Robert Tseng: I guess what I had written was not super helpful, or I’m not really sure, like, I…
442 00:58:08.720 ⇒ 00:58:10.539 Robert Tseng: Yeah, like, I would…
443 00:58:11.020 ⇒ 00:58:17.830 Robert Tseng: I think I would just spend more time and try to get somewhere. Like, now that I’ve already started, I’m gonna try to…
444 00:58:18.250 ⇒ 00:58:19.739 Robert Tseng: Trying to bring it somewhere.
445 00:58:19.840 ⇒ 00:58:28.440 Robert Tseng: But, I understand that, like, an hour has passed, and here we are. Like, we barely built one chart, so…
446 00:58:28.650 ⇒ 00:58:29.580 Robert Tseng: ….
447 00:58:29.840 ⇒ 00:58:36.410 Annie Yu: Are you gonna… are you gonna continue now? If you are, can I stay on for a little bit?
448 00:58:36.410 ⇒ 00:58:44.329 Robert Tseng: Yeah, I’m gonna continue, but I probably, like, I’m having a hard time, like, explaining while also doing, like, I’ll probably just have to….
449 00:58:44.330 ⇒ 00:58:46.679 Annie Yu: Lock in and just, and just do something.
450 00:58:46.850 ⇒ 00:58:52.240 Robert Tseng: You’re welcome to just, like, stay on and do your own thing, or whatever, I don’t know, but, like, I…
451 00:58:52.830 ⇒ 00:58:58.550 Robert Tseng: Like, yeah, I just… I… I do want to try to… try to get… get to… get to something here.
452 00:58:58.960 ⇒ 00:59:00.000 Annie Yu: Yeah, yeah.
453 00:59:00.000 ⇒ 00:59:02.260 Robert Tseng: Okay, cool. So, …
454 00:59:02.490 ⇒ 00:59:12.530 Robert Tseng: I’ll keep sharing my screen, you can look at what I’m doing, whatever, or you can just wait until I, like, share the Zoom recording, up to you. But, I think I have….
455 00:59:12.690 ⇒ 00:59:13.180 Annie Yu: No, I think.
456 00:59:13.180 ⇒ 00:59:13.980 Robert Tseng: about….
457 00:59:13.980 ⇒ 00:59:18.570 Annie Yu: But this, … The, the sparks thing.
458 00:59:19.230 ⇒ 00:59:21.930 Robert Tseng: You do want to figure it out? Okay, sure. Alright, I’m gonna just…
459 00:59:22.460 ⇒ 00:59:26.909 Robert Tseng: I’m gonna shut up and just try to… try to work on it. I’m trying to figure it out.
460 00:59:27.110 ⇒ 00:59:27.780 Robert Tseng: Bye.
461 00:59:35.170 ⇒ 00:59:38.220 Robert Tseng: So where was I? Brush paints… no.
462 00:59:42.800 ⇒ 00:59:45.720 Robert Tseng: The test will be more effective.
463 01:00:21.660 ⇒ 01:00:22.640 Robert Tseng: There we go.
464 01:00:24.220 ⇒ 01:00:34.450 Robert Tseng: Okay, so… Spark Details, Spark Movers, Spark Commission, … scroll…
465 01:01:12.530 ⇒ 01:01:13.640 Robert Tseng: Come on.
466 01:01:14.300 ⇒ 01:01:17.449 Robert Tseng: I could use something that’s live. I don’t really know what’s life.
467 01:01:18.200 ⇒ 01:01:20.499 Robert Tseng: Zero products that matched.
468 01:01:21.140 ⇒ 01:01:22.990 Robert Tseng: Well, what matches?
469 01:01:34.750 ⇒ 01:01:35.560 Robert Tseng: Okay.
470 01:01:36.260 ⇒ 01:01:37.290 Robert Tseng: machine.
471 01:01:51.380 ⇒ 01:01:53.339 Robert Tseng: Those are the launch spark.
472 01:01:54.650 ⇒ 01:01:58.140 Annie Yu: I think it’s LaunchSpark, because that Spark Wizard got.
473 01:01:58.140 ⇒ 01:01:58.890 Robert Tseng: Yeah.
474 01:02:00.820 ⇒ 01:02:04.470 Robert Tseng: Yeah, which is… this is just silly.
475 01:02:04.770 ⇒ 01:02:05.700 Annie Yu: ….
476 01:02:08.580 ⇒ 01:02:17.700 Robert Tseng: Spark creation… Workflow is dependent on… it’s… is triggered.
477 01:02:17.970 ⇒ 01:02:22.629 Robert Tseng: only on button… Presses and quits.
478 01:02:25.460 ⇒ 01:02:34.179 Robert Tseng: This is… These events are unreliable, need to send… What’s this going on?
479 01:02:34.370 ⇒ 01:02:41.270 Robert Tseng: To send this a verified event for when a… Launch.
480 01:02:42.460 ⇒ 01:02:43.540 Robert Tseng: Thank you, watch.
481 01:02:47.690 ⇒ 01:02:49.579 Robert Tseng: So, it was successful.
482 01:02:53.670 ⇒ 01:02:57.060 Robert Tseng: every intermediary step in the spark.
483 01:02:57.600 ⇒ 01:03:01.379 Robert Tseng: Creation, workflow… It’s still cute.
484 01:03:04.490 ⇒ 01:03:06.349 Robert Tseng: I haven’t outlined?
485 01:03:06.570 ⇒ 01:03:16.690 Robert Tseng: Too much noise, and telemetry… Basically, capturing everything, but not being able to interpret anything.
486 01:03:17.560 ⇒ 01:03:23.210 Robert Tseng: Okay, but for the most part, what we can do, well, we can just create a custom funnel.
487 01:03:23.730 ⇒ 01:03:25.370 Robert Tseng: Custom funnel….
488 01:03:25.590 ⇒ 01:03:28.400 Annie Yu: That click. That click is not an event.
489 01:03:29.670 ⇒ 01:03:30.420 Robert Tseng: Sorry?
490 01:03:30.420 ⇒ 01:03:34.399 Annie Yu: You click Launch Spark, it’s not an event that we can track.
491 01:03:35.720 ⇒ 01:03:36.600 Robert Tseng: it’s….
492 01:03:37.860 ⇒ 01:03:38.590 Annie Yu: If….
493 01:03:40.570 ⇒ 01:03:42.940 Robert Tseng: May or may not be.
494 01:03:44.670 ⇒ 01:03:46.699 Robert Tseng: You don’t think it is, Mike?
495 01:03:46.700 ⇒ 01:03:48.910 Annie Yu: No, no, I’m asking, I’m not sure.
496 01:03:48.910 ⇒ 01:03:50.130 Robert Tseng: …
497 01:03:58.250 ⇒ 01:04:00.209 Robert Tseng: Yeah, it might not be.
498 01:04:02.290 ⇒ 01:04:03.350 Robert Tseng: I don’t know.
499 01:04:18.290 ⇒ 01:04:19.280 Robert Tseng: launch.
500 01:04:26.400 ⇒ 01:04:27.520 Robert Tseng: Large.
501 01:04:34.510 ⇒ 01:04:35.880 Robert Tseng: or Spark.
502 01:04:59.150 ⇒ 01:05:00.150 Robert Tseng: Okay, well….
503 01:06:04.580 ⇒ 01:06:05.830 Annie Yu: Yeah, nice.
504 01:06:05.830 ⇒ 01:06:17.320 Robert Tseng: Yeah, now there’s an event there. So, I mean, it’s not best practice to be creating events on the fly, obviously, but I guess we do have some, like, degree flexibility to get to what we want to see.
505 01:06:17.700 ⇒ 01:06:23.499 Robert Tseng: Unfortunately, it doesn’t really backfill things, so it’s not, like, a perfect solve.
506 01:06:31.610 ⇒ 01:06:36.590 Robert Tseng: Okay, well, so, I mean, I guess the answer would just be, well, this is a proxy for creating
507 01:06:36.700 ⇒ 01:06:42.770 Robert Tseng: like, we could just use it as a proxy. So instead of any active event, I could just do button create spark for now.
508 01:06:43.530 ⇒ 01:06:50.210 Robert Tseng: And, … Nothing was created.
509 01:06:51.770 ⇒ 01:06:52.950 Robert Tseng: Incredible.
510 01:06:53.130 ⇒ 01:06:54.000 Robert Tseng: ….
511 01:06:56.060 ⇒ 01:07:01.790 Annie Yu: But if nothing was created, how do the employees Complete sparks.
512 01:07:03.870 ⇒ 01:07:05.200 Robert Tseng: Bye.
513 01:07:05.880 ⇒ 01:07:07.050 Robert Tseng: Don’t know.
514 01:07:08.910 ⇒ 01:07:10.430 Robert Tseng: This must not be right.
515 01:07:26.790 ⇒ 01:07:29.700 Robert Tseng: Oh, well, duh, because the roll is off.
516 01:07:38.280 ⇒ 01:07:43.320 Robert Tseng: retail app, and specifically, So, that’s…
517 01:07:44.150 ⇒ 01:07:46.139 Robert Tseng: You could do a stack column.
518 01:07:47.230 ⇒ 01:07:48.270 Robert Tseng: Spark.
519 01:07:48.770 ⇒ 01:07:52.850 Robert Tseng: Dashboard… with Spark.
520 01:08:01.440 ⇒ 01:08:05.689 Robert Tseng: Wow, so JARS is not creating any sparks. Well, how’s that possible?
521 01:08:05.920 ⇒ 01:08:08.299 Robert Tseng: … Leave it ever.
522 01:08:09.270 ⇒ 01:08:12.369 Annie Yu: So, is there a way to verify if they’re continuing
523 01:08:13.170 ⇒ 01:08:15.580 Annie Yu: To use whatever they had before.
524 01:08:18.020 ⇒ 01:08:24.419 Robert Tseng: … If they’re continuing to use… to go against these parks.
525 01:08:24.420 ⇒ 01:08:26.220 Annie Yu: Oh, sparks, yeah.
526 01:08:26.579 ⇒ 01:08:31.869 Robert Tseng: … Yeah, I shouldn’t save that first.
527 01:08:33.109 ⇒ 01:08:36.919 Robert Tseng: If are they continuing to use the same sparks?
528 01:08:37.529 ⇒ 01:08:47.209 Robert Tseng: Well, yeah, I mean, I guess if we… well, every… every re… there’d be a different Spark ID, probably, for every retailer, and so they’re not really using the same Sparks.
529 01:08:47.339 ⇒ 01:08:48.759 Annie Yu: Yeah. ….
530 01:08:48.759 ⇒ 01:08:50.419 Robert Tseng: So if we want to see…
531 01:08:50.699 ⇒ 01:08:53.969 Robert Tseng: I guess we could just pick… 1…
532 01:08:54.569 ⇒ 01:08:57.439 Robert Tseng: Okay, let’s just pick one retailer.
533 01:08:59.259 ⇒ 01:09:10.979 Robert Tseng: And… we’ll say… off of, like, Spark… Complete… Or….
534 01:09:17.810 ⇒ 01:09:19.870 Robert Tseng: Where, …
535 01:09:23.399 ⇒ 01:09:26.119 Robert Tseng: Is there, like, a Spark ID?
536 01:09:29.140 ⇒ 01:09:35.189 Robert Tseng: … No, these are all user IDs.
537 01:09:49.109 ⇒ 01:09:50.390 Robert Tseng: Yeah, no.
538 01:09:51.700 ⇒ 01:09:55.830 Robert Tseng: what would a Spark look like? They have a Spark name.
539 01:09:58.650 ⇒ 01:10:00.640 Robert Tseng: Spark details…
540 01:10:20.430 ⇒ 01:10:28.499 Robert Tseng: Yeah, this is just a page view, it’s not actually logging anything about the spark. There’s just too many page view events that are, like, not helpful.
541 01:10:30.120 ⇒ 01:10:31.270 Robert Tseng: ….
542 01:10:31.380 ⇒ 01:10:38.880 Annie Yu: During the Spark creation steps, there should be different Page view, right?
543 01:10:40.660 ⇒ 01:10:41.530 Robert Tseng: Yeah.
544 01:10:42.180 ⇒ 01:10:47.250 Robert Tseng: I mean, like, the way that I would describe it is, like, …
545 01:10:50.510 ⇒ 01:10:53.360 Robert Tseng: Data design, …
546 01:10:58.390 ⇒ 01:10:59.850 Robert Tseng: So…
547 01:11:13.190 ⇒ 01:11:14.070 Robert Tseng: Oh.
548 01:11:14.370 ⇒ 01:11:15.330 Robert Tseng: Oh, dear.
549 01:11:15.810 ⇒ 01:11:16.680 Robert Tseng: …
550 01:11:23.510 ⇒ 01:11:27.540 Robert Tseng: But… Already using it, question mark.
551 01:11:32.320 ⇒ 01:11:37.770 Robert Tseng: Insights… Oh, no, no, this is not it.
552 01:11:40.700 ⇒ 01:11:47.500 Robert Tseng: Man, … No, don’t upgrade, Lily.
553 01:11:50.030 ⇒ 01:11:50.870 Robert Tseng: Okay.
554 01:11:52.360 ⇒ 01:11:53.939 Robert Tseng: This is just a template.
555 01:11:54.220 ⇒ 01:12:01.870 Robert Tseng: … Okay, well, I don’t know where my event data design is, like, I…
556 01:12:03.550 ⇒ 01:12:06.929 Robert Tseng: And I can navigate this, but …
557 01:12:07.660 ⇒ 01:12:12.469 Robert Tseng: Yeah, like, if I were… hmm, I want to just edit this directly.
558 01:12:17.500 ⇒ 01:12:18.230 Robert Tseng: Okay.
559 01:12:21.550 ⇒ 01:12:25.300 Robert Tseng: Alright, let’s just say I’m redoing a spark plug.
560 01:12:26.740 ⇒ 01:12:27.430 Robert Tseng: It’s gone.
561 01:12:28.570 ⇒ 01:12:31.499 Robert Tseng: Yeah, so then I would do… oops.
562 01:12:33.500 ⇒ 01:12:35.570 Robert Tseng: Let’s create a, …
563 01:12:39.120 ⇒ 01:12:44.680 Robert Tseng: Retailer, events, all the Spark creation workflow.
564 01:12:45.240 ⇒ 01:12:46.980 Robert Tseng: That would be, like, a spark.
565 01:12:51.720 ⇒ 01:12:52.950 Robert Tseng: Spark started.
566 01:12:53.800 ⇒ 01:12:55.570 Robert Tseng: Spark launched.
567 01:12:56.280 ⇒ 01:12:57.970 Robert Tseng: Some stuff in between.
568 01:12:58.150 ⇒ 01:12:59.010 Robert Tseng: …
569 01:13:08.830 ⇒ 01:13:16.450 Robert Tseng: Yeah, so I would, like, be defining these custom events, right? And, like, this is kind of how I typically start to
570 01:13:16.830 ⇒ 01:13:22.809 Robert Tseng: frame, like, what my ideal tracking plan is when I’m, like, going through a product,
571 01:13:24.050 ⇒ 01:13:28.400 Robert Tseng: Yeah, obviously, you have to take it one workflow at a time, but I think if…
572 01:13:28.580 ⇒ 01:13:33.970 Robert Tseng: If it’s important for us to look at user engagement behavior, from…
573 01:13:35.110 ⇒ 01:13:44.389 Robert Tseng: like, a retailer’s perspective, I think we have to… we have to be able to capture whether or not they’re creating sparks, and then we also have to, from a user perspective.
574 01:13:45.200 ⇒ 01:13:48.649 Robert Tseng: We have to be measuring if they’re,
575 01:13:56.450 ⇒ 01:13:58.799 Robert Tseng: Whoa, that’s not what I wanted to do.
576 01:14:00.090 ⇒ 01:14:01.110 Robert Tseng: …
577 01:14:10.700 ⇒ 01:14:13.560 Robert Tseng: User, events, Spark.
578 01:14:14.990 ⇒ 01:14:17.049 Robert Tseng: Spark walking.
579 01:14:25.100 ⇒ 01:14:25.910 Robert Tseng: Spark.
580 01:14:26.800 ⇒ 01:14:28.439 Robert Tseng: Start it, if you believe.
581 01:14:32.700 ⇒ 01:14:36.610 Robert Tseng: Spark finished, or if I want that. Maybe it’s just two things.
582 01:14:39.690 ⇒ 01:14:41.669 Robert Tseng: Yeah, I think, like, what… what…
583 01:14:42.420 ⇒ 01:14:47.450 Robert Tseng: What Sparkboard clearly suffers from right now is over-tracking. They have way too many things.
584 01:14:47.870 ⇒ 01:14:51.689 Robert Tseng: They’re just tracking everything under the sun, every page view, every click.
585 01:14:51.790 ⇒ 01:14:52.580 Robert Tseng: Edit.
586 01:14:52.770 ⇒ 01:14:56.000 Robert Tseng: I think, typically, the way that I prefer to design
587 01:14:56.300 ⇒ 01:15:03.899 Robert Tseng: event tracking is that it needs to be very focused on the core workflows. And so every…
588 01:15:04.310 ⇒ 01:15:18.770 Robert Tseng: So if I were to be thinking about everything as a funnel, if I were to create a sequence of events for a retailer workflows, like the Smart Creation one, what are the main things that we need to track to get there? There is going to be a sense of, like, …
589 01:15:19.460 ⇒ 01:15:24.700 Robert Tseng: Well, I mean, this one gets complicated, because there’s user events, or, like, employee events, really.
590 01:15:24.830 ⇒ 01:15:29.640 Robert Tseng: There’s retailer events, and then there’s honestly probably, like, ….
591 01:15:30.100 ⇒ 01:15:30.960 Annie Yu: brand.
592 01:15:30.960 ⇒ 01:15:32.160 Robert Tseng: Brand, yeah.
593 01:15:32.650 ⇒ 01:15:33.890 Robert Tseng: So…
594 01:15:34.570 ⇒ 01:15:44.469 Robert Tseng: There’s brand events, I’ll just delete all this because it didn’t really matter. And you do need some general events, so these are what I would typically call our interaction events.
595 01:15:45.200 ⇒ 01:15:46.320 Robert Tseng: ….
596 01:15:46.320 ⇒ 01:15:49.840 Annie Yu: Does that mean? Like, across those roles?
597 01:15:50.400 ⇒ 01:16:01.439 Robert Tseng: Yeah, so this is what, like, a page view would be. Like, so, page view, clicks, hover, estimated, like, those things you do kind of need. Form submissions when things are hovered over, so, …
598 01:16:01.580 ⇒ 01:16:09.699 Robert Tseng: And, like, what are the important things to track about it? So, these are, like, universal events that matter no matter who it is, that’s… that’s interacting with it.
599 01:16:10.230 ⇒ 01:16:14.670 Robert Tseng: … So, I mean, they kind of already do have this, so…
600 01:16:15.090 ⇒ 01:16:22.890 Robert Tseng: Like, it’s not really something I would re-instrument. If anything, it’s missing some detail, because all of this stuff is more or less useless, it’s just, like.
601 01:16:24.340 ⇒ 01:16:27.450 Robert Tseng: I don’t know, it doesn’t tell you anything about the page, so…
602 01:16:27.890 ⇒ 01:16:41.720 Robert Tseng: Like, that could be helpful, because it gives you some measure of engagement, just, like, number of unique page views or whatever, and number of unique clicks and everything. But, like, that’s… that’s just very noisy, so it’s not really getting us to where we need to be.
603 01:16:41.950 ⇒ 01:16:42.920 Robert Tseng: …
604 01:16:43.930 ⇒ 01:16:52.630 Robert Tseng: Yeah, and then, like, the top part is, like, customer-level events, or just, like, user journey events, so… full…
605 01:16:52.750 ⇒ 01:16:53.920 Robert Tseng: I… yeah.
606 01:16:54.020 ⇒ 01:16:57.869 Robert Tseng: Employee, user milestones.
607 01:16:58.100 ⇒ 01:17:05.260 Robert Tseng: So, maybe it’s, like… Completed Spark. Completed first SPARC.
608 01:17:05.710 ⇒ 01:17:13.240 Robert Tseng: … For first, you know, typically this is… this is, like, the more standardized way to do this.
609 01:17:13.670 ⇒ 01:17:21.800 Robert Tseng: … First bar completed… First, maybe first cash out?
610 01:17:22.390 ⇒ 01:17:25.590 Robert Tseng: Like, what are the different value points that, like.
611 01:17:25.900 ⇒ 01:17:30.590 Robert Tseng: matters, like, you know, what does becoming a loyal customer mean, being at risk to churn?
612 01:17:30.940 ⇒ 01:17:33.200 Robert Tseng: … But…
613 01:17:33.910 ⇒ 01:17:48.310 Robert Tseng: yeah, I guess we can kind of, like, think about what those milestones are. This is just, like, a running list that you keep adding to. This is what would go into, like, an ideal, like, employee-customer data model. So if I were going back into Mixpanel at some point.
614 01:17:48.310 ⇒ 01:18:00.190 Robert Tseng: and I were to go to user, and I were looking at the user profiles, this is what I want to see. Like, I wouldn’t… like, all this stuff doesn’t really matter. It’s, like, useless stuff. But, like, ideally, I would be able to see
615 01:18:00.390 ⇒ 01:18:11.729 Robert Tseng: First spark completed, first cash out, da-da-da. Like, this is kind of what we’ve done on Eden. Eden is, like, first order place, time between orders, da-da-da, like, those are, like, these milestones that we care about. So, anyway.
616 01:18:12.350 ⇒ 01:18:16.379 Annie Yu: for each… per event per person.
617 01:18:16.620 ⇒ 01:18:17.460 Annie Yu: Correct.
618 01:18:17.920 ⇒ 01:18:23.399 Robert Tseng: So these are not necessarily tied to specific events. These could be, like, …
619 01:18:24.340 ⇒ 01:18:35.650 Robert Tseng: it could be, like, time-based things, too. It’s like, maybe it’s like, … Say… So… First part completed.
620 01:18:37.960 ⇒ 01:18:40.370 Annie Yu: Would that be, like, Boolean, or…?
621 01:18:40.520 ⇒ 01:18:42.639 Annie Yu: Or, like, a timestamp.
622 01:18:42.640 ⇒ 01:18:43.980 Robert Tseng: So there’s a timestamp, yeah.
623 01:18:44.190 ⇒ 01:18:47.379 Robert Tseng: That’s maybe a timestamp, but maybe it’s, like, …
624 01:18:49.100 ⇒ 01:19:00.599 Robert Tseng: Or… well, yeah, you could make it a Boolean if you want to do true-false, but I think timestamp makes more sense. And then you could do, like, a, time for spark time to completion.
625 01:19:02.070 ⇒ 01:19:09.279 Robert Tseng: Like, if you want to know, like, how long does it take for people to go and complete their first time chat kind of thing, or complete their first part?
626 01:19:09.800 ⇒ 01:19:15.170 Robert Tseng: What else is important? First cash out, first, da-da-da-da. …
627 01:19:17.600 ⇒ 01:19:20.269 Robert Tseng: Yeah, I guess I would have to go back into…
628 01:19:27.580 ⇒ 01:19:35.579 Robert Tseng: just to kind of think about, like, what are… what are those other value points? So, yeah, sparks, urns, events, maybe it’s, like, first…
629 01:19:35.880 ⇒ 01:19:38.259 Robert Tseng: Event completion as well, too.
630 01:19:41.510 ⇒ 01:19:48.290 Robert Tseng: First, events… confusion… …
631 01:20:18.510 ⇒ 01:20:26.440 Robert Tseng: Okay, but this is… these are not tracked events, so this is not, like, super important. So… Yeah.
632 01:20:26.440 ⇒ 01:20:28.050 Annie Yu: Normally set up, okay.
633 01:20:28.210 ⇒ 01:20:28.890 Robert Tseng: Yeah.
634 01:20:29.070 ⇒ 01:20:37.100 Robert Tseng: … Okay, so… I guess… Yeah, I would kind of just…
635 01:20:37.660 ⇒ 01:20:51.970 Robert Tseng: keep building this out in parallel as I’m, like, trying to answer some of these open-ended questions. I should have been doing this documentation work beforehand, because that would have made it easier to kind of, like, wrap my head around it, but I guess, like, the point here is…
636 01:20:52.300 ⇒ 01:21:00.789 Robert Tseng: Hopefully, we’re getting to a point where we’re seeing that, like, yeah, there’s, like, a disconnect between, like, what should be tracked and how it should be.
637 01:21:01.230 ⇒ 01:21:10.120 Robert Tseng: And… yeah, like, not everything is fully where we want it to be, and we have to make some assumptions if, like, things are open-ended, but, like.
638 01:21:10.230 ⇒ 01:21:17.480 Robert Tseng: you know, we can still move things… we can move the conversation forward, I think, to some… to some extent, so…
639 01:21:17.590 ⇒ 01:21:19.019 Robert Tseng: I think, …
640 01:21:20.850 ⇒ 01:21:26.860 Robert Tseng: I mean, that’s all the time I really got for them today. I’m not gonna spend any more time on this right now. …
641 01:21:27.380 ⇒ 01:21:28.740 Robert Tseng: I…
642 01:21:32.650 ⇒ 01:21:36.329 Robert Tseng: I guess this isn’t the easiest one to start with, like, for me.
643 01:21:36.920 ⇒ 01:21:43.959 Robert Tseng: because the relationship with Sparkplug is kind of, like, stalled, I care more about the new clients that we’ve started, so, like.
644 01:21:44.170 ⇒ 01:21:49.530 Robert Tseng: I don’t know, we have, like, a bunch of other new clients. We have, like, an Amplitude, if I go and I log in.
645 01:21:49.740 ⇒ 01:21:54.510 Robert Tseng: I have, one… Looks like on Ellie Mental Health.
646 01:21:56.030 ⇒ 01:21:59.110 Robert Tseng: And README as well, so this is, like, another one.
647 01:21:59.300 ⇒ 01:22:06.199 Robert Tseng: It’s like, okay, great, like, they have a bunch of reports in here, kind of a similar deal, like, I have to go and figure out, like.
648 01:22:06.580 ⇒ 01:22:14.319 Robert Tseng: I have to… I have to explore, like, what’s going on with this… with this client. So, maybe I’ll start here. I haven’t actually started.
649 01:22:14.570 ⇒ 01:22:22.150 Robert Tseng: On my scheduling funnel, … Oops, I saw it.
650 01:22:27.600 ⇒ 01:22:31.280 Robert Tseng: This one is probably more familiar to you, this probably feels more like an Eden.
651 01:22:31.700 ⇒ 01:22:36.060 Robert Tseng: But yeah, their workflow is, yeah, you go and you search for a provider.
652 01:22:36.890 ⇒ 01:22:38.260 Annie Yu: I guess.
653 01:22:38.560 ⇒ 01:22:45.550 Robert Tseng: And then you’re looking at… Provider availability… …
654 01:23:01.880 ⇒ 01:23:02.720 Robert Tseng: Alright.
655 01:23:16.760 ⇒ 01:23:17.740 Robert Tseng: herbs.
656 01:23:25.440 ⇒ 01:23:27.429 Robert Tseng: Come on, I just want to show something.
657 01:23:41.250 ⇒ 01:23:45.230 Robert Tseng: Huh. There are no events being tracked. What the heck is going on here?
658 01:23:46.440 ⇒ 01:23:47.670 Robert Tseng: …
659 01:24:09.390 ⇒ 01:24:17.530 Robert Tseng: Oh, dear. This is another… this is another messed up thing. It’s not even… they’re not tracking anything. Where did they get any of these numbers from?
660 01:24:20.000 ⇒ 01:24:23.420 Robert Tseng: I don’t know. Well, anyway, my point is, …
661 01:24:26.820 ⇒ 01:24:34.620 Robert Tseng: That is another client that we started that I needed to get working on, and …
662 01:24:35.570 ⇒ 01:24:45.790 Robert Tseng: README is another client that we’re doing that I… okay, this one I feel more comfortable with, because I literally just worked on this yesterday, so I could speak to this a lot more clearly.
663 01:24:47.940 ⇒ 01:24:58.120 Robert Tseng: I built out this self-serve conversion space, new user onboarding to conversion analysis, … Excuse me.
664 01:24:58.320 ⇒ 01:25:01.980 Robert Tseng: What IO… What the world.
665 01:25:02.360 ⇒ 01:25:03.780 Robert Tseng: I’m repeating.
666 01:25:19.420 ⇒ 01:25:23.590 Robert Tseng: Yeah, so this is, like, a technical docs company, so I…
667 01:25:24.180 ⇒ 01:25:39.269 Robert Tseng: This one was a lot easier, because I set up all of the track… that tagging and tracking from scratch, more or less, so I… or at least I guided them on how to do it, so I went through the whole event data design process, and they went and they implemented some of the events.
668 01:25:39.380 ⇒ 01:25:44.489 Robert Tseng: So now, you know, the head of growth has all these questions about
669 01:25:44.880 ⇒ 01:25:47.519 Robert Tseng: I can give her some top-level metrics on
670 01:25:49.210 ⇒ 01:25:53.209 Robert Tseng: New users who are starting onboarding, who are completing onboarding, etc.
671 01:25:53.310 ⇒ 01:25:59.020 Robert Tseng: … I suppose, … let’s see…
672 01:26:07.870 ⇒ 01:26:13.689 Robert Tseng: Oh, come on. Like, the debugger chill’s not working for anything. What is going on?
673 01:26:15.110 ⇒ 01:26:18.560 Annie Yu: Do you have to log in for this extension, or…?
674 01:26:19.060 ⇒ 01:26:23.529 Robert Tseng: You usually should just start tracking, it doesn’t really matter, I think.
675 01:26:26.210 ⇒ 01:26:29.680 Robert Tseng: Maybe I’m wrong, I… oh, there we go. Okay.
676 01:26:34.550 ⇒ 01:26:39.019 Robert Tseng: Yeah, sorry, the browser I usually use doesn’t really let me do that, so, …
677 01:26:41.170 ⇒ 01:26:51.130 Robert Tseng: Yeah, I mean, anyway, all these tools have some sort of browser extension, you can go and you can do sign-ups and stuff, and I don’t know, just do TestSSP.
678 01:26:51.890 ⇒ 01:27:01.530 Robert Tseng: I.O, … I’ll test… Yeah, so I was, like, logging all this stuff, and, like.
679 01:27:04.350 ⇒ 01:27:09.789 Robert Tseng: Anyway, like, it’s a similar situation where I had to go in and, like, learn their products, click around.
680 01:27:09.910 ⇒ 01:27:16.220 Robert Tseng: Figure out, like, what’s actually going on, how do we track this? How do we design the events that we need to track?
681 01:27:16.500 ⇒ 01:27:32.060 Robert Tseng: What’s the most important event that we’re trying to get to? That was their whole onboarding process. They care about driving self-serve users to paid users. And so, yeah, there’s a lot of functionality in here, but then let’s just kind of go straight to the billing, billing kind of situation.
682 01:27:32.280 ⇒ 01:27:34.370 Robert Tseng: They’re clicking on this plan.
683 01:27:34.540 ⇒ 01:27:43.150 Robert Tseng: And then we’re looking at users that are upgrading to prepaid users. And so, in the past month, they had 5,000 users come in, finished
684 01:27:43.670 ⇒ 01:27:47.760 Robert Tseng: You know, 20% of them, or 14… 15% of them finish.
685 01:27:48.230 ⇒ 01:27:54.140 Robert Tseng: Onboarding, and then only an even smaller percentage of those end up becoming first-time paying users.
686 01:27:54.300 ⇒ 01:28:10.560 Robert Tseng: But at least I have a very clear funnel of, like, where sign-up started, completed, project created, onboarding finished, and so there’s additional analysis that now the head of growth is asking me, well, why is the drop-off here so big? Why are 50% of people that are starting to sign up not finishing?
687 01:28:10.900 ⇒ 01:28:27.809 Robert Tseng: So, that gives us some more structure for how we approach analysis. So, I would say this is, like, they had been a very more straightforward kind of product analytics setup, because we started from nothing, I gave them some top-level visibility into, you know, their onboarding or early engagement funnel.
688 01:28:27.820 ⇒ 01:28:31.550 Robert Tseng: Their main activation event, which is the paywall.
689 01:28:31.570 ⇒ 01:28:44.070 Robert Tseng: And then we can work backwards from there to figure out, like, what… how do we need to, like, what are… how are users getting there? How do we drive up more, people from… that are ending up going to pay… payments? So…
690 01:28:44.480 ⇒ 01:28:45.820 Robert Tseng: I guess, …
691 01:28:46.130 ⇒ 01:28:52.249 Robert Tseng: Yeah, like, this is… this is an example of something that’s going a little bit more… that’s going more successfully.
692 01:28:52.390 ⇒ 01:28:53.340 Robert Tseng: ….
693 01:28:54.190 ⇒ 01:29:00.189 Annie Yu: I know I’ve kind of been on call with you for a while, I’ve walked through, like, 3 examples of product analytics work.
694 01:29:00.260 ⇒ 01:29:01.120 Robert Tseng: …
695 01:29:01.750 ⇒ 01:29:07.250 Robert Tseng: I guess, kind of, what do you… what do you think… how would you say… what’s different across,
696 01:29:07.840 ⇒ 01:29:18.799 Robert Tseng: Sparkplug, Ellie, and Readme from the three that you just… from you just saw, in terms of, like, the stage we’re at, like, and how you think that we need to approach these three clients differently.
697 01:29:19.750 ⇒ 01:29:26.390 Annie Yu: … I, I would say, I think REAPME is… like…
698 01:29:26.730 ⇒ 01:29:35.750 Annie Yu: very straightforward. I think I… even you just scrolled it down, I can understand what it means, like, each funnel, and Spark Plug is…
699 01:29:35.850 ⇒ 01:29:43.560 Annie Yu: Something more, … I think it sounds more, like, challenging compared to all the others.
700 01:29:43.930 ⇒ 01:29:44.660 Robert Tseng: Yeah.
701 01:29:44.660 ⇒ 01:29:52.190 Annie Yu: I think one thing I do want to ask is, so when we work with MixedPanel and, like, Amplitude, do we just…
702 01:29:52.890 ⇒ 01:30:09.229 Annie Yu: I know that, like, we expected to see some properties, like, product size, like, product count, and things of that sort. So, do we just stay within these tools, and then try to get what we can get?
703 01:30:09.820 ⇒ 01:30:12.189 Annie Yu: Like, those property data.
704 01:30:12.250 ⇒ 01:30:20.980 Robert Tseng: Yeah, well, so I think you’re kind of, like, imagining that you’re not constrained by the tools, and you want to go and get the things that you ideally want.
705 01:30:21.040 ⇒ 01:30:34.479 Robert Tseng: If it’s not there, then I would just note it down. This is part of, like, how we sell into… we sell into more work with the clients. Like, if you think about Eden, for example, Eden started off as a MixedPanel client. It started off as….
706 01:30:35.100 ⇒ 01:30:35.540 Annie Yu: Oh, is that….
707 01:30:35.540 ⇒ 01:30:40.519 Robert Tseng: as… as crazy as spark plug. Just, like, literally nothing in Mixpanel.
708 01:30:40.850 ⇒ 01:30:52.770 Robert Tseng: And I kind of was just going along, trying to, like, figure out, like, what their MixPanel… I mean, their MixPanel still is messy. We’ve… it’s been… it’s been 8 months, and I still haven’t come back to Mixpanel, but it doesn’t matter, because
709 01:30:52.890 ⇒ 01:31:10.799 Robert Tseng: companies think that Nixpanel and Amplitude are going to be, like, the tool that they use for everything, but the reality is it’s not. Like, you do need all this other stuff, right? So we went from Nixpanel, I asked a bunch of questions about what else… what other data that was not there, and that opened up conversations to be like.
710 01:31:10.800 ⇒ 01:31:18.509 Robert Tseng: oh, actually, you know, we can’t look at products in Mixpanel. We don’t… I don’t believe, order… order counts in Mixpanel.
711 01:31:18.910 ⇒ 01:31:34.839 Robert Tseng: And it’s like, why? Where’s that data coming from? Oh, it’s coming from a system called BAST. And then it’s like, okay, well, where’s Bast data going to? Well, we have these webhooks kind of capturing segment, and then they just land in BigQuery. But no, I wouldn’t really know what to do with BigQuery, except for plug it to Looker Studio.
712 01:31:35.180 ⇒ 01:31:42.639 Robert Tseng: Right? And so, like, that kind of helped open the conversation for me to be like, okay, well, actually, what they need is,
713 01:31:42.680 ⇒ 01:31:57.360 Robert Tseng: They need to be able to accurately measure transactions… transactional data, because that’s ultimately what matters. Customer behavior data in Nixpanel is never going to be completely accurate, doesn’t fully matter for a CPG company. It’s, like, good… it’s a nice-to-have, but it’s not a must-have.
714 01:31:57.430 ⇒ 01:32:10.099 Robert Tseng: And so that allowed me, within, you know, the first month, to be able to go back to them and be like, hey, I think you need to invest in this data warehousing thing, like, bring, like, Brainforge on to go and do that. And so then it expanded, right?
715 01:32:10.290 ⇒ 01:32:26.079 Robert Tseng: So I’m just… I’m just saying, like, but all these clients, not all of them will get to that point. I think, you know, Eden is a… is a good use case. You know, they went from paying us 35K a month now. It’s like, you know, it’s… it’s, like, that’s…
716 01:32:26.130 ⇒ 01:32:42.490 Robert Tseng: that’s, like, that’s what I want to… I want to see more clients kind of graduating to that… to that stage. And … I think the needs evolve over time, and that’s part of the dynamic kind of, like, work here. And we’ve done so much stuff for Eden, like, and I… I just think that I…
717 01:32:42.660 ⇒ 01:32:53.530 Robert Tseng: I’m trying to activate these other clients as well. Like, I think README has good opportunity. They started off at, like, 5K a month, and now they’re paying us 10K a month. …
718 01:32:54.330 ⇒ 01:33:01.699 Robert Tseng: like, LE Health, you know, I’m trying to go even higher. Like, I just started working with them this week.
719 01:33:01.940 ⇒ 01:33:06.610 Robert Tseng: And they’re paying us 10K a month, and, like, you know, so I think, like, the…
720 01:33:06.780 ⇒ 01:33:23.409 Robert Tseng: hopefully you’re getting a sense for, like, this is how the business has been growing. Like, I think, you know, 80% of our revenue is coming from clients that are on the data side. The AI work is still kind of, you know, whatever it is, it’s kind of a gamble, in my opinion. So, I…
721 01:33:23.870 ⇒ 01:33:37.870 Robert Tseng: I want, you know, our analysis work to just be really solid and kind of help us graduate to the next stage. The problem is just that I’m… it’s… I’m the only one that’s been starting these clients.
722 01:33:37.870 ⇒ 01:33:48.489 Robert Tseng: and then converting them into bigger contracts. So, I… I do kind of need, like, assistance in kind of doing some of this initial work, where you’re just…
723 01:33:48.850 ⇒ 01:33:57.040 Robert Tseng: kind of doing data discovery, thinking about, like, what’s important for the clients, you’re using these tools, Mixpanel and Amplitude.
724 01:33:57.040 ⇒ 01:34:09.839 Robert Tseng: But this is just the starting point. It’s not gonna… you’re not gonna find everything you need. I don’t think this is what README needs. I think it’ll help to some extent, but I know that README needs more than this, and so I’m hoping that along the way.
725 01:34:09.840 ⇒ 01:34:17.850 Robert Tseng: well, I’ll be able to go, and I’m gonna… I’m gonna push for more from them, and hopefully graduate them to the next tier as well.
726 01:34:17.970 ⇒ 01:34:22.480 Robert Tseng: … But, anyways, I think, that’s…
727 01:34:22.610 ⇒ 01:34:27.450 Robert Tseng: That’s, to me, is one of our main growth strategies for how
728 01:34:27.590 ⇒ 01:34:33.780 Robert Tseng: this business works. Like, we start off by taking on something like this.
729 01:34:33.880 ⇒ 01:34:45.249 Robert Tseng: And we learn a lot about the business, what’s missing along the way, and that helps us to kind of craft the story of, like, what they need next, and we can go and expand that contract.
730 01:34:46.040 ⇒ 01:34:58.620 Annie Yu: So, follow-up question would be… so I guess you kind of answered that. I was gonna ask, with these projects using MixedPanel and Amplitude, was the goal
731 01:34:59.900 ⇒ 01:35:19.649 Annie Yu: figure out, like, because we want to win… go from, like, what happened to what happened and what to do about it, and the what to do about it part, I was gonna ask, was it mainly focusing on, like, optimizing their digital products, like, where the buttons should go, or, how… how should they message about things, so it’s…
732 01:35:20.010 ⇒ 01:35:25.390 Annie Yu: So, are we trying to stay within that… …
733 01:35:25.630 ⇒ 01:35:29.469 Annie Yu: That scope, at least for the early clients.
734 01:35:30.330 ⇒ 01:35:39.410 Robert Tseng: Yeah, that’s a good question. So, I think it depends on, like, who we’re working with. So, like, with README, for example.
735 01:35:40.490 ⇒ 01:35:47.230 Robert Tseng: I show README the initial stuff, and then her stakeholder gets excited from what I’ve shown her, and she’s just like.
736 01:35:47.330 ⇒ 01:36:00.420 Robert Tseng: oh my god, I’ve never seen this before. I have all these ideas. We should try all these different hypotheses. And she’s gonna go and take action on it. Like, that’s… to me, that’s a great… that’s a good… that’s what a good stakeholder should do. She should be like.
737 01:36:00.980 ⇒ 01:36:14.650 Robert Tseng: oh, we need to go and, you know, make launch… the launch button better. She’s gonna go to her designer, tell her designer to try different things. Her designer pushed a recent, update, like, last week, and she wants to see the impact of the update.
738 01:36:14.650 ⇒ 01:36:24.129 Robert Tseng: So that… then we become a partner to them. So, as a stakeholder comes to us with, like, I’m gonna add this feature, I’m gonna try this thing, move this button, add this new workflow.
739 01:36:24.200 ⇒ 01:36:31.970 Robert Tseng: like, then we’re using these tools to kind of, like, help them measure, well, did it actually do anything, right? And, …
740 01:36:32.280 ⇒ 01:36:42.620 Robert Tseng: Yeah, so, like, I think that’s kind of, like, part of her impact, and yeah, like, and that would become basically replaceable at that point. And then she also has other things where
741 01:36:43.020 ⇒ 01:36:50.219 Robert Tseng: This is less measurable of amplitude. But she’s like, maybe the quality of the traffic that we’re bringing in is just bad.
742 01:36:50.890 ⇒ 01:36:53.959 Robert Tseng: That’s why the drop-off is kind of so steep.
743 01:36:54.540 ⇒ 01:37:02.140 Robert Tseng: So maybe we need better messaging from a marketing perspective on our ads, on our emails, all the different channels where we bring users to our product.
744 01:37:03.030 ⇒ 01:37:03.740 Robert Tseng: …
745 01:37:03.970 ⇒ 01:37:13.989 Robert Tseng: And that’s, like, another workstream for her. She’s trying to experiment with better messaging to bring better users, better fitting users to the product from the get-go, and
746 01:37:13.990 ⇒ 01:37:24.920 Robert Tseng: you know, we have been able to see a marginal difference, a 0.2% increase in the number of signups that are kind of, kind of making all the way through, or, I mean, you could see, like, a… yeah, whatever, like.
747 01:37:24.920 ⇒ 01:37:35.630 Robert Tseng: there’s more people that are signing up now than they were before. So, like, that to her… yeah, these are, like, the different levers that she’s doing. She’s trying to… like, and I would say this is a good stakeholder. She understands
748 01:37:35.680 ⇒ 01:37:47.189 Robert Tseng: what she’s reading and, like, what she needs to focus on in order to, like, move the needle on these things. But sometimes they don’t have ideas, and then we can actually make the recommendations. And I’ve seen enough of these funnels
749 01:37:47.190 ⇒ 01:37:57.400 Robert Tseng: to give her some advice, and I can be like, hey, yeah, actually, you should simplify your onboarding flow. Anything that takes more than three to five steps is probably way too long, and that’s why you’re dropping off a lot.
750 01:37:57.870 ⇒ 01:38:05.940 Robert Tseng: Yeah, like, even what you had saw previously with, the situation here, I guess this is, …
751 01:38:06.630 ⇒ 01:38:10.899 Robert Tseng: Oop, I have to, like… Get rid of this again.
752 01:38:15.810 ⇒ 01:38:18.509 Robert Tseng: Yeah, so, like, from here, if I, …
753 01:38:22.090 ⇒ 01:38:24.750 Robert Tseng: I don’t know, that’s a little rough.
754 01:38:28.420 ⇒ 01:38:38.469 Robert Tseng: Yeah, this was the initial sign-up page, this test doc, whatever, this step here, where they basically are creating an instance for… in their platform.
755 01:38:38.680 ⇒ 01:38:41.179 Robert Tseng: I just want to show you one thing.
756 01:38:42.290 ⇒ 01:38:51.230 Robert Tseng: Great. This step right here. They usually… they used to not have these 3 buttons. They had 3 separate workflows for you to go down.
757 01:38:51.310 ⇒ 01:38:59.069 Robert Tseng: you want to upload an upload API, you’d go down to another thing, and it’d be like a two or three-step process for each one of those.
758 01:38:59.070 ⇒ 01:39:12.690 Robert Tseng: I basically told CD, and I was like, look, if you consolidate all three into a single thing, and, like, one page, give people options, don’t send them down three different paths at this point, I think your conversions will go up.
759 01:39:13.050 ⇒ 01:39:13.940 Robert Tseng: So…
760 01:39:14.390 ⇒ 01:39:21.699 Robert Tseng: I gave her that advice, and yeah, I mean, I think that’s led to a 0.6% lift month over month.
761 01:39:21.870 ⇒ 01:39:29.019 Robert Tseng: Which, you know, it’s kind of… I think it’s fun. This is, to me, the fun part. Like, I’m not, like… it wasn’t…
762 01:39:30.410 ⇒ 01:39:46.530 Robert Tseng: it wasn’t, like, that significant, but it’s cool seeing that, like, I made a recommendation, they implemented it, and, like, cool, like, there’s some marginal improvement here at this step. And so we’re just, like, I’m gonna continue to go back and forth with her month to month. I know that she’s looking at this report every week now.
763 01:39:46.620 ⇒ 01:39:54.270 Robert Tseng: And we’re gonna keep trying to figure out, how do we get 50% up to 60%, how do we get 27% up to 40%, and stuff like that.
764 01:39:54.290 ⇒ 01:40:06.460 Robert Tseng: And that to me is, like, a more ideal partnership between, like, data and a stakeholder, where we’re able to be, a thought partner to them and give them recommendations as well.
765 01:40:06.460 ⇒ 01:40:14.969 Robert Tseng: But yes, digital products is more straightforward than, like, something like Eden, where we’re not really involved in how they design… how they create drugs and whatever, and…
766 01:40:14.970 ⇒ 01:40:20.769 Robert Tseng: you know, that’s a whole different type of, type of client. But for now, like, this is…
767 01:40:21.140 ⇒ 01:40:26.920 Robert Tseng: this is what… this is kind of how… how things are going, like, for… with README.
768 01:40:27.290 ⇒ 01:40:28.080 Robert Tseng: Yeah.
769 01:40:29.430 ⇒ 01:40:35.039 Annie Yu: Yeah, no, this is helpful. And… and one, just one last question is…
770 01:40:35.140 ⇒ 01:40:50.019 Annie Yu: So, obviously, like, over the funnel, it’s… like, each rate is always gonna go down, right? So what’s the expected behavior compared to, like, there is something serious wrong, seriously wrong with this step?
771 01:40:50.480 ⇒ 01:41:00.359 Robert Tseng: Yeah, so, I mean, I guess that’s just kind of where you work on enough products and you get a sense of benchmarking. I would say 50% drop-off is not bad. Like, I would say…
772 01:41:01.230 ⇒ 01:41:18.430 Robert Tseng: I, you know, I would expect it to be lower, maybe, like, 30 to 40%, but, you know, this is already excluding all the bot traffic or whatever, so, you know, to me, this is just saying, like, people who get to your sign-up page, like, they’re not a good fit. If they don’t even complete the sign-up, then they’re getting disqualified off of the first page. Like, it’s not good quality traffic.
773 01:41:19.180 ⇒ 01:41:31.599 Robert Tseng: Yeah, I think that’s, you know, how much can you really, you know, innovate on this part? Like, you want to be clear about what your product is, maybe that helps, the messaging there, but also figuring out how to drive higher quality traffic.
774 01:41:31.790 ⇒ 01:41:45.350 Robert Tseng: figuring out what stages to optimize, that’s not really our job. That should be our stakeholder’s job. I think we should have an eye for it, and there’s a bit of taste here, which is why I think product analytics is not just report building, it’s,
775 01:41:45.390 ⇒ 01:41:51.839 Robert Tseng: we do need to have a point of view on, like, what’s the important step to optimize.
776 01:41:51.900 ⇒ 01:42:07.230 Robert Tseng: So, even with README, I don’t feel like I know enough. Does completing onboarding really matter? Does it, like, how… what’s the correlation between this and people who actually pay? Like, I… I don’t fully know. Like, I… I looked at onboarded users and, like, their plan change.
777 01:42:07.440 ⇒ 01:42:12.670 Robert Tseng: So I’m seeing, okay, well, you know, users that,
778 01:42:13.940 ⇒ 01:42:27.640 Robert Tseng: Finish onboarding, you know, more… more of them are attempting to launch, and they’re, you know, buying, and they’re, like, looking through the different pricing tiers, and then there’s a super steep drop-off where nobody is upgrading.
779 01:42:27.680 ⇒ 01:42:34.090 Robert Tseng: Maybe they’re getting priced out, like, there’s a lot of reasons for it. Maybe I should be measuring time between these steps to better understand it.
780 01:42:34.140 ⇒ 01:42:52.159 Robert Tseng: I think Phoebe wants me to spend more time here. She wants to know, like, okay, well, how do I get more users to, like, actually go and convert into paid users? And I think that’s a, you know, a simple funnel analysis is not going to get me there. I’m going to have to do something a bit more, investigative there. So, yeah.
781 01:42:52.390 ⇒ 01:43:02.150 Robert Tseng: I mean, if this… if you feel like this project is more straightforward to you, it’s something that you feel like you can learn in, and … like, yeah, I… I think we can talk about
782 01:43:02.340 ⇒ 01:43:16.920 Robert Tseng: like, yeah, maybe I should just have you work… work here and… and… and do… work… work on README. Like, maybe it’s a better… better one for you to get started in, and like, you know, there’s not… you’re not gonna be frustrated as much, obviously, as, like, something like a…
783 01:43:16.940 ⇒ 01:43:24.190 Robert Tseng: like a spark plug, and I’m okay with that. So, I don’t know, like, we can talk about that.
784 01:43:24.190 ⇒ 01:43:30.740 Annie Yu: make the judgment there, because I feel like you know more than I do at this point.
785 01:43:30.740 ⇒ 01:43:31.470 Robert Tseng: Okay.
786 01:43:31.960 ⇒ 01:43:32.710 Robert Tseng: …
787 01:43:33.340 ⇒ 01:43:52.980 Robert Tseng: Yeah, I… I mean, to me, all three of those need to be worked on, so I don’t really care which one. Like, if it’s… if it’s gonna be README, and I… and I guess, kinda… I’m leaning more towards that now, after I was walking through this. I wasn’t expecting it from this conversation, but, I think I… you know, let me just sit on that for the day, and…
788 01:43:53.380 ⇒ 01:43:56.149 Robert Tseng: I’ll, I’ll let you know. But, …
789 01:43:56.610 ⇒ 01:44:04.650 Robert Tseng: Okay. I, you know, I’ve… I think we’ve been… we’ve covered a lot of ground. I… Gotta move on.
790 01:44:04.780 ⇒ 01:44:15.509 Robert Tseng: But let me know if you have any other questions, and if, you know, you can doodle on this stuff. All the logins are in 1Pass, you can log into anything if you want to see it. Yeah.
791 01:44:19.160 ⇒ 01:44:25.969 Annie Yu: I… I think… Unless they’re in the shared vault.
792 01:44:27.230 ⇒ 01:44:34.020 Robert Tseng: Oh, right, you’re not added to every client, you’re only added to a few, so I might have to get, Rico to add you to some of these.
793 01:44:35.050 ⇒ 01:44:45.000 Annie Yu: Okay, then is there anything else from, Ellie’s Sparkplug’s side that you would want me to explore? Like, if I can help you in any way?
794 01:44:45.330 ⇒ 01:44:50.169 Robert Tseng: … I would say, …
795 01:44:53.550 ⇒ 01:44:57.449 Robert Tseng: Yeah, give me, give me, like, the rest of the day to decide, like.
796 01:44:58.720 ⇒ 01:45:10.179 Robert Tseng: where I can put you, like, I… if you’re gonna… like I said, I… I don’t… I don’t care which one you help me on, like, I just… I just thought SparkBlug was the one that I literally was spending no time on.
797 01:45:10.450 ⇒ 01:45:11.360 Robert Tseng: But…
798 01:45:11.980 ⇒ 01:45:23.779 Robert Tseng: if you can… if you can work on any of these, and then I… I think that would give me some more time to go and… and do some of the spark plug stuff. So, I…
799 01:45:23.920 ⇒ 01:45:36.249 Robert Tseng: I feel like I’m leaning towards putting you… or, like, to kind of giving you the opportunity to go and, do some of the follow-ups to README, because it seems a bit more straightforward. …
800 01:45:36.490 ⇒ 01:45:40.710 Robert Tseng: But, yeah, I don’t have… I don’t have an answer for you right now.
801 01:45:41.200 ⇒ 01:45:44.069 Annie Yu: Okay, yeah, yeah. No, that sounds good.
802 01:45:44.300 ⇒ 01:45:45.820 Robert Tseng: Okay, cool.
803 01:45:46.090 ⇒ 01:45:48.500 Annie Yu: This is more straightforward, so….
804 01:45:48.500 ⇒ 01:45:49.020 Robert Tseng: Yeah.
805 01:45:49.700 ⇒ 01:45:50.450 Robert Tseng: Okay.
806 01:45:50.700 ⇒ 01:45:53.200 Annie Yu: Okay, thank you very much, Robert.
807 01:45:53.200 ⇒ 01:45:53.820 Robert Tseng: I didn’t spite.