Meeting Title: Great Forge Analytics Role Discussion Date: 2025-09-09 Meeting participants: Robert Tseng, Maria Andrade
WEBVTT
1 00:00:21.660 ⇒ 00:00:22.990 Maria Andrade: Hello!
2 00:00:23.940 ⇒ 00:00:24.950 Robert Tseng: Hi, Maria.
3 00:00:25.360 ⇒ 00:00:26.340 Maria Andrade: How are you?
4 00:00:26.340 ⇒ 00:00:27.560 Robert Tseng: I’m good, how are you?
5 00:00:27.700 ⇒ 00:00:29.849 Maria Andrade: I’m good. Are you in New York?
6 00:00:30.080 ⇒ 00:00:31.420 Robert Tseng: I’m in New York, yeah.
7 00:00:31.620 ⇒ 00:00:33.410 Maria Andrade: Amazing. I’m in Austin.
8 00:00:33.410 ⇒ 00:00:37.049 Robert Tseng: Yeah, I heard from Utah. You were in New York for some time, too, right?
9 00:00:37.210 ⇒ 00:00:38.410 Maria Andrade: Yes, yes.
10 00:00:38.590 ⇒ 00:00:39.290 Robert Tseng: Okay.
11 00:00:40.450 ⇒ 00:00:41.430 Robert Tseng: Cool.
12 00:00:41.430 ⇒ 00:00:42.669 Maria Andrade: Yeah, nice to meet you.
13 00:00:42.670 ⇒ 00:00:49.000 Robert Tseng: Yeah, good to meet you. I guess, Tom caught me up a bit on your background, I think…
14 00:00:49.370 ⇒ 00:00:53.329 Robert Tseng: some of the highlights that he mentioned. He said that you’re…
15 00:00:53.440 ⇒ 00:01:00.179 Robert Tseng: really strong marketing analytics. Obviously we’re kind of with Bumble for a while, and, yeah, I guess…
16 00:01:00.340 ⇒ 00:01:02.819 Robert Tseng: Seems like B2C is…
17 00:01:02.930 ⇒ 00:01:13.840 Robert Tseng: you’re more specialized in B2C, then I guess I’m not really sure how much B2B experience you have. And then, as far as, like, technical skills-wise.
18 00:01:13.950 ⇒ 00:01:18.579 Robert Tseng: you have a lot of experience working with Looker, dbt.
19 00:01:18.690 ⇒ 00:01:24.549 Robert Tseng: And… what else did he mention? Yeah, I guess, like.
20 00:01:24.730 ⇒ 00:01:42.179 Robert Tseng: I’ve… I mean, I guess I’m curious to fill in the blank on, like, kind of your experience and where you sit on a data team. And then I would say I’m also interested, I know that maybe talked to you about our need for product analytics, and kind of tried to
21 00:01:43.070 ⇒ 00:01:49.330 Robert Tseng: Describe how your experience, you feel like, would be able to fill, kind of, our needs on that role.
22 00:01:49.500 ⇒ 00:01:55.790 Robert Tseng: So yeah, I know that was kind of, like, about what was trying to, like, put the pieces in your story together.
23 00:01:56.050 ⇒ 00:02:00.180 Robert Tseng: On this call, and happy to answer any questions about Great Forge.
24 00:02:00.730 ⇒ 00:02:13.369 Maria Andrade: Amazing, okay, yeah, I think you did a great top-level review, honestly. Okay. But yeah, I’ll just give you a little bit of spiel and kind of go into more detail of my background, and then go deeper into product marketing, so…
25 00:02:13.370 ⇒ 00:02:13.780 Robert Tseng: Sure.
26 00:02:13.780 ⇒ 00:02:38.780 Maria Andrade: Yes, started off at Bumble as an intern. I was, like, the 36th employee in college, and during my time there, first was, like, in marketing and social media, and when I was a marketing intern, I really saw that, like, loved Excel, loved automating things, like VBAs, just making systems easier for people, and I worked specifically with community marketing, which is, like, a little bit harder to measure, and so we were trying to find creative ways
27 00:02:38.780 ⇒ 00:02:59.130 Maria Andrade: to help steer the direction the right, so we would do, like, regression analysis based on, places where we had ambassador programs, or we would do, some, like, customer satisfaction analysis and run SurveyMonkey tests. So it was more ambiguous as the company was really small at this stage.
28 00:02:59.420 ⇒ 00:03:24.410 Maria Andrade: I was still, like, in a really large growth. While it’s growing, I then transitioned more into product analytics. I just realized that I really enjoyed analytics, wanted to do more of it, and I was a product analyst for 3 years, maybe a little bit more. In my time there, I spent on Bumble BFF. So if you’re not familiar with Bumble BFF… I am. Yeah, you are. I really am a huge advocate for it.
29 00:03:24.410 ⇒ 00:03:26.399 Maria Andrade: I’m a strong believer in…
30 00:03:26.400 ⇒ 00:03:44.780 Maria Andrade: what it does, but previously it was part of the Bumble Day app, and so what my team did, it was me and my manager as the two analysts on the team, and we just did the whole rollout of launching a standalone app, and so this was a lot of
31 00:03:44.940 ⇒ 00:03:58.709 Maria Andrade: like, a really fun experience, probably one of the career highlights that I have, A, because it worked like a startup in the sense of you got to test, and you got to, like, launch fast, you just got to have more flexibility that you didn’t have on Bumble Date.
32 00:03:58.810 ⇒ 00:04:22.560 Maria Andrade: And so this would mean, like, rolling out product features, like, meeting in person, group chats, like, extending the one-to-one connection, focused there a lot on A-B testing, user group testing, and then did a lot of go-to-market analysis, as well as any, like, cross-contamination. We were trying to make sure that anything that was spent on BFF wasn’t cannibalizing revenue.
33 00:04:22.560 ⇒ 00:04:35.670 Maria Andrade: users or anything to Bumble Day. So, while I was a product analyst, I did a little bit of everything on there. And then in my last year, I spent as a marketing analyst. And so, marketing, it was just something that I was
34 00:04:35.670 ⇒ 00:04:46.360 Maria Andrade: like, had done before, and it kind of was like a Pandora’s box at Bumble, and so my manager had actually requested for me to go back on the team. Basically.
35 00:04:46.580 ⇒ 00:05:07.749 Maria Andrade: during my time at Product Analytics, Bumble had a 30% layoff, and they laid off all the analytics for BFF, and so my trust was to either stay as a product analyst for one of our, like, white labels, or date, or move over to marketing, and so I moved back over to marketing to, like, expand in my Rolodex a little bit. And this is when…
36 00:05:07.750 ⇒ 00:05:27.189 Maria Andrade: we had, like, no marketing data pipeline, like, there was no source that could tell us how much money was being spent where, and we didn’t have data engineers at the time or anything, so that’s where we kind of got, like, our hands dirty. Dbt pipelines, set up APIs, even did some, like, SFTP buckets, worked out a system.
37 00:05:27.190 ⇒ 00:05:36.639 Maria Andrade: That really just was able to get all our money into one place to then use our MMM. And so we had brought our MMM in-house,
38 00:05:36.860 ⇒ 00:05:37.470 Robert Tseng: Hmm.
39 00:05:37.470 ⇒ 00:05:44.109 Maria Andrade: Previously, we were using Forloop for, as our agency for it, but when we were looking at the data approach.
40 00:05:44.110 ⇒ 00:06:08.019 Maria Andrade: we just couldn’t understand the groupings that they did, and there was just, like, a lot of leniency of how they used the data that didn’t make sense for, like, our business need at the time. And so our… it was, like, a senior machine learning data scientist built the model and then onboarded me on it. It was, like, a lightweight, model, and we used Prof… Prophecy, I believe, which is the seasonality package.
41 00:06:08.040 ⇒ 00:06:32.480 Maria Andrade: To really bring it to life, and then after we historically got, 3 years of data, and so that was, like, why the whole data pipeline was a huge part of it. We were able to just make a fully functioning dashboard, and we did monthly refreshes of the MMM, and so then we partnered with all sorts of stakeholders, from influencer to integrated marketing to
42 00:06:32.480 ⇒ 00:06:34.219 Maria Andrade: Paid social to…
43 00:06:34.220 ⇒ 00:06:50.879 Maria Andrade: just, like, paid, like, media buys, and we’re able to say, like, what’s working, what’s not, what’s the CPR, look at the ad stock, and kind of were just able to make more, strategic investment decisions. We also back-tested a lot of what the MMM was saying, because
44 00:06:51.200 ⇒ 00:07:07.020 Maria Andrade: as data people, we all know that, like, you can’t take one source as full truth, like, and we tried to triangulate it with a lot of, like, A-B tests, or, investment blackouts, or withholding tests, and so that was what I spent my last year there really doing.
45 00:07:07.140 ⇒ 00:07:27.759 Maria Andrade: And then they had another 30% layoffs in end of June, early July, which impacted all the marketing analytics team, unfortunately. But yeah, that’s basically my time there, and then what I’m doing right now is kind of just exploring for options, trying to get some AI certifications. I really do see the role of, and I’m sure you feel the same way.
46 00:07:27.760 ⇒ 00:07:34.250 Maria Andrade: AI is going to completely change what a data analyst looks like in 5 years, even more in 10 years, and kind of how
47 00:07:34.250 ⇒ 00:07:42.340 Maria Andrade: To understand it at a deeper level, to use it in, like, a sophisticated manner, not just to use it for… just to use it, you know, for the sake of it.
48 00:07:42.340 ⇒ 00:07:46.099 Maria Andrade: So that’s currently where I am.
49 00:07:46.750 ⇒ 00:07:50.170 Robert Tseng: Okay, cool, yeah, that makes for the walkthrough in more depth.
50 00:07:50.330 ⇒ 00:08:02.220 Robert Tseng: Yeah, I mean, I think I… so… I think that what was interesting about what you said, a few things, yeah, I think definitely, your kind of…
51 00:08:03.320 ⇒ 00:08:04.090 Robert Tseng: just…
52 00:08:04.300 ⇒ 00:08:12.349 Robert Tseng: generalist analyst skill set, and just being able to help bubble from an early stage, kind of going through, like, kind of the early stages of different
53 00:08:13.280 ⇒ 00:08:15.990 Robert Tseng: Degrees of maturity in their analytics sport.
54 00:08:16.180 ⇒ 00:08:21.100 Robert Tseng: I think that’s interesting to me, because that’s largely what we do with our clients.
55 00:08:21.300 ⇒ 00:08:25.490 Robert Tseng: Clients we work with, they’re kind of in this growth stage, I would say, probably, like.
56 00:08:25.550 ⇒ 00:08:33.240 Robert Tseng: Maybe 10 million a year in revenue when we meet them, and then, you know, our goal is just to grow with them until they, you know, go to $100 million.
57 00:08:33.299 ⇒ 00:08:42.419 Robert Tseng: 100 billion plus. And so, there’s just, like, some sort of problems that when… that you encounter when you’re kind of, at that stage.
58 00:08:42.419 ⇒ 00:08:59.659 Robert Tseng: What attribution is a big one, kind of like, it’s not just, like, campaign management across, like, a few channels, but now they’re really trying to go really broadly, maybe marketing budgets have increased, and then also wanting to, like, understand the complexity of, like, how do these different channels, like, impact each other, so…
59 00:08:59.660 ⇒ 00:09:15.829 Robert Tseng: Maybe it’s not all the way to maybe it’s just kind of multi-touch attribution to some extent at first, but kind of, like, getting some of that scaffolding set up is a really common request for a lot of our clients. So, yeah, I think that’s kind of one thing I’ll call out on the marketing side.
60 00:09:15.830 ⇒ 00:09:22.729 Robert Tseng: The product analytics side, yeah, you know, for… we don’t work with too many app companies, to be honest, like, I think
61 00:09:22.940 ⇒ 00:09:39.990 Robert Tseng: Or maybe we just haven’t been really good at it, because we haven’t had somebody who’s really good at that. So, I just think about all the apps that we’ve worked with the past year, we haven’t really retained them beyond, like, 2 or 3 months, to be honest. So, I think maybe it’s just, like, a lack of expertise on our side, and, like, not really knowing, like, how to
62 00:09:39.990 ⇒ 00:09:48.930 Robert Tseng: continue to run with those clients, but it does seem like the app user lifecycle is a lot more up and down than
63 00:09:49.100 ⇒ 00:10:06.440 Robert Tseng: you know, traditional B2B SaaS, or CPG. So, I just think that, like, for B2C products, like, consumer behavior is changing all the time, like, best practices or whatever, from, like, a few years ago, like, you know, sometimes consumer taste just, like, changes. There’s just, like, all these confounding things that you could just…
64 00:10:06.440 ⇒ 00:10:20.330 Robert Tseng: spent, like, a long time digging into, which seems like you did have a lot of opportunity through that level, which is… which is great, because, I think maybe we lack some of that breadth on our team, because we have, like, more… mostly, like, a team of specialists right now.
65 00:10:20.560 ⇒ 00:10:28.069 Robert Tseng: So I think that’s kind of my… those are my initial reactions to kind of, like, what you’ve described.
66 00:10:29.490 ⇒ 00:10:36.780 Robert Tseng: I think what could be helpful, and you let me know if this is interesting for me to talk through, I like to just be, like, kind of…
67 00:10:36.890 ⇒ 00:10:54.300 Robert Tseng: talk about some of our specific needs, kind of… I’m not gonna be going off of the job description. I think that it is what it is. I think you get… you get the whole gist of, like, what we’re trying to do there, but, yeah, I think for this hire specifically,
68 00:10:54.660 ⇒ 00:10:59.870 Robert Tseng: I’m really looking for somebody who can kind of come alongside me and
69 00:11:00.140 ⇒ 00:11:05.649 Robert Tseng: Really, like, kind of more roadmap, like, what the next
70 00:11:07.890 ⇒ 00:11:23.990 Robert Tseng: 6 months to a year look like for, for a client. So, kind of like an analytics architect, if you want… a data architect, if you want to think about it that way. So, I think data engineering, analytics engineering, we’re pretty solid on, so we don’t really need you to come in and do dbt modeling, whatever, like, we have a team to do that.
71 00:11:24.000 ⇒ 00:11:34.610 Robert Tseng: And then on the analyst side, we already… we hired… we just hired, we just hired somebody who kind of is kind of filling in the gaps on doing BI and, some of these ad hoc work.
72 00:11:34.620 ⇒ 00:11:38.670 Robert Tseng: But I think we’re still missing somebody who can really, like.
73 00:11:38.760 ⇒ 00:11:54.800 Robert Tseng: get the lay of the land of a client, and then within, like, their first month, be able to put a… put out a six to, you know, six-month roadmap for a client, and so that they would be, you know, that’s, you know, compelling enough that they would be able to extend, to, like, you know, extend a contract with us, so…
74 00:11:54.870 ⇒ 00:12:07.680 Robert Tseng: I can… maybe if it’s helpful, I can walk through, kind of, how we typically structure our engagements, and then we can maybe talk through a couple, like, a hypo of, like, what that would look like, after… at a certain point.
75 00:12:08.530 ⇒ 00:12:10.149 Maria Andrade: Yeah, I would absolutely love it, yeah.
76 00:12:10.150 ⇒ 00:12:12.020 Robert Tseng: Okay, cool.
77 00:12:12.050 ⇒ 00:12:31.220 Robert Tseng: So yeah, let’s kind of, go with… so recently, we’ve been getting a lot of traction on the, like, telehealth side. There have been a lot of, like, B2C kind of companies in that space, so both on the products, we work with, like, a GLP-1 supplements company, kind of like a hint her competitor, and then we also work with, like, mental health, telehealth, so, like, that talk space, or,
78 00:12:32.920 ⇒ 00:12:33.450 Maria Andrade: health, or.
79 00:12:33.450 ⇒ 00:12:40.179 Robert Tseng: Yeah, yeah, oh yeah, exactly. It’s kind of my competitor. So, yeah, so I think for both of those situations,
80 00:12:40.610 ⇒ 00:12:45.360 Robert Tseng: Yeah, like, you know, the first month maybe has just been, like, okay, they,
81 00:12:45.490 ⇒ 00:12:56.770 Robert Tseng: Actually, they are separate businesses, so let me just pick one. So let’s just pick the one that’s more like a GOP1 supplement company. We started with them, they originally kind of started working with us because they wanted to
82 00:12:56.770 ⇒ 00:13:06.759 Robert Tseng: they were kind of looking at it from a product analytics perspective. Actually, they had mixed panel, and it was really much just like that. Come in here and tell us, like.
83 00:13:06.910 ⇒ 00:13:23.530 Robert Tseng: how do we drive more, like, users through, like, our funnel, so they can, you know, book… book more, schedule more bookings with our, with our doctors and get approvals so they can go and purchase products? This is a bit… it’s, it’s not a direct, D2C,
84 00:13:23.770 ⇒ 00:13:31.439 Robert Tseng: purchase, funnel, like, they need to go through, like, this… they actually have to go through, like, an appointment before they can actually purchase product.
85 00:13:31.690 ⇒ 00:13:47.010 Robert Tseng: So it was a bit more of a complex… more of a complicated funnel, compared to other, like, consumer, like, DDC consumer companies I’ve worked with, because it was, like, both, like, a booking system that you had to go and, like, actually schedule with providers, and then… then there was a checkout flow after that.
86 00:13:47.650 ⇒ 00:13:55.839 Robert Tseng: And so, we kind of went in there, we stood up some of that in the first month, realized that there’s actually a lot more gaps, and we couldn’t actually,
87 00:13:56.200 ⇒ 00:14:06.520 Robert Tseng: you know, turns out, like, the booking… the bookings data was, like, really unreliable. They weren’t capturing a lot of the stuff that was coming through the intakes, because it was with, like, a third-party provider. It wasn’t, like, a native, like.
88 00:14:06.620 ⇒ 00:14:21.620 Robert Tseng: system that they had set up. And then, like, their e-comm, like, EMS, or, order management OMS system, was kind of, like, clunky as well. Like, there’s just a lot of… turns out, like, we built out this funnel, but there’s just so many gaps in the data along the way.
89 00:14:21.930 ⇒ 00:14:44.390 Robert Tseng: And so, within the first month, we were able to basically deliver, like, that initial report of, like, okay, this is what, like, a more robust funnel report would look like, but then also here are all the recommendations for what you need for data cleanliness that you’re not there yet, so I wouldn’t really trust you. And, that… that ended up helping us to sell into, like, this full data engineering work that took another 3 months.
90 00:14:44.410 ⇒ 00:15:00.549 Robert Tseng: And so, I guess I’m curious, like, you know, if you were to kind of put yourselves back in the shoes, you know, first month discovery, you’re given something limited scope, of like, hey, like, we just want you to come in and basically help us drive more conversions and drive more users top of funnel.
91 00:15:00.550 ⇒ 00:15:10.889 Robert Tseng: What would you say are the opportunities that you would be looking for, to, like, uncover in that first month as well, to be able to build out, like, a 3-6 month?
92 00:15:11.390 ⇒ 00:15:19.349 Maria Andrade: Absolutely, and not to, like, toot your own horn or anything, but I feel like you did everything extremely well. There was certain instances at Bumble when
93 00:15:19.380 ⇒ 00:15:21.940 Maria Andrade: Our data was just, like, completely…
94 00:15:21.940 ⇒ 00:15:41.390 Maria Andrade: wacky, like, we were definitely building the ship and flying it at the same time, or building a plane and flying it at the same time, which is always how those companies work. So I think, I mean, initially what you did was probably, like, a funnel analysis, just to analyze it, and then look at those points, but during that process, you were able to see, like, obviously, we can’t trust the data, and this is how
95 00:15:41.390 ⇒ 00:15:48.810 Maria Andrade: you would go about and do the data engineering work to really figure out where those gaps are. That being said,
96 00:15:49.390 ⇒ 00:16:02.419 Maria Andrade: I feel like whenever you would have to, like, sell those points and then do a comparison analysis at the end of it, just to be, like, this is the pick-me-up that we saw from there. I’m trying to compare it to one of my experiences, like, I’ve been born
97 00:16:02.420 ⇒ 00:16:06.219 Maria Andrade: I remember one time, they were like, hey, would you…
98 00:16:06.220 ⇒ 00:16:13.430 Maria Andrade: it was a request from Whitney herself, and she said, can you tell me how many Bumble babies are made? And I’m just like, well, obviously, no.
99 00:16:13.430 ⇒ 00:16:28.810 Maria Andrade: Yeah, we don’t really track that. And so, what I did was just look at conversations that had, like, two replies back and forth, and either shared a social media handle, Instagram, or phone number. And so, kind of able to…
100 00:16:28.880 ⇒ 00:16:39.099 Maria Andrade: give some sort of response to that. But going back to your example, figuring out where the gaps are in the tracking, fix the tracking, especially if it comes to MixedPanel and, like.
101 00:16:39.320 ⇒ 00:16:57.769 Maria Andrade: have one of the data engineers go through every single step, and then, like, show them that front. And the second part, I would look at any APIs that we’re able to get from that third-party user. And so, if the API for wherever the checkout happens, and see how reliable that data is… Let’s say there’s no APIs. There’s no APIs.
102 00:16:57.770 ⇒ 00:16:58.090 Robert Tseng: Cookie.
103 00:16:58.090 ⇒ 00:17:01.229 Maria Andrade: And so how that data is just, like, exists.
104 00:17:01.770 ⇒ 00:17:04.410 Robert Tseng: Yeah, it’s just webhooks, yeah.
105 00:17:04.410 ⇒ 00:17:15.189 Maria Andrade: Yeah, and it’s just… so just to complete… they, like, go through the process of making an appointment, and then you just have, like, another, like, a binary that says that they made a purchase or not.
106 00:17:15.190 ⇒ 00:17:15.829 Robert Tseng: Yeah.
107 00:17:16.150 ⇒ 00:17:21.830 Maria Andrade: Okay, and so you’re able to see that they made a purchase, but do we trust the data?
108 00:17:22.690 ⇒ 00:17:36.259 Robert Tseng: Well, yeah, because it’s not an API, it’s kind of just that the third-party provider just… just pushes webhooks, and it maybe, you know, it doesn’t come… it just comes in batches, and so it’s not really, like…
109 00:17:36.530 ⇒ 00:17:38.079 Robert Tseng: It’s not as reliable as you think.
110 00:17:38.080 ⇒ 00:17:39.079 Maria Andrade: Yeah, it’s like…
111 00:17:39.080 ⇒ 00:17:39.720 Robert Tseng: introduction.
112 00:17:39.720 ⇒ 00:17:44.260 Maria Andrade: can do, right? It’s… and so then, I think I would also put it at…
113 00:17:44.540 ⇒ 00:18:04.359 Maria Andrade: kind of, like, a… not a maturity curve, but then also talk to the client and be like, okay, this is where we’re confident with the data until this point of, like, booking the appointment. We know that the appointment happened, and we can… we’re also confident in that. And the purchasing point, if that’s, like, where they want the conversions to make, then what I would do after we’re all clean with the data.
114 00:18:04.360 ⇒ 00:18:06.860 Maria Andrade: Is go back to, like, the last…
115 00:18:06.860 ⇒ 00:18:11.329 Maria Andrade: Not the low-hanging fruit, but the people who are already all the way down the funnel, and…
116 00:18:11.640 ⇒ 00:18:15.939 Maria Andrade: Maybe we do, like, a pop-up screen, like, survey test, and be like.
117 00:18:16.420 ⇒ 00:18:39.579 Maria Andrade: did you make a purchase, yes or no, like, and have reviews on there of what type of appointments that they’re making, or what type of, medication they’re being prescribed, to see if maybe there’s, like, a higher rate… higher rate of, like, GLP-1s versus, like, anxiety medication, or depending on, like, hair loss, if it’s, like, a him-and-hers competitor, and kind of mix
118 00:18:39.580 ⇒ 00:18:42.750 Maria Andrade: If the goal is to bring up the sales.
119 00:18:43.610 ⇒ 00:18:57.849 Maria Andrade: To see which of the appointments is causing the most conversions, and then also make more digital ads, maybe targeting people who would buy those prescriptions, so, like, look at the demographics of that, as well as looking at the ones that have the lowest
120 00:18:57.920 ⇒ 00:19:08.950 Maria Andrade: conversion rate, and then kind of assess the strategy there. Like, are we trying to increase these, or is this something that we know people don’t feel comfortable purchasing online, and if they don’t feel purchasing… feel confident…
121 00:19:08.950 ⇒ 00:19:20.640 Maria Andrade: confident purchasing these online, or maybe through the service, what can we get there to make them more secure? So maybe there’s that little bit of, like, a research analytics or research function there that speaks more to the consumers to figure out
122 00:19:21.250 ⇒ 00:19:27.770 Maria Andrade: What is going on, after, like, we get everything we can from the data that’s provided.
123 00:19:28.160 ⇒ 00:19:35.329 Robert Tseng: Okay, yeah, no, I think that’s an interesting kind of, like, point to double-thick on. I think there’s a couple nuances, I mean, on the health side.
124 00:19:35.570 ⇒ 00:19:47.909 Robert Tseng: One, like, matching, users to, like, their products, to, like, the product, if they choose medication, you know, before they check out, or, like, it’s actually… they can’t… you can’t do that.
125 00:19:47.910 ⇒ 00:19:48.390 Maria Andrade: Okay.
126 00:19:48.390 ⇒ 00:20:03.840 Robert Tseng: violation of the HI. So it’s, like, there’s… there are some limitations on the health side that we have to, like, keep in mind, but as far as, like, I think I like the idea of injecting something to be able to ask the user for more information, especially if we can’t get it daily from the… from the system.
127 00:20:03.840 ⇒ 00:20:11.430 Robert Tseng: And, yeah, I think that’s a really important partnership to build with, like, the team, especially, you know.
128 00:20:11.470 ⇒ 00:20:21.120 Robert Tseng: client may have, like, CRO or marketing specialist, like, and it’s often, like, kind of where we get, like, I think that’s… that’s… that’s oftentimes where we get stuck, because,
129 00:20:21.360 ⇒ 00:20:30.140 Robert Tseng: Yeah, like, ultimately, we’re not the ones that are kind of running the campaigns, and we’re kind of working with some other, like, third party, probably, and, yeah, there’s, like, this back and forth of, like.
130 00:20:30.170 ⇒ 00:20:45.799 Robert Tseng: they want to just run, you know, media buyers just want to, like, keep running ads, whatever. They don’t really want to create more friction, because if they create more friction, then they think that conversions will go down. That’s going to make them look bad. And so, I think what you’ve described ends up being, like, perhaps that’s a good idea, but then…
131 00:20:45.800 ⇒ 00:20:58.899 Robert Tseng: just by adding an extra step, or, like, kind of creating something in the, in the, you know, purchase funnel, you know, I can imagine, like, the, this type of client be, like, would challenge that, and you, you, that would be a, that’d be a…
132 00:20:58.900 ⇒ 00:21:07.580 Robert Tseng: like, that’s probably, like, a tense relationship that, you know, you would have to… that you would have to navigate. So, yeah, I’m not saying there’s a right or a right answer.
133 00:21:07.580 ⇒ 00:21:09.210 Maria Andrade: No, no, for sure. I think it was…
134 00:21:09.210 ⇒ 00:21:10.020 Robert Tseng: Calling and calling.
135 00:21:10.020 ⇒ 00:21:17.399 Maria Andrade: Yeah, I think also, like, just working with clients like that, you have to be so aware of, like, the tools that they use, like, for
136 00:21:17.400 ⇒ 00:21:37.699 Maria Andrade: about Bumble, like, we had our in-house CRM team, we didn’t use, like, Braids or anything, and so we could do, like, a pop-add survey, like, post-purchase pretty easily, because, like, our system was set up like that. And so, yeah, it’s just deciding what’s important to the client. We don’t have to add anything that will stop them before the purchasing point, but maybe after.
137 00:21:37.700 ⇒ 00:21:50.739 Maria Andrade: And yeah, I think it’s just having conversations of, like, what do you have available? What levers can we pull, and do you feel comfortable pulling? What levers do you have that you don’t feel comfortable pulling that maybe we can persuade you to, if, like, the risk is…
138 00:21:50.990 ⇒ 00:22:01.689 Maria Andrade: big enough to take, because I do think that is a portion of it. Like, at Bumble, we also experience this, that a lot of people want very concrete statistical results, but don’t have the time or space.
139 00:22:01.690 ⇒ 00:22:02.090 Robert Tseng: Yeah.
140 00:22:02.460 ⇒ 00:22:17.559 Maria Andrade: And, like, you just have to be like, we can do this, and it can be directional, and I don’t think either of us will necessarily be satisfied with the answer that it gives us, but we could still, again, directional and make choices from this. Or we could just spend, like, a little bit more time, maybe we do, like.
141 00:22:17.870 ⇒ 00:22:29.140 Maria Andrade: a month, or I guess it also has to depend on, like, the TAM and, like, what type of, like, statistical audience we’re trying to reach, but I think it’s all about having conversations and making sure that everybody
142 00:22:29.320 ⇒ 00:22:39.409 Maria Andrade: Is aligned on the vision and, like, feels confident in the method, because that’s something that always… like, not everybody’s gonna be happy, but there is gonna have to be some give and take on either side.
143 00:22:39.740 ⇒ 00:22:58.320 Robert Tseng: Okay, yeah, no, I think that’s, that’s totally fair. So, I think we can just pause there on that example. So, yeah, I think, I think what we kind of walked through was kind of the scenario where, yeah, I mean, I would kind of want somebody, like, there to kind of make those decisions, and…
144 00:22:58.420 ⇒ 00:23:16.350 Robert Tseng: like, to kind of be, like, a thought partner to a client. So, maybe you, you know, I think the way that you thought about the situation was great, and I think we anticipated, like, kind of, like, some of the tensions and things that we have to… we have to be, like, aware of that the client come back with.
145 00:23:16.370 ⇒ 00:23:21.430 Robert Tseng: But yeah, I do think that ultimately it’s about, like, trying to drive
146 00:23:21.890 ⇒ 00:23:35.530 Robert Tseng: Well, using what you’ve seen, like, works from, like, a data team, like, what… what could we actually do to push, like, proactively push recommendations back to the client in a way that, yeah, like, maybe, like, you know, obviously.
147 00:23:35.530 ⇒ 00:23:41.910 Robert Tseng: if they were to adopt it, you know, we can convince them that… and it will actually have a bigger impact, right? Because I think
148 00:23:41.910 ⇒ 00:24:05.269 Robert Tseng: the problem with having a team of just specialists and data engineers, right now is that we’re just pretty much just reactive to what the client wants, or they’ll ask a bunch of things, ask for a bunch of things. If that ends up filling our roadmap. We don’t really have enough people that are kind of at this position to be able to kind of push back and set our own agenda. So, I think that’s kind of, maybe, like, a, you know, very short sneak peek into
149 00:24:05.280 ⇒ 00:24:20.579 Robert Tseng: kind of the problems that I think about across our clients all the time. Yeah, I think, like, you know, you’ll… you know, and I don’t know if that’s interesting. Maybe you wanted… if you wanted to be more of, like, an IC person knocking out, like, engineering work, like, maybe this isn’t really the.
150 00:24:20.580 ⇒ 00:24:21.600 Maria Andrade: Yeah, no, I agree.
151 00:24:21.600 ⇒ 00:24:22.469 Robert Tseng: not now.
152 00:24:22.730 ⇒ 00:24:26.850 Maria Andrade: That is not, necessarily what I see my career going in as at all.
153 00:24:26.850 ⇒ 00:24:27.310 Robert Tseng: Yeah.
154 00:24:27.310 ⇒ 00:24:43.020 Maria Andrade: I definitely, AAC, as I mentioned, AI changing analytics work so much that I definitely want to be more of, like, a strategy, like, operations kind of thought partner in that sense, and that’s kind of what I was doing at Bumble towards the end, working a lot with, like, the different channel owners and seeing how we.
155 00:24:43.020 ⇒ 00:24:43.560 Robert Tseng: Yeah.
156 00:24:43.560 ⇒ 00:25:02.989 Maria Andrade: make their own, like, for influencers, an example, it’s like, we were spending millions of dollars on, like, four huge influencers, right? But that’s all their budget for the semester, and, like, ROI was bad, like, ad stock was decent compared to, like, a Google AdWords, just talking about, like, MMM measurements and seeing those changes, and then we’re like, okay, what if we…
157 00:25:03.620 ⇒ 00:25:10.390 Maria Andrade: Give that same budget to 200 influencers, and they’re micro-influencers, and just assessing how that impacts strategy.
158 00:25:10.690 ⇒ 00:25:18.920 Maria Andrade: I think those are all just… you have to think creatively and make sure nobody’s upset, especially when working with partners like that.
159 00:25:19.260 ⇒ 00:25:19.810 Robert Tseng: Yup.
160 00:25:20.940 ⇒ 00:25:26.620 Robert Tseng: Okay. Yeah, no, I mean, that’s… that’s totally… I mean, I think that’s great. I think, your…
161 00:25:26.810 ⇒ 00:25:33.540 Robert Tseng: I guess what I… am interested in doing, and just… I’m just, like, thinking out loud,
162 00:25:35.410 ⇒ 00:25:37.889 Robert Tseng: Yeah, it does seem like you can, like.
163 00:25:39.500 ⇒ 00:25:42.670 Robert Tseng: I’m trying to think of how it constructions.
164 00:25:47.840 ⇒ 00:25:57.989 Robert Tseng: Yeah, I think… give me a couple days to kind of just, like, discuss with UTOM, but, like, just preliminary thoughts, it may look a little bit different by the time I send you a follow-up, but, like.
165 00:25:59.210 ⇒ 00:26:07.450 Robert Tseng: I think being able to have you kind of advise on, like, a client first would probably be a good
166 00:26:07.580 ⇒ 00:26:12.300 Robert Tseng: Good way for you to kind of just get a sense of, like, what…
167 00:26:13.570 ⇒ 00:26:19.089 Robert Tseng: Being a thought partner to our, like, clients would be… would look like.
168 00:26:19.370 ⇒ 00:26:29.999 Robert Tseng: Yeah, like, I really don’t need you to be, like, running ad hoc analyses to do engineering work, at least not right now. I would… I do want to try to, like.
169 00:26:30.400 ⇒ 00:26:42.629 Robert Tseng: explore this a bit more to see, like, yeah, basically, if I can plug you in as, like, that head of data on, like, two or three clients, like, that would be ideal. But, like, let’s just kind of start with one.
170 00:26:42.760 ⇒ 00:26:48.059 Robert Tseng: And yeah, like, the objective… the scope is really just to
171 00:26:48.590 ⇒ 00:26:58.129 Robert Tseng: kind of work with marketing partners, specifically, like, VP of Marketing, CMO, like, those are typically people that we’re working with, and, kind of…
172 00:26:59.110 ⇒ 00:27:01.759 Robert Tseng: Understand their roadmap, and…
173 00:27:01.870 ⇒ 00:27:15.449 Robert Tseng: You know, their team asks for a lot of different things, so, like, from within our team, we’d be able to tell you the types of… we’d be able to show you the roadmap of the things that we’re working on to support them. But then, you know, hopefully you have a point of view as well on, like.
174 00:27:15.490 ⇒ 00:27:32.210 Robert Tseng: yeah, like, are… is this the right… are we… are we prioritizing things the right way? Like, are… what’s… what are the other things that we should push back on and, to, like, actually, build something that’s complementary to what the stakeholders really need?
175 00:27:32.340 ⇒ 00:27:33.570 Robert Tseng: Absolutely.
176 00:27:33.570 ⇒ 00:27:39.450 Maria Andrade: get bogged down in, like, those ad hoc requests, for sure. And, like, yeah.
177 00:27:40.110 ⇒ 00:27:49.969 Robert Tseng: Okay. If that kind of sounds interesting to you, I would like to… yeah, I mean, I’d like to kind of put something in front of you and see if you’d be interested in working.
178 00:27:49.970 ⇒ 00:28:07.489 Maria Andrade: Yeah, absolutely. I’m interested in the way that you’re describing this role. I think it’s really interesting. I do think, also, just working with so many different types of people in my past, from people who are very data literate to people who have told me that they’re allergic to graphs, like, you just have to find a way to speak to everybody in their own element.
179 00:28:07.770 ⇒ 00:28:19.790 Robert Tseng: Yeah, yeah, I do like… I do feel like you’re… you’re definitely a good… you’re a good communicator, I feel comfortable in, like, you being client-facing, so I think that was something that I was also, like, kind of trying to assess, but…
180 00:28:19.920 ⇒ 00:28:24.710 Robert Tseng: Yeah, I know that may not have been the role that, like, John talked to you about, but…
181 00:28:24.980 ⇒ 00:28:29.329 Robert Tseng: I don’t know, I guess our needs are changing, so, like, I think this seems to be…
182 00:28:29.650 ⇒ 00:28:32.590 Robert Tseng: A possible, way forward.
183 00:28:34.450 ⇒ 00:28:43.270 Robert Tseng: Yeah, let me… let me… just give me a couple days to kind of come back to you with something, but I think this was… this was good. I think a bit different from what I expected coming of the call.
184 00:28:43.820 ⇒ 00:28:51.910 Maria Andrade: Amazing, yeah, I mean, same here. I loved meeting with you, Tom, and just hearing more about the company. I am a big Urban Stems purchaser, and so that’s.
185 00:28:51.910 ⇒ 00:28:52.870 Robert Tseng: Oh, no way, okay.
186 00:28:52.870 ⇒ 00:29:07.080 Maria Andrade: Yeah, yeah, yeah. And so familiar, and I’ve seen, like, obviously the ABC, billboards all over Texas. I mean, we have critters everywhere. So I definitely think you’re hitting companies also at such a good stage of that growth, because
187 00:29:07.080 ⇒ 00:29:14.380 Maria Andrade: it’s more difficult when people are trying to, like, downsize, and you’re trying to figure out where to cut. Right now, it’s, like, how to cut. Those are the fun questions to answer.
188 00:29:14.380 ⇒ 00:29:29.580 Robert Tseng: Yeah, sorry, I know we’re out of time, but if you could go over a minute or two, like, let me just kind of… I have a couple other thoughts. So yeah, I think what you pointed out is great. So, like, we… we were intentionally not doing, like, cost-cutting kind of, like, work anymore. Like, we are targeting
189 00:29:29.600 ⇒ 00:29:37.549 Robert Tseng: And we are… we’re strategically trying to sell to, like, go-to-market leaders, which is why, like, heads of marketing are the people that we want to go after.
190 00:29:37.610 ⇒ 00:29:56.950 Robert Tseng: ABC is a bit of a different… just because, like, our internal champion was, like, the head of ops, but we’re trying to move over to that side of the org, too. So what I would consider to be, like, success in this role, which we can kind of think about, like, what this ends up looking like if, like, longer term, but, yeah, like, if we can basically, like.
191 00:29:57.080 ⇒ 00:30:06.269 Robert Tseng: We are basically the data engineering team for Urban Sems, and we’re the data team for ABC. But, like, that to me is not enough. Like, I want us to be able to…
192 00:30:06.270 ⇒ 00:30:09.649 Robert Tseng: He’s viewed as, like, a…
193 00:30:09.650 ⇒ 00:30:26.539 Robert Tseng: strategic marketing, or, like, I want us to be considered, like, a go-to-market partner as well, because that’ll help me to structure the, engagements to be… I mean, basically to grow these accounts more. And so, the closer that we can get to revenue, and then be able to have a way forward where
194 00:30:26.560 ⇒ 00:30:43.289 Robert Tseng: I’m just thinking, like, 3-6 months from now, you know, for all of our clients, yeah, like, we have our core data engineering and, like, BI work that we do. That’s, like, you know, we run it efficiently, we’re able to basically displace a lot of the existing, kind of, like, data work that they already have.
195 00:30:43.290 ⇒ 00:31:01.739 Robert Tseng: But to be able to be like, yeah, and then if we… if you want to work with us, and, we advise you on, like, marketing analytics, and there’s, like, a very clear, like, UBITDA, like, impact, then, like, we can have, like, we can… we can… we can go a bit higher, on, like, the work that we’re doing. So,
196 00:31:01.820 ⇒ 00:31:17.550 Robert Tseng: I… I know I’ve said a lot of things, and that’s… this is kind of… I’m… that’s what I’m thinking about these days. So, yeah, anyway, like, if that’s… that’s something that you’re interested in, I’d love to kind of keep… keep this conversation going, so…
197 00:31:17.550 ⇒ 00:31:24.510 Maria Andrade: Absolutely, I’d be super interested to, yeah, like, shadow or hop into one of your clients and really understand the process and how the team works.
198 00:31:24.720 ⇒ 00:31:25.060 Robert Tseng: Okay.
199 00:31:25.060 ⇒ 00:31:32.010 Maria Andrade: It does seem like you guys are very, like, data-stacked and have, like, a lot of talent there, which is always my biggest worry, because I feel like.
200 00:31:32.010 ⇒ 00:31:32.490 Robert Tseng: chaotic.
201 00:31:32.490 ⇒ 00:31:37.289 Maria Andrade: They’re just like, oh, we have nothing, give me everything, and so it’s good to know.
202 00:31:37.760 ⇒ 00:31:47.130 Robert Tseng: Okay, yeah. All right, well, thank you for your time, Maria. Yeah, I… if you have any other questions, just, like, let me know, and yeah, we’ll… I’ll be in touch soon.
203 00:31:47.450 ⇒ 00:31:48.859 Maria Andrade: Sounds good. Have a great one.
204 00:31:48.860 ⇒ 00:31:50.350 Robert Tseng: Cool, thanks. Bye.