Meeting Title: Robert x Katie - Product Analytics Date: 2026-04-29 Meeting participants: Katie’s Claap Recorder, Robert Tseng, Greg Stoutenburg, Katie Ellig, louis
WEBVTT
1 00:01:21.140 ⇒ 00:01:22.120 Greg Stoutenburg: There are…
2 00:01:22.700 ⇒ 00:01:23.430 Robert Tseng: Hey, Greg
3 00:01:25.980 ⇒ 00:01:31.549 Robert Tseng: Sorry, I didn’t give you much heads up. I wish I just sent you that message, like, at least 30 minutes ago.
4 00:01:31.990 ⇒ 00:01:32.939 Greg Stoutenburg: Let it rip.
5 00:01:33.490 ⇒ 00:01:45.950 Robert Tseng: Yeah, but I do have a few things pulled up. I can flash, like, an Eden tracking plan if they wanted to see a visual of that. I also have, like, the event data design, so, I think they just probably want to understand, like, how we… how we work.
6 00:01:46.200 ⇒ 00:01:52.960 Robert Tseng: Yeah. Yeah, I mean, I think the, contexts…
7 00:01:53.290 ⇒ 00:01:57.979 Robert Tseng: that I sent you kind of gives you a sense of what their current stack is. They have segments set up.
8 00:02:00.250 ⇒ 00:02:10.580 Robert Tseng: Yeah, and then… I don’t know if you got a chance to click around their product. I think their website changed even in the last time I talked to them, so it’s, like, accounting software.
9 00:02:10.639 ⇒ 00:02:24.710 Greg Stoutenburg: Yeah, yeah, yeah, yeah. I got it. And yeah, and for a particular industry, and the… just, you know, her recorder’s on. Yeah, and so, yes, that’s right. That’s what I saw, especially for, for, for ag. Yeah, and I spent some time as well.
10 00:02:24.710 ⇒ 00:02:28.410 Robert Tseng: Beyond Ag, by the way, so that’s why they kind of changed their site. Yeah.
11 00:02:28.410 ⇒ 00:02:41.010 Greg Stoutenburg: Yeah. Yeah, I was looking at the… I was looking at the initial motion from, where you… where you sign up for the free trial, and where you get hit with the paywall, and thinking through activation for them, so…
12 00:02:41.010 ⇒ 00:02:41.460 Robert Tseng: Perfect.
13 00:02:41.460 ⇒ 00:02:54.539 Greg Stoutenburg: I think I already have some ideas, and then, as well, just, like, it, you know, you’re… yeah, I mean, flying blind without product analytics, so I think, I think getting… getting something in place to really understand
14 00:02:54.640 ⇒ 00:02:57.400 Greg Stoutenburg: Even just where the drop-offs are,
15 00:02:57.440 ⇒ 00:03:08.220 Greg Stoutenburg: you know, I’ve got my gut and preferences and things like that for changes that could be made to that flow, but would really like to just see where they’re at right now, and then start working on what to… what to do to improve.
16 00:03:08.220 ⇒ 00:03:18.909 Greg Stoutenburg: And I think that the potential for what that improvement looks like, especially just in the activation flow that I’m able to see, it could be really significant, and I’m able to… I can compare what I saw from,
17 00:03:19.020 ⇒ 00:03:21.600 Greg Stoutenburg: Oh! Hi, Katie. Missed mid-rant.
18 00:03:21.970 ⇒ 00:03:26.329 Katie Ellig: Hey! I know how that goes.
19 00:03:28.500 ⇒ 00:03:33.060 Robert Tseng: Yeah, we were just getting excited talking about Ambrooks and things we were already seeing, so, yeah.
20 00:03:33.060 ⇒ 00:03:40.659 Katie Ellig: Yep. Amazing. I love, I love a good, a good rant about, improvement opportunities. Yeah, yeah.
21 00:03:40.660 ⇒ 00:03:44.989 Greg Stoutenburg: Yeah. It’s the best kind of ranch is just show up and find somebody else midstream on, right?
22 00:03:45.130 ⇒ 00:03:50.810 Katie Ellig: Absolutely. Yeah, well, great to meet you, Greg, and good to see you, Rob.
23 00:03:50.810 ⇒ 00:03:52.829 Robert Tseng: Yeah, there’s a few.
24 00:03:52.940 ⇒ 00:03:54.190 Robert Tseng: Hey Louis, good to meet you.
25 00:03:54.190 ⇒ 00:03:55.229 Katie Ellig: Yeah.
26 00:03:55.230 ⇒ 00:03:55.840 louis: run.
27 00:03:55.840 ⇒ 00:03:56.550 Katie Ellig: Good.
28 00:03:57.600 ⇒ 00:04:03.010 Katie Ellig: Maybe if we want to do a quick… a quick round of intros before we get into things, we’d love that.
29 00:04:04.220 ⇒ 00:04:07.050 Robert Tseng: I can kick it off,
30 00:04:07.270 ⇒ 00:04:17.950 Robert Tseng: Hey Louis, I run Brainforge. We’re a data and ad consultancy. We basically help organizations with context engineering, getting them ready to
31 00:04:18.050 ⇒ 00:04:31.900 Robert Tseng: build their data stack and plug AI tools to kind of help them grow. I think product analytics is one of our kind of pillars for, like, what we do, and a crucial part to what we consider context.
32 00:04:32.190 ⇒ 00:04:41.469 Robert Tseng: And Greg is our product analytics lead. And so, just wanted him on this call. You know, Kate and I have chatted a bit, you know, more informally about, kind of, Ambrook.
33 00:04:41.470 ⇒ 00:04:54.000 Robert Tseng: And really incredible growth that your team has been seeing, and know that maybe, hey, there’s an opportunity for us to at least maybe advise on, like, kind of where you’re at in terms of,
34 00:04:54.020 ⇒ 00:05:07.500 Robert Tseng: you know, getting… getting some of the initial, product analytics infrastructure set up. So yeah, just looking to be helpful on this call, better learn the needs, but also kind of hand it off to Greg to introduce himself, his background, and, yeah.
35 00:05:08.260 ⇒ 00:05:14.679 Greg Stoutenburg: Yeah, yep, yeah, thanks for having me on. I’m Greg Sautenberg, I’m a product analytics lead at Brainforge.
36 00:05:14.700 ⇒ 00:05:23.200 Greg Stoutenburg: I come from a background of… I mean, the short version of the story is, once I was finished being a philosophy professor, I became a product growth strategist.
37 00:05:23.200 ⇒ 00:05:38.040 Greg Stoutenburg: at tech companies for the last 5 years, and so, yeah, I mean, where I really made my first sort of big intro to tech was working specifically on product analytics, and especially on user activation and retention and conversion to paid.
38 00:05:38.040 ⇒ 00:05:55.089 Greg Stoutenburg: At, at, at SaaS companies. So, that’s what I’ve been up to for the last little while, and, since coming to Brainforge, it’s been awesome that I’m able to do similar work just across so many, so many clients. So, Robert’s given me some of the background, and, you know, as well, you know, Katie, we exchanged some emails.
39 00:05:55.090 ⇒ 00:06:02.770 Greg Stoutenburg: On what you’re up to and what you’re looking for. So, excited to, excited to dive in with you and explore where the possibilities could take us.
40 00:06:04.340 ⇒ 00:06:12.290 Katie Ellig: Awesome, yeah, good to meet you, Greg. And, I’m Katie, I lead Customer Success at Ambrook, so, like.
41 00:06:12.380 ⇒ 00:06:27.529 Katie Ellig: Rob might have shared, we’ve grown a lot over the last two years, which has been really exciting, so gone from, like, 20 customers to 7,000 customers, and, are starting to get to the scale where it’s more and more
42 00:06:27.800 ⇒ 00:06:45.339 Katie Ellig: useful for us to have, data about exactly how customers are, are using the product, and having that be translated into our, you know, Lewis and I were just talking this morning about, our onboarding process and, how we make sure that customers are,
43 00:06:45.560 ⇒ 00:07:01.589 Katie Ellig: set up for success in the different parts of the product that we want them to be using, so, was just interested. Rob and I have chatted more informally, and was interested to learn a little bit more about, like, what you all do specifically for folks in the product analytics,
44 00:07:01.680 ⇒ 00:07:09.420 Katie Ellig: area of things, and yeah, wanted to bring Lewis, to the call as well, so let him introduce himself.
45 00:07:13.130 ⇒ 00:07:14.199 Katie Ellig: You’re on mute, Louis.
46 00:07:15.960 ⇒ 00:07:24.600 louis: Yeah, I’m Louis, I lead the product marketing efforts here at Nbrook. I’m working very closely with pretty much everything marketing, but mostly focused on, sort of.
47 00:07:25.110 ⇒ 00:07:29.749 louis: Yeah, things top of mind are sort of attribution, campaign, content performance,
48 00:07:30.140 ⇒ 00:07:35.969 louis: research, and then, yeah, yeah, anything that kind of touches products, I’m onboarding.
49 00:07:36.110 ⇒ 00:07:47.519 louis: lifecycle perspective. So, yeah, before Emmerich, I was… I’ve been in a number of, kind of, early-stage product marketing roles, mostly in the AI developer tooling space, so I’m quite familiar with
50 00:07:47.620 ⇒ 00:07:51.700 louis: most of the stack, and yeah, I would love to, you know, dive a little bit more into, sort of, what is the…
51 00:07:53.050 ⇒ 00:08:01.929 louis: Sort of what’s… what’s your stack, and sort of what’s… what’s sort of, like, the, sort of the unique insight that you guys bring in terms of, like, the services layer, and, yeah, keep it… keep it pretty conversational.
52 00:08:04.920 ⇒ 00:08:27.420 Robert Tseng: Cool. So, I mean, I’ll probably just kind of set the stage, and then I’ll let kind of Greg do more of the talking. But, I mean, as far as, like, our stack goes, I think we’re pretty, agnostic. We’ve used all these different tools. We’re a systems integrator, basically, like, a premier partner for Amplitude, for Mixpanel. We use PostHog internally, so we kind of understand what all the latest capabilities are from a tooling standpoint.
53 00:08:27.420 ⇒ 00:08:32.430 Robert Tseng: I think we’re very opinionated about, kind of, how we set up these tools, how to not over-engineer it.
54 00:08:32.429 ⇒ 00:08:36.739 Robert Tseng: We don’t believe in tracking everything under the sun and then drawing a needle from a haystack.
55 00:08:36.740 ⇒ 00:08:50.569 Robert Tseng: But we do… I mean, I think we’re a mix of quants and qual, and so I do think that there is a baseline set of measurement that we need to establish, and I think we really are opinionated about setting up key, kind of.
56 00:08:50.570 ⇒ 00:09:14.969 Robert Tseng: milestones, defining things in workflows, and so you’ll… I think if you have any sort of, like, you know, maybe… I’m sure both of you work with engineers, so, like, I think that’s the rigor that we bring to, kind of this type of development. And then Greg has a lot of, kind of, yeah, just taste and experience from just being able to, kind of, work with many different types of SaaS companies, knowing kind of where the key friction points are, and, yeah, so it’s usually kind of a blend of these two things that we type of…
57 00:09:14.970 ⇒ 00:09:34.970 Robert Tseng: we would bring into an engagement. So, at a high-level kind of structure is typically we would make sure that, we define maybe, like, a couple core workflows, just to start, for a client that’s never kind of set up product analytics before. We go end-to-end from tracking to implementation, building out core reporting, so that you can at least, like.
58 00:09:34.970 ⇒ 00:09:45.849 Robert Tseng: test with a basic experiment and be able to know kind of a difference, between what the previous funnel was and afterwards. So, we prefer to do it in kind of these kind of, like, modular,
59 00:09:45.850 ⇒ 00:09:49.879 Robert Tseng: In a modular way, so that you can, you know, if it’s… if you…
60 00:09:49.930 ⇒ 00:10:03.189 Robert Tseng: now that you’ve seen it run through one time, then you can kind of keep iterating on it. If we need to train someone internally, we can do a handoff there. But yeah, I’ll let kind of Greg kind of go a little bit more into detail of, like, what this… what the process actually looks like.
61 00:10:03.740 ⇒ 00:10:28.570 Greg Stoutenburg: Yeah, so what we’d be looking for is, to start with, you know, every company, every product has something that the user is supposed to do with it. And so, we go from the standpoint of what would someone be trying to accomplish in this point in your product? What are they trying to do? So you set that goal, and you kind of work backwards to what the essential steps are leading up to that. So, as an example, one of the first things someone would try to do with Ambrook is just sign up for
62 00:10:28.570 ⇒ 00:10:32.360 Greg Stoutenburg: And get into the product, right? So, then, if that’s the goal.
63 00:10:32.510 ⇒ 00:10:50.049 Greg Stoutenburg: we map it out and work back, what are the incremental steps that someone has to take to get there. Now, we’re not going to look at literally every single thing, as Robert was suggesting, right? Sometimes people get really excited a few years ago about, like, auto-tracking capability of various, product analytics tools, but then they end up just
64 00:10:50.050 ⇒ 00:10:57.279 Greg Stoutenburg: Tracking every single thing, and then, you know, you go and you try to do some analysis, you type in a word that you think is gonna lead to some particular event.
65 00:10:57.280 ⇒ 00:11:16.069 Greg Stoutenburg: you’re wrong, because it was just something that’s picking up on CSS structures on a page. So, we’re going to be more strategic than that, and be more opinionated about the things that are likely to matter. Map it out, instrument those, and then begin tracking it. So, for example, if someone’s goal is to begin to use Ambrook, then we’re going to look at the journey from landing the web page.
66 00:11:16.150 ⇒ 00:11:22.609 Greg Stoutenburg: signing up, Clearing the paywall, landing in the product, and performing some first meaningful action.
67 00:11:23.390 ⇒ 00:11:36.000 Greg Stoutenburg: Which action? We’ll talk with you. We’ll rely on your expertise of your own product, of what someone would be likely to try to do right away. That matters, right? So we measure to that as a goal point, and we call that one workflow.
68 00:11:36.300 ⇒ 00:11:48.550 Greg Stoutenburg: And we’ll do that for other workflows that we see to be likely to be of value that someone could be using your product for. We map out all of those, and we track those. We set goals, and we then have a baseline for experimentation.
69 00:11:48.630 ⇒ 00:11:52.890 Greg Stoutenburg: So, as an example of what that might look like, again, just on that core workflow.
70 00:11:52.890 ⇒ 00:12:12.330 Greg Stoutenburg: Something that we might experiment with is removing that paywall until someone has done something in the application. And then we can compare, right, what does user activation look like when someone’s completing that first valuable action? What does user activation and conversion look like on the early paywall model versus on the later paywall model?
71 00:12:12.440 ⇒ 00:12:18.800 Greg Stoutenburg: And that sort of approach is the one that we’ll take to the core workflows in the product.
72 00:12:19.380 ⇒ 00:12:21.669 Greg Stoutenburg: Should I… should I keep going with that?
73 00:12:21.820 ⇒ 00:12:33.199 Greg Stoutenburg: The, it’s tool agnostic. There are, you know, we’ve mentioned PostHog, Amplitude, Mixpanel. It’s, there are various tools that we’ll be able to accomplish this with.
74 00:12:33.240 ⇒ 00:12:52.900 Greg Stoutenburg: And, but the main thing is just having that opinionated, principled take on what we’ll be measuring. And then the glorious future that it can lead us to right now is, you know, there are a lot of things that you don’t know about what might happen if various changes were made, because we’re not tracking that before and after, as far as user engagement goes.
75 00:12:52.900 ⇒ 00:13:17.249 Greg Stoutenburg: So, like, we can… we can help someone like Lewis knock the ball out of the park by, you know, really being able to find out what are those core engagement levers, where when someone performs some action or some group of action… actions in some period of time, now they get an email, right? Now they… now they ended up in some other, you know, engagement sequence, whether that’s in the product or outside of it. And so, in this way, it can really empower marketing as well.
76 00:13:17.250 ⇒ 00:13:18.859 Greg Stoutenburg: Through the power of product analytics.
77 00:13:21.050 ⇒ 00:13:34.779 Katie Ellig: I’m curious, what, you know, when I joined, you were, talking about some things with, with Rob. I’m curious what you already were looking at or thinking about in the realm of e-book.
78 00:13:35.350 ⇒ 00:13:39.819 Greg Stoutenburg: Yeah, for me, the fir- first big thing, so, I,
79 00:13:39.930 ⇒ 00:13:49.939 Greg Stoutenburg: my… what got me excited about software is… is… is user journeys, and looking at exactly the kind of thing that I’m talking about now. So, for…
80 00:13:49.940 ⇒ 00:14:03.000 Greg Stoutenburg: well, I could give a bunch of background about that, but it’s not super important. When I was at a company called FlowFuse, we, we rolled out a free trial, and we’re very excited about it, like, hey, freemium, freemium, everyone loves freemium, right?
81 00:14:03.050 ⇒ 00:14:20.639 Greg Stoutenburg: But the Stripe paywall early on was just a killer. People did not want to put in any payment information before they used it, and when we took that down, sign-ups went up, like, 20x. And, so, like, things like that are things I’m curious about, right? Like, when it comes to a free trial,
82 00:14:21.050 ⇒ 00:14:36.750 Greg Stoutenburg: I’m curious about what the bounds around your trial should be. Often it’s just a period of time, but there are things that you can experiment with as well, right? Like, maybe your trial goes to 14 days if, someone doesn’t perform
83 00:14:36.870 ⇒ 00:14:43.319 Greg Stoutenburg: any actions, and then at the end of that 14 days, they just get a reminder, like, hey, it looks like you didn’t really use this trial, do you need a week? And, you know.
84 00:14:43.570 ⇒ 00:15:02.799 Greg Stoutenburg: Or, maybe once someone has performed an action for free, we go, hey, great, glad to see you’re getting value from this, you can use this for free for another 7 days, and then you’ll make a decision, right? So there are things you can do around access to the trial, and then around the duration of it, based on features that are being used. Those are some places where I think that,
85 00:15:02.800 ⇒ 00:15:05.259 Greg Stoutenburg: We’d want to experiment with,
86 00:15:05.260 ⇒ 00:15:20.759 Greg Stoutenburg: sort of immediately. Now, I’m hypothesizing, so you said, like, what are some of the thoughts that I have? It could turn out that you measure some of these things, and you go, actually, everything’s fine, don’t mess with those things. Mess with these other things here. But, you know, as far as… as far as the impact of product analytics.
87 00:15:20.860 ⇒ 00:15:32.070 Greg Stoutenburg: what can seem like small strategic changes to early moments in a user journey can turn into really significant improvements in, like, 30-day retention. And it’s your users who are sticking around who pay.
88 00:15:34.150 ⇒ 00:15:34.900 Greg Stoutenburg: Nope.
89 00:15:36.680 ⇒ 00:15:55.279 Katie Ellig: Cool, yeah, and I guess I’m curious for other, like, early-stage companies, let’s say Series A, Series B, what some of the things that, like, what engagements with them have looked like that have been really impactful, and generally where you’re… where you’re starting from there.
90 00:15:56.610 ⇒ 00:16:00.179 Greg Stoutenburg: Yeah, I’ll let Robert speak to the breadth of the business and previous engagements like this.
91 00:16:01.340 ⇒ 00:16:21.310 Robert Tseng: Yeah, sure. So, I think, like, I mean, just kind of going… trying to align with some of the stack that I’ve already heard you kind of set up. So, I mean, it’s great that you already have segment, kind of set up, and you’re already kind of landing that event dream somewhere. And so, yeah, oftentimes at the Series A stage, like, we’re working with companies that have… that have nothing. If anything, they might have
92 00:16:21.420 ⇒ 00:16:35.270 Robert Tseng: yeah, they might have some CDP set up, and they’re starting to stream in data, but it’s not structured in any way, it’s not really kind of, like, tied to any sort of kind of workflows, there’s no, like, business modeling around that. And so I think that’s, you know, getting this
93 00:16:35.950 ⇒ 00:16:37.649 Robert Tseng: Doing this initial…
94 00:16:37.900 ⇒ 00:16:45.390 Robert Tseng: Kind of, like, planning, design kind of phases is super important, just to, like, really empower
95 00:16:45.390 ⇒ 00:16:51.970 Robert Tseng: Like, the business users, who are really the people who understand the product the best, to really be able to start thinking, in this, like, kind of, like.
96 00:16:51.970 ⇒ 00:17:07.960 Robert Tseng: PLG growth way. So I think that’s… that’s really kind of the biggest unlock for us, that, like, yes, we have certain artifacts that we’ll… that we would create with the clients, the tracking plan, event data design, and I can show you kind of, like, snippets of what these exercises look like.
97 00:17:07.960 ⇒ 00:17:24.009 Robert Tseng: And also, like, you know, what does… what does good… what does good look like? I think is oftentimes, like, a big question, because early-stage companies are curious to benchmark their… their performance, against, you know, you know, against other companies.
98 00:17:24.010 ⇒ 00:17:27.350 Robert Tseng: So, like, we’d be able to kind of provide some guidance around all of that.
99 00:17:27.349 ⇒ 00:17:46.170 Robert Tseng: I think then medium-term, I’m taking… I’m thinking, like, you know, past… past the first month, past the first three months, where, like, they’re starting to get some momentum. I think then every organization is different, and what we really enjoy working with early-stage companies is that, we get to learn, you know, you know, for your… for your highest performers, highest, you know.
100 00:17:46.190 ⇒ 00:18:06.170 Robert Tseng: operators, like, they’re all consuming data in different ways, and so we’re actually not married to, like, a particular tool. Like, whether your team goes off and rips and builds a bunch of reports in Mixpanel doesn’t particularly matter. We kind of build things up in an agnostic way, and I kind of alluded to that, before, where from, like, from our kind of perspective as a context engineering firm.
101 00:18:06.460 ⇒ 00:18:11.939 Robert Tseng: We believe that, you know, there’s different types of data, and they all have different half-lives, right? So, like.
102 00:18:11.940 ⇒ 00:18:30.730 Robert Tseng: Product analytics covers a slice of the world, where you see things in moments, like current state snapshots, like you’re able to see at the session level. Product analytics, whether you get it from a tool or, you know, you’ll be able to see, like, you know, session level kind of behavior. You’ll be able to see across different sessions when users come in and out.
103 00:18:30.730 ⇒ 00:18:49.570 Robert Tseng: Over a period of, you know, typically you’re looking at days, weeks, maybe… maybe month over month, perhaps, and then you’re also wanting to measure their journey over, like, a longer period of time, and you’re… and so I think these are all kind of just… it’s just one bucket of data. Eventually, business users start to ask questions, well, I want to start understanding, like, kind of.
104 00:18:49.570 ⇒ 00:18:56.300 Robert Tseng: relationship history with other, accounts, if this is, like, a B2B SaaS platform like yours, where your end user is, like.
105 00:18:56.300 ⇒ 00:18:59.300 Robert Tseng: One of probably multiple users in an account.
106 00:18:59.300 ⇒ 00:19:16.169 Robert Tseng: They’re wanting to bring in, other transaction data, you know, and, just, like, kind of other stuff that’s, like, kind of locked up in different data structures across the organization, and that’s where I think, like, the biggest, kind of the next phase, really kind of unlocks the most.
107 00:19:16.170 ⇒ 00:19:40.539 Robert Tseng: Where, yeah, we start to really blend all this data together, make it accessible, whether you’re surfacing it through, you know, agentic workspaces now, and you’re just, like, plugging into the MCP of these different tools, you’re able to ask questions, and really… and it’s able to pull data quickly from a bunch of different sources all in one place. Or if you actually want to build, like, kind of more fixed reporting, in a different kind of setting, custom internal tooling, or, you know.
108 00:19:40.540 ⇒ 00:19:56.280 Robert Tseng: whatever business intelligence you use currently. So, I think those are just, like, some, you know, examples of, like, what kind of the early, like, kind of the immediate kind of value is, and then, like, kind of midterm, like, you know, three, three months out that we’ve seen with clients that work with us.
109 00:19:58.420 ⇒ 00:20:03.600 Katie Ellig: Well, yeah, before I jump in with other questions, Lewis, anything, on your mind?
110 00:20:07.730 ⇒ 00:20:18.439 louis: No, not too much. I think, at least on the marketing front, I think we actually have a pretty solid handle. We brought in a lot of specialists as well to kind of get quite forensic on, sort of.
111 00:20:19.020 ⇒ 00:20:31.730 louis: on lifecycle analysis, sort of like the event, sort of… event analysis, but I would be probably interested… there’s a level of granularity that we do want to get to, kind of, on the per-user personalization piece. I think, for example, if our PLG
112 00:20:31.970 ⇒ 00:20:47.829 louis: Oftentimes, the buying committee is just one person, so we just sort of need to do… we need to cover a lot of ground there. It’s the person who ends up buying Anbrook is the one who ends up using it. Obviously, we’re sort of moving up market into these higher, you know, these higher ECB segments, there’s a little bit of some nuance there, but
113 00:20:47.980 ⇒ 00:20:53.940 louis: at least for the next 6 to 9 months, we’re going to be very focused. I think for us to be able to kind of get to, like, the…
114 00:20:55.240 ⇒ 00:21:06.049 louis: Yeah, like, per user level personalization, whether that’s lifecycle or onboarding, and being able to signal and kind of build generative content, based on… based on that would be…
115 00:21:06.300 ⇒ 00:21:10.409 louis: Would be definitely a game changer. So, that kind of ties into…
116 00:21:10.660 ⇒ 00:21:17.019 louis: Yeah, onboarding, as Katie mentioned, and then obviously with some marketing considerations as well, but
117 00:21:17.150 ⇒ 00:21:19.370 louis: But yeah, we can chat about that a little bit later.
118 00:21:20.560 ⇒ 00:21:31.120 Robert Tseng: I’m curious if we could ask a bit more about, kind of, what your current capabilities are on the lifecycle side, like, what have you set up that’s working well? Just want to better understand the baseline, for at least for you.
119 00:21:31.820 ⇒ 00:21:42.699 louis: We have… we pipe in some… so we have a good amount of behavioral information… behavioral, information from… in the product that we… we bring back into HubSpot, and that we can send off different. I think it’s still… it’s…
120 00:21:42.880 ⇒ 00:21:46.860 louis: It’s not… it’s not so, so sophisticated, but again, we don’t need
121 00:21:47.040 ⇒ 00:22:01.340 louis: something, like, ultra-sophisticated, because there’s only about 3 or 4 signals that can help us move… move leads, through the funnel. Yeah. I think it’s mostly… it’s mostly around, sort of, there’s a lot more personalization that we want to do. There’s a lot of nuance with our customers that we want to be able to…
122 00:22:01.480 ⇒ 00:22:13.360 louis: To capture that there’s sort of, like, a general contractor that also runs a ranch, or vice versa, and that we want to be able to feed content that’s a little bit more, like, that hits both kind of use cases, and so we,
123 00:22:13.580 ⇒ 00:22:27.669 louis: we’re able to do that cleanly. I think on, like, on the paid and demand gen side, we’re using an agency that’s quite piped in. They brought their own analytics team, and so we have pretty good visibility on spend, on CAC, on channel performance.
124 00:22:27.960 ⇒ 00:22:47.419 louis: I think we… we want to… we’d want to love that level of granularity within the product and the product itself, and we basically would be able to… if we’re able to connect, like, also top of funnel and, like, what we’re seeing on top of funnel, all the way through past… past post sales into product experience, that could be pretty interesting as well, and see if there’s any… any patterns that sort of emerge that would be, like.
125 00:22:48.870 ⇒ 00:23:04.590 louis: we brought in an Ambrook Wallet user that came in through this group of campaigns that was very sort of, like, pain-focused. They have actually a higher activation rate. I mean, that would be something that would be pretty interesting to run some analysis on, but again,
126 00:23:04.840 ⇒ 00:23:10.349 louis: That’s kind of a tall order to… in my experience, to, like, kind of marry those two systems.
127 00:23:10.350 ⇒ 00:23:10.740 Robert Tseng: Sure.
128 00:23:10.740 ⇒ 00:23:14.480 louis: Especially because they kind of run… they’re a little bit different, but yeah,
129 00:23:16.070 ⇒ 00:23:17.619 louis: Does that answer your question a little bit?
130 00:23:18.210 ⇒ 00:23:20.590 Robert Tseng: Yeah, yeah, no, thank you, that’s, that’s, that’s super helpful.
131 00:23:22.430 ⇒ 00:23:28.140 Robert Tseng: And then, Katie, I know I maybe kind of cut you off earlier when I was asking that question. You were going to jump in. Okay.
132 00:23:28.420 ⇒ 00:23:29.880 Robert Tseng: Yeah.
133 00:23:30.410 ⇒ 00:23:43.870 Katie Ellig: Yeah, I was just gonna say, I think, just to bliss on what Louis was saying, I think top of mind for… for me is that, journey for after someone gets into the product, and being able to connect that to
134 00:23:43.870 ⇒ 00:24:03.629 Katie Ellig: how and why they came into the product in the first place, so that we have that really solid loop of, like, is the set of ads, or is, like, this type of user targeting actually leading to the actions that we would expect that group of users to take in the product, would definitely be something that’s interesting for us. I think we do have, like.
135 00:24:04.030 ⇒ 00:24:21.029 Katie Ellig: relatively good signals in terms of, like, once a user is in the product, are they taking these actions? How fast are they taking these actions? That sort of thing, but kind of putting the whole story together, or being able to do, like, hyper…
136 00:24:21.030 ⇒ 00:24:24.919 Katie Ellig: Personalization, could be kind of interesting.
137 00:24:26.810 ⇒ 00:24:40.060 Robert Tseng: So, I guess what I’m hearing is, yeah, definitely, kind of, personalization as far as priorities for now for this team is, like, really focused on, kind of, like, attribution and top of funnel, like, kind of what you can understand about them early on.
138 00:24:40.060 ⇒ 00:24:46.020 Robert Tseng: It seems like, yeah, as far as, like, what they’re… what users are doing once they’re in their product,
139 00:24:46.020 ⇒ 00:25:02.870 Robert Tseng: I guess, like, you have good signals on… on… on that. Seems like, you know, Lewis is already getting some stuff on HubSpot, able to record it, but, like, being able to go deeper on personalization for, especially for early funnel users, I think is… is… seems like it’s a priority for the team.
140 00:25:05.310 ⇒ 00:25:06.719 Katie Ellig: Yep, I think that’s right.
141 00:25:07.240 ⇒ 00:25:07.780 Robert Tseng: Yeah.
142 00:25:11.020 ⇒ 00:25:28.979 Greg Stoutenburg: Yeah, great. Well, I mean, with a well-instrumented product analytics stack, then you’ll be in a very good position to do that, because you’ll see who’s doing what and when, and you can make the right kind of outreach where it’s relevant. You can segment your users in much more granular ways than might be possible otherwise. Well, than is possible otherwise.
143 00:25:28.990 ⇒ 00:25:31.950 Greg Stoutenburg: So, yeah, I think that that’s something that we can deliver.
144 00:25:33.760 ⇒ 00:25:51.300 Katie Ellig: Yeah, I’m curious, what an engagement typically looks like for you guys. I know, Rob, you mentioned you try to focus on, like, very modular, focused deliverables, so yeah, I guess I’m curious, like, timeline scoping, how you usually run that sort of thing.
145 00:25:51.840 ⇒ 00:26:08.950 Robert Tseng: Yeah, I mean, even from this, I think it already narrowed the scope, which is great. I think, definitely trying to focus on early signals, is… yeah, I feel like that’s… that’s quite narrow. We can come back to you with kind of a proposal, just something to, like, we can kind of scratch up and… and we can maybe jump on another call later to go
146 00:26:08.950 ⇒ 00:26:19.369 Robert Tseng: kind of go more… go in more detail. But yeah, I mean, I feel like as far as, like, phasing goes, typically we like to do what’s just kind of a discovery kind of strategy, kind of,
147 00:26:19.440 ⇒ 00:26:24.910 Robert Tseng: Audit, where we just go around, poke around, go in, poke around, really understand the systems that you’re already…
148 00:26:24.950 ⇒ 00:26:35.010 Robert Tseng: they’re using, and… yeah, I think that helps us to kind of define, like, what… what is the… what is the… what is the sequence of events that we’re gonna be… we’re gonna be focused on?
149 00:26:35.010 ⇒ 00:26:39.589 Robert Tseng: Yeah, there’s no, like, tooling selection aside, like, that can all kind of come…
150 00:26:39.590 ⇒ 00:26:57.449 Robert Tseng: down, like, a little bit later as we better understand, like, kind of your team, because I think certain tools are better for certain types of users. So, yeah, I think that’s probably my recommendation for, like, how we could move forward. It’d be a short, tightly scoped, you know, could be less… pretty much, typically we do listen less than 10 hours.
151 00:26:57.470 ⇒ 00:26:59.879 Robert Tseng: And yeah, we could just…
152 00:26:59.990 ⇒ 00:27:15.740 Robert Tseng: you know, once we go in, then we’ll be able to come back to you and be like, okay, do they need engineering support in terms of, like, we actually do the instrumentation and the events? Do we also need reporting support, where we actually, after the events are instrumented, we would build the initial set of reports, and we’re doing more, kind of, like.
153 00:27:15.980 ⇒ 00:27:21.650 Robert Tseng: kind of enablement or coaching… coaching there. So yeah, I think there’s a few different, like.
154 00:27:21.700 ⇒ 00:27:40.090 Robert Tseng: things that we try to figure out within that first discovery audit period as well, to see, like, what type of, like, medium-term support will they need, like, after we kind of hand you something that, you know, you could either just take that and be like, that’s good, we’ll implement it ourselves, or, you know, if you want to lean on our team for more support, we try to
155 00:27:40.090 ⇒ 00:27:53.269 Robert Tseng: We can kind of… we’ve… we’ve pretty much built product analytics functions from scratch, you know, fully turnkey, but also have also kind of more… kind of picked different pieces and… and just done specific roles, with, with teams that have
156 00:27:53.270 ⇒ 00:28:00.149 Robert Tseng: really strong engineering support already. So, those are some of the things that come to mind for, like, what I would like to figure out in a first phase.
157 00:28:02.350 ⇒ 00:28:06.299 Katie Ellig: Yeah, I definitely think, this is, for sure…
158 00:28:06.380 ⇒ 00:28:25.530 Katie Ellig: in the exploratory phase for us, so, I would love to… if you could share just, like, a snippet about what you just kind of talked through of, like, what the process would look like, and kind of your initial suggestion for that, like, 10-hour phase. I think that would be super helpful for us.
159 00:28:25.530 ⇒ 00:28:45.950 Katie Ellig: look at and, see if, if not now, also maybe, in the future, as we’re doing a lot of thinking about, like, what customized onboarding looks like for users, and how we’re having a really strong loop from, what people are doing in the product back to the top of the funnel, so I think that would be super helpful for us to have.
160 00:28:47.460 ⇒ 00:29:03.530 Robert Tseng: Okay, yeah, no, that’s great. Maybe one more question that I have regarding, yeah, if we could kind of… do you have… do you have hypotheses about some of these early signals that, like, it’s like you have a gut for, like, what you’re trying… a gut feeling for, like, what you’re trying to prove out, but you’re just not able to see it yet?
161 00:29:03.680 ⇒ 00:29:14.180 Robert Tseng: I mean, I feel like sometimes being able to know how you think about things kind of helps us to contextualize what we’re, you know, writing a bit better as well.
162 00:29:16.800 ⇒ 00:29:29.209 Katie Ellig: I think maybe, like, a couple of notes from me, one hypothesis that we’re interested in is, like, historically, we have, had most of our, and I think Robbie might have talked about this, but, we…
163 00:29:29.420 ⇒ 00:29:39.880 Katie Ellig: have had most of our users coming to Ambrook as an accounting software. We’re increasingly, targeting users on the basis of, like, Ambrook Wallet and payments.
164 00:29:39.880 ⇒ 00:30:01.919 Katie Ellig: being the main, like, the primary or, like, the first way they’re interacting with the product, and so really understanding, like, how do those behaviors differ? Are the users that we’re targeting for… with those campaigns also taking the actions in the product of, like, activating the wallet and making payments that we would be expecting based on, that
165 00:30:01.920 ⇒ 00:30:04.950 Katie Ellig: That targeting is maybe, like, one area of
166 00:30:04.960 ⇒ 00:30:10.190 Katie Ellig: hypothesis that we’re… we’re investigating, like, interested in… in getting deeper.
167 00:30:10.480 ⇒ 00:30:11.940 Katie Ellig: on right now.
168 00:30:11.950 ⇒ 00:30:26.420 Katie Ellig: And then I think as we’re also expanding to multiple verticals, kind of from, like, very ag-focused to now trucking, construction, real estate, etc, understanding how, like, the,
169 00:30:26.420 ⇒ 00:30:36.910 Katie Ellig: you know, behaviors of those users are differing? Are we properly, like, tailoring, you know, how are we tailoring onboarding based off of those, like, different,
170 00:30:37.140 ⇒ 00:30:50.859 Katie Ellig: user signals that we… we may or may not be already collecting, but we’re not necessarily personalizing the onboarding process for right now, and, what that… what that looks like. So those are, like, a couple, maybe, areas, on my mind.
171 00:30:51.620 ⇒ 00:31:02.990 Robert Tseng: Got it. I mean, as far as those, like, verticals, I mean, these are, like, pretty firmographic details, like, Louis, you’re catching those signals already through, or, like, I guess, like, it’s less about identifying, hey, they’re coming
172 00:31:02.990 ⇒ 00:31:14.889 Robert Tseng: this… they’re coming from this, you know, sector, it’s more like, okay, well, what do we actually need to design as a workflow for this particular vertical? Is that kind of… is that a better way to put it? Okay, sure.
173 00:31:14.890 ⇒ 00:31:19.400 louis: I think we’d probably, like, maybe, like,
174 00:31:19.740 ⇒ 00:31:34.550 louis: I think we want to, like, further map, like, we could probably do an exercise, like, mapping out our current state, like, in terms of, like, signal performance, sort of our workflows. I think with this type of work, there’s a level of change management to make sure that people, like, we co-build with the teams, and that they are actually using these reports, we’re mapping out.
175 00:31:34.550 ⇒ 00:31:35.030 Robert Tseng: Absolutely.
176 00:31:35.030 ⇒ 00:31:44.850 louis: something that they’re gonna regularly access in action. I’ve been in a lot of engagements where, like, we just bring in… we bring in consultants, and they bring… they build out some beautiful stuff, and then, like, six months later.
177 00:31:45.060 ⇒ 00:31:58.679 louis: we have a new team building out in-house, and so I want to avoid that, and be like, let’s just, like, co-build, and so maybe we have this, like, current state of affairs, and then I think we can work backwards from our future state. There’s, like… and we could probably swing to the fences there, and be like.
178 00:31:58.700 ⇒ 00:32:19.809 louis: and if you guys could help us get there, there is, like, for example, we can go quite deep on per-user personalization. That does include signals, that also… that also does include semantics as well, and there’s a way to, like, things that we… things that we feed in intercom as well, like, there’s, like, hey, I’m a general contract, I would love to capture that somehow in a kind of programmatic way, and be able to feed that, and have a full view, and then have that sort of…
179 00:32:19.810 ⇒ 00:32:22.510 louis: Either it’s enrichment, or it’s something else that comes in, but…
180 00:32:22.670 ⇒ 00:32:31.879 louis: I think there’s a lot of, like, kind of interesting things now with AI that we… that probably can speed a lot of things up, and, like, give us this, like, incredible, really rich level of context,
181 00:32:31.950 ⇒ 00:32:48.699 louis: we definitely need help building that out. We don’t, you know, so that’s why I think, like, companies like yours is, you know, could be, like, interesting, and so maybe from there, there’s sort of, like, this gap analysis of, like, helping us think through, like, if we want to… we’re here, and we want to go there, like, what do we need to think about from a process, from a skills, from a kind of,
182 00:32:48.700 ⇒ 00:32:59.350 louis: From a tooling, and, you know, the more, kind of, insight you guys could provide there, it’s like, the better it’d be like, this feels very good, like, let’s… let’s just start, let’s start building on this, because this is something that we,
183 00:32:59.620 ⇒ 00:33:01.139 louis: We don’t want to be behind on.
184 00:33:02.630 ⇒ 00:33:22.160 Robert Tseng: Yeah, totally. Okay, cool. I know we’re… we’re at time, so I want to be respectful. Yeah, I mean, we will… we’ll come together. This was super great. Appreciate, kind of a lot of the context, taking the time out. And yeah, we’ll come back. I guess I’ll… we’re emails with Katie, but I guess maybe we’ll… we’ll loop Lewis in as well.
185 00:33:22.870 ⇒ 00:33:23.380 Greg Stoutenburg: Yeah.
186 00:33:23.740 ⇒ 00:33:25.340 Greg Stoutenburg: Thanks, Al, this was great, really appreciate it.
187 00:33:25.340 ⇒ 00:33:26.030 louis: committing.
188 00:33:26.470 ⇒ 00:33:27.200 Robert Tseng: Alright, thanks, David.
189 00:33:27.780 ⇒ 00:33:28.849 Greg Stoutenburg: Talk to you. Thanks.
190 00:33:29.320 ⇒ 00:33:29.970 Greg Stoutenburg: Bye.