Meeting Title: Brainforge x LMNT: Data Initiatives Date: 2025-11-04 Meeting participants: Uttam Kumaran, Dan O’Keefe, Robert Tseng, Shivani Amar, Jason Wu
WEBVTT
1 00:01:27.590 ⇒ 00:01:28.280 Dan O’Keefe: How’s it going?
2 00:01:28.960 ⇒ 00:01:30.080 Uttam Kumaran: How’s it going?
3 00:01:30.310 ⇒ 00:01:31.370 Dan O’Keefe: Dude, how are you?
4 00:01:31.790 ⇒ 00:01:32.670 Uttam Kumaran: Good.
5 00:01:33.110 ⇒ 00:01:34.350 Uttam Kumaran: How’s the week been?
6 00:01:35.410 ⇒ 00:01:40.860 Dan O’Keefe: Not bad. Just,
7 00:01:41.590 ⇒ 00:01:51.720 Dan O’Keefe: a little crazy going into, like… November’s a little crazy to, like, get stuff done before, year-end, before it, like, quiets down for the holidays and stuff.
8 00:01:52.030 ⇒ 00:02:01.619 Uttam Kumaran: Do you guys have a… like, is there… is there very… is there a lot of seasonality? Because it’s, like, in terms of Black Friday, like, for a lot of our e-commerce customers, this is, like.
9 00:02:02.260 ⇒ 00:02:07.309 Uttam Kumaran: this is, like, the Super Bowl coming up, but I was wondering, like, what are you guys seeing?
10 00:02:07.730 ⇒ 00:02:15.530 Dan O’Keefe: So we have a little bit of seasonality in the business, just as, people are outside.
11 00:02:15.530 ⇒ 00:02:16.730 Uttam Kumaran: They’re doing stuff.
12 00:02:16.730 ⇒ 00:02:22.670 Dan O’Keefe: Yeah, like, when it’s hot… when it’s hotter and people are outside more, it’s… so, like, the summer is big for us.
13 00:02:22.830 ⇒ 00:02:23.170 Uttam Kumaran: Yeah.
14 00:02:23.170 ⇒ 00:02:27.170 Dan O’Keefe: And then we don’t really do any sort of,
15 00:02:27.980 ⇒ 00:02:31.680 Dan O’Keefe: like, discounting or anything, so there’s not… we don’t… we don’t, like…
16 00:02:31.840 ⇒ 00:02:39.690 Dan O’Keefe: you know, do a bunch of, like, Black Friday stuff, so… The summer is definitely…
17 00:02:40.180 ⇒ 00:02:44.629 Dan O’Keefe: Busier time for us, but… I don’t know, it all… it all kind of feels busy, too.
18 00:02:46.810 ⇒ 00:02:48.039 Shivani Amar: Hey Utam, good to see you.
19 00:02:48.040 ⇒ 00:02:50.110 Uttam Kumaran: Hey, Shivani. Hi. Nice to see you.
20 00:02:52.240 ⇒ 00:02:54.720 Shivani Amar: I guess we’re waiting on Jason, right, Dan?
21 00:02:54.720 ⇒ 00:02:55.320 Dan O’Keefe: Yeah.
22 00:02:55.320 ⇒ 00:02:58.089 Shivani Amar: Okay, cool. And then Robert is on your team with them?
23 00:02:58.090 ⇒ 00:03:16.369 Uttam Kumaran: Yeah, so Robert and I both run Brainforce together. Oh, nice. Robert… I mean, both of us, have a lot of background in data. Robert more on the analysis side, but also, you know, worked a lot in… he worked at Flexport for a while, a lot on the logistics, shipping side, so in case any questions come up there.
24 00:03:16.600 ⇒ 00:03:21.330 Uttam Kumaran: And kind of doing some forecasting stuff for clients, so yeah, just wanted to have him.
25 00:03:21.730 ⇒ 00:03:23.230 Uttam Kumaran: On the line with us.
26 00:03:23.530 ⇒ 00:03:24.470 Uttam Kumaran: He’s on it.
27 00:03:24.470 ⇒ 00:03:32.649 Robert Tseng: Both of you. Yeah, I’m on a train one stop away from New York, so I’m gonna stay camera off for now, but, wanted to be on this call.
28 00:03:33.070 ⇒ 00:03:48.529 Shivani Amar: No worries. Cool. Well, it looks like Jason’s just, like, wrapping a call, but I think we can go ahead and get started. It sounds like you got to know Dan a little bit, and obviously we connected last time with him, but it sounds like you had some guiding questions for the conversation, so maybe we can dive in?
29 00:03:48.800 ⇒ 00:03:58.980 Uttam Kumaran: Yeah, so, and, you know, feel free to stop me if we’re going too quick, or if we want to kind of highlight somewhere, but generally what I arrived at at the last call is, you know, there is…
30 00:03:59.210 ⇒ 00:04:12.139 Uttam Kumaran: And it’s very common to where we kind of come into companies, is there’s sort of disparate reporting infrastructure, there’s probably some key KPIs that are either difficult to answer right now, or impossible.
31 00:04:12.140 ⇒ 00:04:31.620 Uttam Kumaran: There is, you know, probably a core structure that reporting is happening, but maybe it’s not ideal, and there’s also probably, like, key areas where you guys want to enable, you know, reporting. Dan, for context, Brainforge, we’re a data and AI consultancy. Sort of, we’ve worked with a ton of folks in e-com.
32 00:04:31.620 ⇒ 00:04:38.300 Uttam Kumaran: in TPG. Kind of most relevant is I did a stint at Athletic Greens, like, maybe, like, 5 or 6 years ago.
33 00:04:38.330 ⇒ 00:04:42.830 Uttam Kumaran: We’ve worked… we work with this company called Urban Stems right now, they’re in the flower business.
34 00:04:42.910 ⇒ 00:05:01.039 Uttam Kumaran: sort of kind of a wide mode of customers across CPG ecom, and then we also do a lot of work in SaaS and a few other industries, so kind of very familiar with your type of business. Of course, my questions today are probably going to be more around twofolds, like, one, learning a little bit more about
35 00:05:01.130 ⇒ 00:05:13.880 Uttam Kumaran: the business and the KPIs, so maybe we could start there, and then also have some questions on what’s hurting in terms of the reporting infrastructure. So maybe if we want to start there, we’d love to just hear about
36 00:05:14.020 ⇒ 00:05:14.960 Uttam Kumaran: Kind of, like.
37 00:05:15.090 ⇒ 00:05:32.740 Uttam Kumaran: how the business, like, the revenue is segmented, like, you know, to give you a sense, like, of course, I’m familiar with the product. I see you guys in multiple different areas. Shivani, you mentioned a little bit about owning distribution more, getting into, like, I think, wholesale, retail. I would love to just hear a little bit about, like, how…
38 00:05:33.250 ⇒ 00:05:39.340 Uttam Kumaran: Revenue is segmented through channels, and maybe we can… Just start there.
39 00:05:40.940 ⇒ 00:05:42.389 Shivani Amar: Dan, do you want to share a little bit?
40 00:05:42.390 ⇒ 00:05:46.800 Dan O’Keefe: Yeah, I’m happy to kind of walk through. So,
41 00:05:47.360 ⇒ 00:06:01.580 Dan O’Keefe: up until this year, Element was, you know, entirely, shopify, Amazon, wholesale. Our wholesale clients are…
42 00:06:02.060 ⇒ 00:06:03.690 Dan O’Keefe: mostly, like.
43 00:06:04.320 ⇒ 00:06:13.190 Dan O’Keefe: service through, Shopify, with the exception of, like, a few larger wholesalers, where they’re just at the point where it makes sense to,
44 00:06:13.680 ⇒ 00:06:19.600 Dan O’Keefe: Like, ship those orders via freight, which those orders are, like, kind of manually
45 00:06:20.080 ⇒ 00:06:27.120 Dan O’Keefe: inputted to our… our 3PL provider. So, and then obviously this year, we…
46 00:06:27.420 ⇒ 00:06:34.329 Dan O’Keefe: went online at Target, Walmart, we… Also…
47 00:06:34.980 ⇒ 00:06:44.400 Dan O’Keefe: have, like, Vitamin Shop as… actually, that was our, our kind of… sorry, like, foray into retail, probably, like, 24…
48 00:06:44.550 ⇒ 00:06:46.299 Dan O’Keefe: Yeah, in 2024.
49 00:06:46.420 ⇒ 00:06:55.520 Dan O’Keefe: So the revenue is really segmented by channel, so it’s, like, think of it as Shopify, Amazon, wholesale.
50 00:06:55.930 ⇒ 00:06:58.820 Dan O’Keefe: And then, like, Target, Walmart.
51 00:07:01.190 ⇒ 00:07:02.130 Shivani Amar: Did he freeze?
52 00:07:02.670 ⇒ 00:07:03.719 Uttam Kumaran: Phrase a little bit.
53 00:07:03.720 ⇒ 00:07:05.579 Shivani Amar: Yeah, so…
54 00:07:05.580 ⇒ 00:07:07.400 Dan O’Keefe: the,
55 00:07:09.260 ⇒ 00:07:20.990 Dan O’Keefe: The self-distribution is another piece of that. It’s smaller, but, and then within that, it’s broken down by, like, sparkling and, our drink mix, our core product, yeah.
56 00:07:21.420 ⇒ 00:07:22.020 Uttam Kumaran: Okay.
57 00:07:22.260 ⇒ 00:07:27.349 Uttam Kumaran: And then tell me a little bit about, like, what the goals are coming into the next year.
58 00:07:27.490 ⇒ 00:07:30.480 Uttam Kumaran: You know, again, for us, we…
59 00:07:30.620 ⇒ 00:07:42.090 Uttam Kumaran: are typically focused on, in e-commerce, the revenue side or the profit equation. And of course, the profit equation can include everything from the dollar coming in all the way to bottom line, whether it’s
60 00:07:42.120 ⇒ 00:07:51.820 Uttam Kumaran: You know, we’ve done analysis on cost of goods, you know, shipping, inventory, and then all the way sort of down. But, like, tell me kind of where…
61 00:07:51.860 ⇒ 00:07:53.890 Uttam Kumaran: The focus of the businesses.
62 00:07:53.910 ⇒ 00:08:08.739 Uttam Kumaran: And then of course, like, we’re a data company, so to do any work in our world, we have to have extremely clear metrics, and that’s what… what we affect. But ultimately, what we’re trying to enable is not just, like, pretty dashboards.
63 00:08:08.740 ⇒ 00:08:19.100 Uttam Kumaran: But it’s that everybody on your team can make more decisions and more accurate decisions. And so in that way, you know, we’re trying to get aligned as close to possible to outcomes, where if we help you
64 00:08:19.180 ⇒ 00:08:35.920 Uttam Kumaran: ideally, you know, typically what we see is, like, okay, we’re trying to affect a high-level metric, but we don’t know how to segment, we don’t know how to break it down evenly to find out where to put pressure. So, do you guys have, like, kind of clear goals for next year that we could start to use to sort of break down?
65 00:08:36.240 ⇒ 00:08:40.599 Shivani Amar: So, we do, we have our OKR in terms of, like, an overarching, hey, Jason.
66 00:08:42.240 ⇒ 00:09:02.140 Shivani Amar: overarching, OKRs in terms of, like, revenue we’re trying to hit for 2026, let’s say, right? But then, it’s like, how do you actually get there? So, Dan just broke down, like, the different ways we think about revenue across product type and then channel. And so, like, if we take wholesale as an example, right? Like, the wholesale team has maybe, like.
67 00:09:02.190 ⇒ 00:09:22.689 Shivani Amar: some thousands of partners on the wholesale side, and they’re trying to increase that for next year. And so for them, like, what kind of data insights might be helpful could be, like, which wholesale partners are about to, like, haven’t ordered in a while, which wholesale partners are our biggest accounts, and how do we tend to them a little bit more? And what they’re doing right now is using a system of, like.
68 00:09:22.690 ⇒ 00:09:32.139 Shivani Amar: Jason, you can speak to this more, but, like, they’re going into Shopify and, like, tagging different partners. We don’t have a B instance right now set up in Shopify, so they’re almost, like.
69 00:09:32.140 ⇒ 00:09:33.620 Shivani Amar: You know, like, you could have…
70 00:09:33.720 ⇒ 00:09:51.009 Shivani Amar: one gym with 5 locations having, like, 5 different accounts in Shopify with 5 different addresses, and then they’re, like, ordering for each gym right now, and then we’re tagging it as a wholesaler, and we’re trying to… so if the person owning, like, the wholesale OKR is trying to think about, like.
71 00:09:51.190 ⇒ 00:10:01.270 Shivani Amar: how do I segment a bit better? How do I think about trying to drive different behavior change? It’s, like, a lot of manual pulling from Shopify right now, and trying to, like, triangulate things.
72 00:10:01.370 ⇒ 00:10:04.270 Shivani Amar: Sure. Whereas, if you think about…
73 00:10:04.560 ⇒ 00:10:11.989 Shivani Amar: Our self-distribution folks, who are, like, they’ve launched self-distribution in a few different markets.
74 00:10:11.990 ⇒ 00:10:28.929 Shivani Amar: they’re trying to think about, like, where are… like, let me understand, like, different things about our geographies right now, so I can understand which would be the next market I would go into for self-distribution and why. And so, like, to give you an example, I was just talking to that person who does self-distribution, and he was like.
75 00:10:28.930 ⇒ 00:10:37.919 Shivani Amar: like, New York, maybe we don’t want to do self-distribution because there’s, like, a high cost of living, so we want to, like, partner with people, right? Sure. And it’s like, if I’m looking at retail data in New York.
76 00:10:37.920 ⇒ 00:10:57.040 Shivani Amar: it doesn’t look that strong, but that’s probably because people who are buying Element right now in New York are doing it via our website, via Shopify, or Amazon. Like, they’re buying online, people in New York City are less likely to go to Target, and there’s no water here. So, like, I think eventually you’d want to be able to, like, triangulate all of this stuff, and not just have, like.
77 00:10:57.330 ⇒ 00:11:10.219 Shivani Amar: retail, wholesale, and to be so separate, right? So if he’s trying to think about things geographically, he wants to understand how to tie together retail plus e-commerce, for example.
78 00:11:10.970 ⇒ 00:11:18.680 Uttam Kumaran: Makes sense. I mean, it’s kind of like what we do for several people, it’s just stitching omni-channel together. Yeah. So, that can be… that takes…
79 00:11:18.680 ⇒ 00:11:38.500 Uttam Kumaran: sort of what, Shivani, you proposed, which is a core data warehouse where you can land all the data. And for Shopify and our digital platforms, it’s a lot easier, but of course, when we talk about wholesale or more manual platforms, that’s where, you know, our team comes in to help you actually get that data structured in one place, and then, ultimately, it’s modeling that, so making sure you can see in one view
80 00:11:38.500 ⇒ 00:11:54.600 Uttam Kumaran: the sales from every… every order from every single channel with the dimensionality that your team needs, whether it’s geo, the product SKUs, information about the customer. And then lastly, sort of where we operate, is we take on even supporting some of that analysis work.
81 00:11:54.620 ⇒ 00:12:06.409 Uttam Kumaran: So, we have… we have companies, for example, that are in… that are omni-channel, but have a big retail presence, and they’re trying to understand campaign segmentation on the marketing side, and they’re like, hey, we’ve never segment… maybe it’s been…
82 00:12:06.680 ⇒ 00:12:15.500 Uttam Kumaran: six, seven years since we’ve segmented our customers, right? And understanding, like, which… how can we use data to actually build those segments? And of course, segments
83 00:12:15.500 ⇒ 00:12:31.379 Uttam Kumaran: are not just, like, to attribute your money to, but we have to act on them differently. And so, an exercise that we’ll do for them is, like, okay, let’s look at your data, let’s look at your existing segments, and let’s propose, hey, you have these customer tiers. And that’s the kind of work that we do. So it is a lot of what you mentioned, which is.
84 00:12:31.380 ⇒ 00:12:46.729 Uttam Kumaran: Setting up a core data warehouse, landing all that data in one place, modeling it, and then working with the stakeholders in the business to either enable them to self-service report, or if there are cases where we can support, then we kind of work with them on the analysis itself.
85 00:12:47.130 ⇒ 00:13:00.209 Uttam Kumaran: I guess my next question is gonna be, like, what is walking… like, if you guys were to have a question, like, something you mentioned, Shivani, today, like, hey, we want to know… we’re deciding to go into this market, why or why not?
86 00:13:00.400 ⇒ 00:13:02.590 Uttam Kumaran: Walk me through, like, how that happens.
87 00:13:02.880 ⇒ 00:13:06.189 Uttam Kumaran: And is there data involved right now? And if so, like.
88 00:13:06.370 ⇒ 00:13:24.469 Uttam Kumaran: like, what is the process? Because you mentioned there are… there is some tooling that maybe has been tried, you know, and I know people are getting data out of the UIs themselves, right? Out of the… I’m sure, the Amazon UI or the Shopify UI, but, like, how… how would you go about answering that question today?
89 00:13:24.740 ⇒ 00:13:33.470 Shivani Amar: So, an example of, like, the geographic question, the person I was just talking to around self-distribution, so it’s like, we’re sold online every… like, everywhere, right?
90 00:13:33.470 ⇒ 00:13:34.230 Uttam Kumaran: Yes.
91 00:13:34.230 ⇒ 00:13:49.290 Shivani Amar: But, like, let’s say we’re trying to figure out where do we want to do more self-distribution, what is the market dynamics there? Like, the person who would make that decision is looking at data related to our retail sales and trying to, like, triangulate stuff by looking at data that comes from spins.
92 00:13:49.290 ⇒ 00:13:59.460 Shivani Amar: Have you heard of Spins? He’s going into Spins and trying to look at, like, geographic data, but this isn’t triangular… Spins data is not triangulated right now with Shopify’s data.
93 00:13:59.460 ⇒ 00:14:00.200 Uttam Kumaran: Yes.
94 00:14:00.200 ⇒ 00:14:09.509 Shivani Amar: And, like, maybe somebody is working on stitching that data together, like, there might… that might already exist. I don’t… I’m new to Elements, so I’m like, somebody might have said, let me, like, combine.
95 00:14:09.770 ⇒ 00:14:11.209 Uttam Kumaran: They might have done it once at some point.
96 00:14:11.210 ⇒ 00:14:21.549 Shivani Amar: Yeah, it’s like, there might be, like, analysis that’s done ad hoc to say, like, let me combine Shopify data with Spin’s data to see, like, what I’m gleaning. There’s also, like, we have…
97 00:14:21.550 ⇒ 00:14:32.510 Shivani Amar: Spins data might tell you about point of sale, like, the sale to the actual end customer from retail, but then, like, we have something called Emerson that’s giving us data, around…
98 00:14:32.530 ⇒ 00:14:38.760 Shivani Amar: like, the purchase orders from the retailers. Yeah. So it’s like, that’s, like, another step in the chain, right?
99 00:14:38.760 ⇒ 00:14:40.080 Uttam Kumaran: Yeah, makes sense.
100 00:14:40.850 ⇒ 00:14:41.520 Dan O’Keefe: Okay.
101 00:14:41.520 ⇒ 00:14:42.409 Uttam Kumaran: Yeah, go ahead.
102 00:14:42.410 ⇒ 00:14:46.489 Dan O’Keefe: Yeah, I was just gonna say, yeah, so, like, a lot of this stuff happens in…
103 00:14:48.400 ⇒ 00:14:50.630 Dan O’Keefe: A good amount of it come…
104 00:14:50.870 ⇒ 00:14:58.469 Dan O’Keefe: Comes to me in what ends up being, like, a one-time ad hoc type analysis, which is useful
105 00:14:58.960 ⇒ 00:15:09.510 Dan O’Keefe: for… at, like, a point in time. But it’s… it very quickly becomes, like, stale, and it lacks… like…
106 00:15:11.070 ⇒ 00:15:24.500 Dan O’Keefe: because you’re dealing with… especially on, like, the e-commerce side, where you’re dealing with, like, large amounts of data as it relates to just, like, the sheer number of orders in any given month on Amazon or Shopify, where it becomes very difficult to, like.
107 00:15:24.690 ⇒ 00:15:27.760 Dan O’Keefe: Step back and really see, okay.
108 00:15:28.130 ⇒ 00:15:32.730 Dan O’Keefe: Any sort of, like, real trends, because it’s just everything…
109 00:15:33.610 ⇒ 00:15:37.100 Dan O’Keefe: Because of the limitations of just working out of a spreadsheet, it becomes.
110 00:15:37.100 ⇒ 00:15:37.440 Uttam Kumaran: Yes.
111 00:15:37.440 ⇒ 00:15:40.049 Dan O’Keefe: You almost have to look at things like.
112 00:15:40.340 ⇒ 00:15:55.350 Dan O’Keefe: last month compared to this month, which doesn’t give you the view of, like, stepping back and being like, oh, there’s a real trend here of something going on, but you’re just too close to it, and it’s… like, that’s how it’s… but it sounds like, like.
113 00:15:55.830 ⇒ 00:15:57.719 Dan O’Keefe: Yeah. Like, a lot of the stuff…
114 00:15:57.950 ⇒ 00:16:12.769 Uttam Kumaran: I mean, the form factor itself of doing… now, given your size, and the amount of business, the amount of channels, the amount of initiatives, it’s not possible to do this, and even if you were to get it right once, there’s no rigor, and can it be used for, like.
115 00:16:12.950 ⇒ 00:16:27.659 Uttam Kumaran: actual… like, actuals going forward, right? And that’s… that’s really difficult. I guess maybe a question for Jason, like, what… what do you… like, what is your perspective on, like, sort of the… the ecosystem right now around data? Kind of curious if…
116 00:16:28.100 ⇒ 00:16:31.860 Uttam Kumaran: like, if you guys have stepped foot into any tools so far, I mean.
117 00:16:32.150 ⇒ 00:16:34.229 Uttam Kumaran: I kind of mentioned a couple things that…
118 00:16:34.240 ⇒ 00:16:44.479 Uttam Kumaran: we’d look to propose. The other piece about us is we’re pretty tool agnostic, like, I don’t like saying, hey, we have to come in and implement 100 tools to get the job done.
119 00:16:44.480 ⇒ 00:16:54.850 Uttam Kumaran: there are, of course, some things that really make this job faster, but all of it is a conversation about budget, BI especially, right? If we’re going to implement a business intelligence tool.
120 00:16:54.880 ⇒ 00:17:07.140 Uttam Kumaran: for me, it’s important to know who’s… are people using this? Are most people consuming data? Are they just gonna export it to a Google Sheet? Like, are people gonna actually be exploring and running analysis? So, yeah, I’m just curious from your perspective, like.
121 00:17:07.300 ⇒ 00:17:11.230 Uttam Kumaran: What you think about the ecosystem of data tools, right now.
122 00:17:11.230 ⇒ 00:17:11.900 Jason Wu: Yeah.
123 00:17:12.450 ⇒ 00:17:22.390 Jason Wu: It’s pretty disparate, so I would say the one thing that we do have that’s pretty reasonable is,
124 00:17:23.069 ⇒ 00:17:33.610 Jason Wu: our acquisition reporting, so our partnership team, uses, a few dashboards that have been set up by another vendor, to do, acquisition
125 00:17:33.830 ⇒ 00:17:49.890 Jason Wu: Attribution. Okay. It’s, essentially a combination of our Shopify data, and GA data, and a couple other things that, help attributes back to, like, kind of, like, the right partner pages for that.
126 00:17:49.890 ⇒ 00:17:54.909 Uttam Kumaran: Is that, is that through, like, Northbeam, or is that just through, like, they built something, their own?
127 00:17:55.080 ⇒ 00:17:59.880 Jason Wu: It’s built something on their own. It all resides on a BigQuery database.
128 00:17:59.880 ⇒ 00:18:00.230 Uttam Kumaran: Integrate.
129 00:18:00.290 ⇒ 00:18:07.680 Jason Wu: With, like, Looker in the front end, but it pretty much kind of exclusively, like, been developed around, around.
130 00:18:07.680 ⇒ 00:18:08.190 Uttam Kumaran: I see.
131 00:18:08.190 ⇒ 00:18:09.860 Jason Wu: Acquisition reporting in itself.
132 00:18:09.970 ⇒ 00:18:26.000 Jason Wu: I think… and there’s a spin-off of that, I should say, where the team that did that also did some looker dashboards for some customer support type reports as well, that they get there as well. I think… and then everything else is kind of like…
133 00:18:26.400 ⇒ 00:18:38.680 Jason Wu: it kind of depends on the particular business holder, stakeholder, in terms of where they’re getting the data. So, for example, we have a person that’s responsible for all of our Amazon metrics and, like, paid ads.
134 00:18:38.720 ⇒ 00:18:48.349 Jason Wu: And I know he pulls data from multiple places, including some of those local dashboards and other sources that he has.
135 00:18:48.360 ⇒ 00:19:06.179 Jason Wu: when it comes to, you know, all the reports that Dan’s pulling, you know, that’s from a variety of different sources as well. And I think… I think that’s just… in my mind, the way I look at it is, you know, we’re probably at a size right now where it’s, like, everything should be ingested in one spot, and we’ve just got, like, that…
136 00:19:06.670 ⇒ 00:19:15.459 Jason Wu: essentially kind of centrally located, right? And I think that’s just kind of the strategy that we don’t necessarily have right now, that we should be marching towards.
137 00:19:16.440 ⇒ 00:19:33.859 Uttam Kumaran: Yeah, I mean, it’s actually great that you at least have… you guys are aware of, sort of, like, okay, at least in one area, and this is what we typically find, is that one or two teams have found the need to kind of go the distance, but they’re just solving their, like, one unique use case, right? So, for us, our job, Jason, is to kind of
138 00:19:34.030 ⇒ 00:19:46.529 Uttam Kumaran: show you, like, what are the options in, like, picking the right BI tool, picking the right ETL tool, and making sure that you choose this infrastructure so that it’s gonna last, sort of, based on the growth of the company.
139 00:19:46.530 ⇒ 00:19:54.609 Uttam Kumaran: And so, whether we can work with BigQuery or whatever it is, but I think that’s where we do a lot of support as well, is just helping
140 00:19:54.630 ⇒ 00:20:00.480 Uttam Kumaran: to, like, choose that infrastructure. But you’re right, like, kind of if you talk about the gamut of sources we’re talking about, there’s one
141 00:20:00.480 ⇒ 00:20:15.239 Uttam Kumaran: really robust APIs coming out of Shopify, coming out of your ad sources, right? Those are typically the best, most robust. You also are going to have stuff coming from, like, for example, FedEx and things like that. These just might be, like, CSVs that get emailed to somebody once a week.
142 00:20:15.240 ⇒ 00:20:29.149 Uttam Kumaran: Right? And then you kind of have the range in between. For us, it doesn’t really matter, it’s actually more about, like, making sure that all of those are documented, and that we bring it into one place. At that point, we can then tell you, hey, we don’t have
143 00:20:29.200 ⇒ 00:20:36.289 Uttam Kumaran: like, geo-fidelity from one source. Okay, so either we go and call that vendor and have to figure out, do you guys have
144 00:20:36.410 ⇒ 00:20:48.830 Uttam Kumaran: geo, can you add that as a column, or we’re just kind of stuck. I mean, we’ve done work with 3PLs, for example, where, yeah, we have to get them to build us an integration to send us the data, and they will do that, but it’s,
145 00:20:48.850 ⇒ 00:21:03.049 Uttam Kumaran: that’s what our team kind of goes and calls them and makes sure that that’s what happens. But ultimately, it’s like, can we end up with a clear schema around customers, around orders, around transactions, that the team as a whole
146 00:21:03.050 ⇒ 00:21:14.909 Uttam Kumaran: can say this is our source of truth for transactions across any sales channel. And the folks that just focus on Shopify can still do… they still and they should do a lot just within Shopify, right? We’re not here to…
147 00:21:14.970 ⇒ 00:21:22.620 Uttam Kumaran: completely replace those things. It’s actually… there’s a lot of stuff in Shopify that, if you’re just focused on that, you should go do there, but…
148 00:21:22.910 ⇒ 00:21:37.930 Uttam Kumaran: this omni-channel problem is something that we’ve dealt with across a lot of clients, where we have to bring this data together. And then finally, it’s like, okay, once we bring it together, what are the decisions that we want to make? Like, can we start to calculate things that maybe we didn’t have calculations on? Can we start doing
149 00:21:37.930 ⇒ 00:21:49.299 Uttam Kumaran: period of period analysis across past years, across channels where maybe we had no data, right? Another piece that I was going to talk about was, on the e-comm side, like.
150 00:21:49.300 ⇒ 00:22:08.290 Uttam Kumaran: one thing that wasn’t mentioned was sort of, tracking on just the website. So we do a lot of work on Amplitude and Mixpanel and GA, just, like, how are people coming through the conversion funnel? Of course, where are they coming from? And then, like, what is the LTV based on the products that they buy, and things like that. But again, all of that is, like, you have to have
151 00:22:08.820 ⇒ 00:22:12.679 Uttam Kumaran: Analytics data landed somewhere to even start on that analysis.
152 00:22:12.980 ⇒ 00:22:27.979 Uttam Kumaran: But ultimately, it’s like, it seems clear that there’s people in the company that can run these, which is great. And so our job is to kind of enable them, you know, make sure that if they’re going for a source of truth with customers, transaction data, web events, like wholesale data.
153 00:22:28.070 ⇒ 00:22:38.949 Uttam Kumaran: another kind of work we often do is, like, SKU taxonomies, like, how are we naming our SKUs? They’re going to one place to do that, and it’s kind of like a core team agrees on the definitions and things like that.
154 00:22:39.110 ⇒ 00:22:39.980 Uttam Kumaran: So…
155 00:22:40.390 ⇒ 00:22:49.649 Uttam Kumaran: That’s… that’s kind of where we land. To talk about, like, kind of what our typical structure is here is, one, we typically start with all of our clients on, like, a one-month discovery.
156 00:22:49.670 ⇒ 00:23:03.500 Uttam Kumaran: Where it’s a mix of sort of meeting with the folks here on this call, with other folks, understanding, like, what all the core priorities are, but also getting a really good understanding of the customer, of, like, the products and the sales channels.
157 00:23:03.500 ⇒ 00:23:12.030 Uttam Kumaran: In parallel, we also build. So, Jason, we probably worked with you to build out, like, our initial data warehouse to choose and build.
158 00:23:12.110 ⇒ 00:23:27.390 Uttam Kumaran: bring in an ETL tool to land all the data in one place, and then build, like, a core reporting schema, which would have a couple of those core tables that I mentioned, like the source of truth for customers, source of truth for transactions, for orders.
159 00:23:27.500 ⇒ 00:23:32.790 Uttam Kumaran: And then at that point, like, at the end of the month, from our side, what we typically deliver is one, like.
160 00:23:32.900 ⇒ 00:23:42.950 Uttam Kumaran: Of course, the work to develop that, we usually develop an architecture diagram of, like, what is now the data architecture, because we may be the only people that have come in and sort of asked
161 00:23:43.220 ⇒ 00:23:47.430 Uttam Kumaran: All of these questions around data, and then we deliver sort of a roadmap.
162 00:23:47.470 ⇒ 00:24:07.010 Uttam Kumaran: like, what is the analytics roadmap for 3 to 6 months? And at that point, we usually give, you know, our clients a decision whether you want to do that with us, whether we want to pick off a piece, or you’re free to shop that around. Like, that’s usually our month of work, given that there’s still a lot of discovery to be done here. Like, what do you…
163 00:24:07.550 ⇒ 00:24:10.550 Uttam Kumaran: What do you think about… That, roughly.
164 00:24:12.360 ⇒ 00:24:16.580 Jason Wu: Shivan, I don’t know what your thoughts are on this, but… You know, with the month
165 00:24:16.750 ⇒ 00:24:27.889 Jason Wu: where my mind goes is, like, what’s, like, the one or two use cases that we want to… that we currently don’t support, right? So, I kind of mentioned, like, you know, we’ve got a pretty reasonable dashboard for, like, acquisition.
166 00:24:28.020 ⇒ 00:24:30.450 Uttam Kumaran: I don’t know if…
167 00:24:31.000 ⇒ 00:24:33.730 Jason Wu: It’s as compelling to kind of say, well, here’s how we do it.
168 00:24:33.730 ⇒ 00:24:35.200 Uttam Kumaran: Right, because… I think…
169 00:24:35.200 ⇒ 00:24:37.550 Jason Wu: No one on our side is saying that it’s broken, right?
170 00:24:37.550 ⇒ 00:24:38.880 Uttam Kumaran: Okay, perfect.
171 00:24:38.880 ⇒ 00:24:44.189 Jason Wu: I think… I think what we’re trying to figure out is, well, what else is just, like…
172 00:24:45.470 ⇒ 00:25:03.530 Jason Wu: kind of the octopus of, like, you know, data that we have there, like, what else makes sense to consolidate, in a way, and, like, where we can get value out of, like, that infrastructure, right? To say, okay, I’m sold on, like, this because it now solves this… this question I’m trying to answer, right? Versus trying to recreate something that we already have.
173 00:25:03.890 ⇒ 00:25:18.429 Uttam Kumaran: That’s perfect. In fact, I think we’d prefer that. If you can point us only to the direction that matters, for example, if you’re like, hey, at the end of the month, we want to try to drive towards a simple, unified transactions dashboard, then that’s what we’ll work backwards from.
174 00:25:18.430 ⇒ 00:25:25.020 Uttam Kumaran: The nice thing is, like, we’ll make these healthy infrastructure decisions, you know, on these data tools in order to achieve that.
175 00:25:25.020 ⇒ 00:25:42.309 Uttam Kumaran: But we certainly want to drive towards that. We’re… we want to drive towards either an analysis output or a dashboard output that sort of wrangles this data in one place. Ideally, we have something to compare to in terms of how long it may have taken before, and ideally, it enables net new reporting.
176 00:25:42.310 ⇒ 00:25:46.879 Uttam Kumaran: Like, if the conversion side and the acquisition side is fine, then no need to…
177 00:25:46.910 ⇒ 00:25:51.340 Uttam Kumaran: for us to kind of go spin our wheels there. So I think we’re on the same page.
178 00:25:51.340 ⇒ 00:26:13.390 Shivani Amar: Like, Jason, the place that Phil told me to prioritize, because with the implementation of NetSuite, we’re gonna get much more data around inventory and supply chain and stuff like that, and, like, COGS, margins, like, that’s gonna just come with NetSuite in time, and so he was saying, like, focus more on the revenue side, so focus more on, like, the insights that Laura, and…
179 00:26:13.390 ⇒ 00:26:17.969 Shivani Amar: Laura and Paul, and…
180 00:26:17.970 ⇒ 00:26:31.970 Shivani Amar: you know, folks on that team are, like, trying to glean from, like, piecing together bits of, like, our… triangulating data around, like, our revenue. And so, like, maybe it’s instead of using the Shopify tags for things that, like, Laura’s doing and having it be so manual, it’s like, she’s able to…
181 00:26:31.970 ⇒ 00:26:38.959 Shivani Amar: see a list of all of our wholesale partners and see, like, who’s about to churn, who are their top wholesale partners that I want to, like.
182 00:26:38.960 ⇒ 00:26:44.430 Shivani Amar: lean in to that relationship, or whatever else, right? So, like, that might be an example.
183 00:26:44.490 ⇒ 00:26:50.360 Shivani Amar: And I think we can come up with what those, like, few priority questions are before we, like, kick off the work.
184 00:26:50.360 ⇒ 00:26:50.930 Uttam Kumaran: Yeah.
185 00:26:51.950 ⇒ 00:27:04.349 Dan O’Keefe: Yeah, I think that definitely is a good place. I think wholesale feels like a right place to start on this, just because that is… I don’t want to, you know, speak for the group, but it’s…
186 00:27:04.980 ⇒ 00:27:14.760 Dan O’Keefe: it’s a mix of Shopify and then, like, manual processes on, like, the larger orders, which makes it relatively unclear what the, like, overall
187 00:27:15.110 ⇒ 00:27:21.580 Dan O’Keefe: looks like, and then it’s also, it is very…
188 00:27:23.000 ⇒ 00:27:25.939 Dan O’Keefe: Or, you know, it could just be my lack of, like.
189 00:27:26.550 ⇒ 00:27:34.389 Dan O’Keefe: you know, not, like, understanding of the… the Shopify, like, platform, but it is quite difficult to, like.
190 00:27:35.010 ⇒ 00:27:42.869 Dan O’Keefe: You know, if we just wanted to pull a list of, like, hey, we know these are, like, 50 biggest wholesale, customers, like.
191 00:27:43.060 ⇒ 00:27:44.919 Dan O’Keefe: What have they ordered in the past?
192 00:27:45.110 ⇒ 00:27:50.609 Dan O’Keefe: 12 months on, like, a SKU-level basis, like, that is a very… actually, like.
193 00:27:50.810 ⇒ 00:28:01.130 Dan O’Keefe: did it recently. A very manual pull of, like, basically going into each customer’s, like, account number, seeing what they ordered,
194 00:28:01.480 ⇒ 00:28:02.500 Dan O’Keefe: So, like.
195 00:28:03.060 ⇒ 00:28:21.809 Uttam Kumaran: You can’t do… well, you can’t so… you can’t segment customers based on, like, element, right? Like, and so, therefore, you can’t say, pull everything from a segment. You have to rely on, like, basically manual tagging, and then say, pull these tags. And, like, that’s… that’s not a… that’s not a process, you know?
196 00:28:22.260 ⇒ 00:28:27.330 Shivani Amar: Yeah. I think related to that, it’s kind of like… it’s like the account management around these wholesale partners, right?
197 00:28:27.330 ⇒ 00:28:27.740 Uttam Kumaran: Yes.
198 00:28:27.740 ⇒ 00:28:52.409 Shivani Amar: That’s like a… that’s like one slice to help the wholesale team, but if we, like, zoom out from there, I think it’s also, like, what we talked about, geographies, it’s like, if I were to say which markets are we doing well in, but you have, like, wholesale partners, e-commerce revenue happening, and retail, it’s like, I want to be able to understand all of that as a whole, instead of, like, in the disparate fashion. Retail is good in Texas, not good in New York.
199 00:28:52.410 ⇒ 00:28:58.189 Shivani Amar: but e-commerce is good in New York, not good in Texas, or I’m making things up, but, like, I’m like, what’s my…
200 00:28:58.190 ⇒ 00:28:59.960 Shivani Amar: What’s my revenue by geography?
201 00:28:59.960 ⇒ 00:29:00.390 Uttam Kumaran: Yes.
202 00:29:00.390 ⇒ 00:29:17.409 Shivani Amar: what’s my revenue by SKU biography? Like, the different cuts, right? Yeah. And so, I think it’s, like, leaning more into, like, the revenue side of things. And then eventually, once we get our ERP set up, it’ll be… whether we have a head of data by then, or we’re continuing to work together, then it’ll be a lot more of, like, the cost side.
203 00:29:17.760 ⇒ 00:29:18.970 Uttam Kumaran: Yeah, we have a.
204 00:29:18.970 ⇒ 00:29:19.620 Shivani Amar: Right now.
205 00:29:19.780 ⇒ 00:29:34.349 Uttam Kumaran: Yeah, we had a client that moved to NetSuite, like, earlier this year, and all the work… part of the work we did for them is all the cost-profit margin equation. Exactly. And then they’re also doing all their fulfillment through there. So, yeah, I mean, if that seems like the situation, like, for us.
206 00:29:34.370 ⇒ 00:29:44.280 Uttam Kumaran: we… one is, I always tell her to point us at, like, the nastiest thing to go after. And if wholesale seems like the thing that’s the most disparate.
207 00:29:44.310 ⇒ 00:29:46.909 Uttam Kumaran: Like, clearly there’s probably some…
208 00:29:47.260 ⇒ 00:29:54.239 Uttam Kumaran: wins to be made, or some decisions to be made, then that’s a great place to start. We typically like to think about a basket of work
209 00:29:54.290 ⇒ 00:30:11.880 Uttam Kumaran: that takes about a month, and so you guys can get a sense of, like, what it’s like to work with us. We get some wins. We can also bite off a bit more and go after a couple different areas, and I can kind of put together a little bit of a list. In parallel, though, again, we would look to make some of these data decisions on where.
210 00:30:11.880 ⇒ 00:30:27.410 Shivani Amar: Yeah, so talk me through that a little bit, and I don’t know if other folks have a hard stop, but, like, you’re kind of like, hey, in month one, we’ll have, like, built a data warehouse, so it’s like, how long do you think the decision-making around deciding if it’s Snowflake versus, like, some.
211 00:30:27.410 ⇒ 00:30:34.720 Uttam Kumaran: It’s… it’s… yeah, I guess it’s… it depends on, our discovery, and it depends on,
212 00:30:35.720 ⇒ 00:30:55.059 Uttam Kumaran: it’s just kind of unique. It depends on, like, what’s been the strategy so far on procuring IT. For us, we’ve negotiated with every vendor, so we’re totally on your side in terms of getting the best deals, but it really matters about the budget, it matters about, like, who’s going to be using and what the volumes are. We have some pretty common recommendations, like, you already mentioned Snowflake.
213 00:30:55.350 ⇒ 00:30:56.819 Shivani Amar: You said Mother Duck last time.
214 00:30:56.820 ⇒ 00:31:04.969 Uttam Kumaran: Yeah, we… so I can put these options, and again, if it’s just us and Jason making the decisions, then we would put together all the…
215 00:31:05.060 ⇒ 00:31:17.359 Uttam Kumaran: sort of, like, here’s the trade-offs, and everything has trials, and we can extend those, so it’s not like… these aren’t, like, life-changing decisions, but we kind of want to make them, and then move on. And so, yes, it’s typically…
216 00:31:17.480 ⇒ 00:31:31.799 Uttam Kumaran: Again, this is where it takes… it’s gonna take a little bit of discovery to figure out truly how many sources of data we’re talking about, how complicated there are, but we want to kind of have, like, a north star of, like, hey, we want, like, a wholesale dashboard, and then we work backwards from there.
217 00:31:31.940 ⇒ 00:31:35.940 Uttam Kumaran: And… Like, I would say it’s probably 1-2 months of work.
218 00:31:35.990 ⇒ 00:31:43.090 Uttam Kumaran: The setting up a warehouse, setting up an ETL tool, plugging in the common stuff is not that bad. Like, that’s probably a few weeks.
219 00:31:43.110 ⇒ 00:31:48.549 Uttam Kumaran: Making the decisions on software can be anywhere from, like, 1 to 3 weeks of work.
220 00:31:48.560 ⇒ 00:32:06.840 Uttam Kumaran: These are all kind of, like, staggered, though. They’re not, like, we just do software and then we do this. In parallel, we’re having meetings to learn about the business and, like, learn about all the sources of data. Additionally, like, we can… in the meantime, if we’re building stuff and it’s not ready, we can continue to produce analysis that are urgent.
221 00:32:06.840 ⇒ 00:32:08.690 Uttam Kumaran: You know, support,
222 00:32:08.760 ⇒ 00:32:15.840 Uttam Kumaran: And so, it’s a little bit of scoping once we’re in. Like, we can only do so much discovery beforehand, but I think
223 00:32:16.220 ⇒ 00:32:22.939 Uttam Kumaran: Jason, then you’re right in terms of, like, trying to focus on something that is about a month or two months of work to chew off.
224 00:32:23.310 ⇒ 00:32:32.199 Uttam Kumaran: we will talk to everyone in the process, and then we can also find out, okay, what are the lift on other items? You know, that’s typically how we do things.
225 00:32:32.540 ⇒ 00:32:38.559 Shivani Amar: I think, like, let’s not, like, be, like, fully, like, okay, it’s wholesale, like, I don’t feel comfortable today being like, that’s a.
226 00:32:38.560 ⇒ 00:32:38.930 Uttam Kumaran: Sure.
227 00:32:38.930 ⇒ 00:32:57.589 Shivani Amar: that was made. I’m like, I want to triangulate… I’m having conversations with folks across the revenue team to try and, like, surface up what I think the top questions are, and I think it’s more about, like… like, if I think about, like, a leadership level question, it’s probably about combining data that has retail, wholesale.
228 00:32:57.590 ⇒ 00:33:08.829 Uttam Kumaran: It has those dimensions on it. So then that’s… that’s essentially it. If it’s like, hey, we need a set of omni-channel tables with this dimensionality to answer these questions.
229 00:33:09.150 ⇒ 00:33:15.179 Uttam Kumaran: What, kind of, if you talk about, like, what we would do, initially, we would go in and scope, okay, what are all the sources?
230 00:33:15.370 ⇒ 00:33:17.870 Uttam Kumaran: Who owns those, like, touchpoints?
231 00:33:17.870 ⇒ 00:33:18.460 Shivani Amar: Yeah.
232 00:33:18.460 ⇒ 00:33:22.620 Uttam Kumaran: We go find out what data is available, either that’s talking to the people internally, or if they’re like.
233 00:33:22.620 ⇒ 00:33:27.620 Shivani Amar: We never asked if they had an API, we go call them and figure out what’s the situation.
234 00:33:27.620 ⇒ 00:33:42.700 Uttam Kumaran: And then, basically, we’re like, okay, we can then figure out, like, the lift, how fast can we get to something usable? You know, whether it’s even just, for example, Shopify and Amazon is a lot, I would say, easier lift to produce reporting off of than some of the sources would know
235 00:33:42.920 ⇒ 00:33:49.639 Uttam Kumaran: you know, data. And so, it would… it would depend. And yeah, if you have the list, then that’s really helpful.
236 00:33:49.640 ⇒ 00:33:56.079 Shivani Amar: That’s, like, an example so far from conversations, right? Spins is, like, the point-of-sales data, Shopify is, like, e-commerce.
237 00:33:56.080 ⇒ 00:34:03.530 Uttam Kumaran: For example, even if you can combine Shopify, Amazon together, and just produce something, that is still, like, quite tremendous, so that’s how we would…
238 00:34:03.780 ⇒ 00:34:08.599 Shivani Amar: Dan, like, Dan will tell you that, like, Shopify and Amazon don’t always, like, it’s like they have different…
239 00:34:08.600 ⇒ 00:34:12.780 Uttam Kumaran: Well, no, there’s not… they don’t handle refunds the same, they don’t handle, yeah, it’s like…
240 00:34:12.780 ⇒ 00:34:32.329 Dan O’Keefe: Yeah, you know, I… it’s something that I don’t… and it’s like, as we’re going through this, it’s, like, kind of triaging, like, what is the most, like, ROI, high-impact stuff, and then there’s stuff that’s, like, me just sitting within finance that kind of bugs me, that’s, like, you know, which I do think it just creates general,
241 00:34:33.020 ⇒ 00:34:41.820 Dan O’Keefe: confusion within the org on, like, hey, you know, if I’m looking at, like, Amazon Seller Central, and it… I pull, like.
242 00:34:42.080 ⇒ 00:34:47.250 Dan O’Keefe: like, ordered products for the number, for the month. Why does that, like.
243 00:34:47.520 ⇒ 00:34:54.800 Dan O’Keefe: that number, like, a… which should be, like, a gross revenue number, not tied with, like, what I’m seeing at the end of the month.
244 00:34:54.800 ⇒ 00:34:55.989 Uttam Kumaran: Yes, exactly.
245 00:34:56.330 ⇒ 00:34:58.850 Dan O’Keefe: Which, I think, from just, like, our…
246 00:35:00.370 ⇒ 00:35:11.529 Dan O’Keefe: Edification and getting everybody on the same page of, like, okay, if we’re gonna have a source of truth as far as what is the data, like, it’s just,
247 00:35:12.130 ⇒ 00:35:17.519 Dan O’Keefe: a smaller piece within that, of getting everyone on the same page, of what all these definitions of…
248 00:35:17.520 ⇒ 00:35:31.270 Uttam Kumaran: Like, what is the definition of a transaction? When you do a refund, how does it get affected to pass revenue? So all those things, like, if you’re sourcing from the platforms, they have definitions that you can’t control. And the reason about doing a data warehouse is
249 00:35:31.270 ⇒ 00:35:40.400 Uttam Kumaran: The raw data sits as, like, an order comes in with a couple of dates, and you build from there, and you build a logic of, like, if there’s a return, do you reconcile it in the month of…
250 00:35:40.460 ⇒ 00:35:56.889 Uttam Kumaran: when it was returned, does it affect the current month? Like, you can start to build that logic out. But again, that’s something that… the reason why you can’t do that in the platform, because that’s what it is. They have a one-size-fits-all sort of approach, and you can’t tweak these, and it builds distrust, as you mentioned, like.
251 00:35:56.890 ⇒ 00:36:03.110 Uttam Kumaran: someone will say, well, I went into Shopify, and the numbers don’t line up with our export here. And there’s no, like, nobody can…
252 00:36:03.180 ⇒ 00:36:05.470 Uttam Kumaran: There’s, like, no rhyme or reason, you know?
253 00:36:05.660 ⇒ 00:36:14.299 Dan O’Keefe: Yeah, it’s not, like, a materiality on, like, the overall, like, numbers. Yes, for sure. It’s just… intuitively, if I’m somebody that… yeah, if it doesn’t…
254 00:36:14.970 ⇒ 00:36:17.759 Dan O’Keefe: as somebody that’s like, I want everything to tie out all.
255 00:36:17.760 ⇒ 00:36:18.600 Uttam Kumaran: Yes.
256 00:36:18.830 ⇒ 00:36:24.670 Dan O’Keefe: it pisses me off when it doesn’t, and then it’s like, okay, then, do I even trust these numbers at all?
257 00:36:25.030 ⇒ 00:36:43.980 Uttam Kumaran: Yeah, and you don’t know, like, is it an error, or is it just the fact that they treat something differently, right? And again, this is, like, you’re right, on totality, as long as it gets logged, but it is, like, okay, is this logged in this month versus last month? And you have to make a decision on it, and small fluctuations in that logic can change a lot, so…
258 00:36:44.180 ⇒ 00:36:44.980 Dan O’Keefe: Definitely.
259 00:36:45.260 ⇒ 00:36:51.929 Shivani Amar: And I think with the… I don’t know, so, like, when I… when you think about discovery output, right? Yes. Like, the recommended stack.
260 00:36:52.150 ⇒ 00:37:03.339 Shivani Amar: Yes. Okay. There’s maybe, like, a dashboard or two solving urgent questions, but then I think that there’s also this piece around, like, which metrics, and, like, that’s where I don’t know, like.
261 00:37:03.340 ⇒ 00:37:04.760 Uttam Kumaran: So this is where we… we would…
262 00:37:04.760 ⇒ 00:37:08.549 Shivani Amar: Metrics are most important. How do… where are we seeing definitional gaps?
263 00:37:08.690 ⇒ 00:37:09.460 Uttam Kumaran: Exactly.
264 00:37:09.460 ⇒ 00:37:13.430 Shivani Amar: Even, like, teeing up where we as element stakeholders need to get aligned.
265 00:37:13.780 ⇒ 00:37:30.309 Uttam Kumaran: So that’s what we would do, so as part of, like, we would produce for you, kind of, like, what is a 6-month analytics roadmap? Like, if you were to get ahead of data, this is, like, their directive, right? And there’s not only questions on aligning definitions, but at that point, we would have met all the stakeholders, so you know who owns
266 00:37:30.310 ⇒ 00:37:41.250 Uttam Kumaran: like, the decision-making, what are the core KPIs that drive this business? And then lastly, it’s our recommendations, and doing this work for a bunch of, you know, really high-volume e-comm.
267 00:37:41.250 ⇒ 00:37:58.599 Uttam Kumaran: what do we see that’s working across the board? And then that’s sort of, like, our deliverable. You know, I think given this scope, it’s probably a minimum a month of work, if not a little bit more, to kind of drive there. Like, we… some things, no matter how many people you throw at it, it’s just…
268 00:37:58.600 ⇒ 00:38:04.770 Uttam Kumaran: There’s just momentum, but, like, our process is not a, okay, we sign in, and then we just, like, leave for a month.
269 00:38:04.800 ⇒ 00:38:22.649 Uttam Kumaran: It’s like, we really have to come in and meet everybody and learn the story to kind of say, if you are going to go on an analytics sort of journey here, here’s what, over the next six months, you should do, and why, and how much it costs, and kind of who’s involved. But we need a theme, right? And we need… for our team, it’s like, I want to drive towards something like…
270 00:38:22.890 ⇒ 00:38:40.149 Uttam Kumaran: That we can wrap a little bit of a bow around and say, okay, at least we have a analysis on wholesale, or we have an analysis on revenue by segmentation, and we can, no matter what, at the end of the month, we want to drive towards that, and we will say, hey, this is, like, what it took to get there. Here are the broken pieces on why
271 00:38:40.280 ⇒ 00:38:45.440 Uttam Kumaran: it took us a… it takes a month, and how can we do that in a week? Next time, how can we do that in a day, you know? Yeah.
272 00:38:45.770 ⇒ 00:38:52.999 Shivani Amar: And I think the reason I’m, like, not wanting to just do wholesale is because they can, like… it’s manual, but they’re, like, gleaning these insights by.
273 00:38:53.000 ⇒ 00:38:53.370 Uttam Kumaran: Yes.
274 00:38:53.370 ⇒ 00:39:08.890 Shivani Amar: Shopify, but, like, to me, the beauty is when you’re actually, like, linking systems. Like, you’re actually… like, to me, when I hear, let’s link Shopify with Amazon, I’m actually, like, that’s, like, the beauty, that’s, like, the beauty of this work, versus just making a manual process faster.
275 00:39:08.890 ⇒ 00:39:24.480 Uttam Kumaran: I think part of the reason it’s nice to fix on one is mainly because we need to get buy-in from everybody, right? The fact that they should switch their reporting cadence and source of truth to this system. And so commonly, what we find is we have to kind of do both.
276 00:39:24.550 ⇒ 00:39:39.140 Uttam Kumaran: we sort of build this larger, like, okay, we’re getting source of truth, we’re having the right tools, but we’re also, like, solving that person’s problem today, and, like, they’re really happy, and their, like, job is easier. And, like, that’s… that’s why I’m trying to narrow us in that we’re… it’s… we have to do both.
277 00:39:39.140 ⇒ 00:39:39.690 Shivani Amar: Okay, great.
278 00:39:39.690 ⇒ 00:39:56.370 Uttam Kumaran: sit and build for 6 months. We also can’t… we also can’t get just back into, like, developing something ad hoc every day. Yeah. So, we have to solve the people’s problems to build their trust, and to get ownership and buy, and get their next questions. And similarly, we’re building the platform, you know, as we go.
279 00:39:56.370 ⇒ 00:40:02.359 Shivani Amar: Great. Okay, cool. So, I… what I’m doing in the meantime is, like, my next episode is I’m meeting a lot of the leaders.
280 00:40:02.360 ⇒ 00:40:02.900 Uttam Kumaran: Right, okay.
281 00:40:02.900 ⇒ 00:40:06.120 Shivani Amar: We started at Elements, so I’m, like, meeting a lot of the leaders on.
282 00:40:06.120 ⇒ 00:40:06.770 Uttam Kumaran: Perfect.
283 00:40:06.770 ⇒ 00:40:08.429 Shivani Amar: Revenue side of the business.
284 00:40:08.530 ⇒ 00:40:19.710 Shivani Amar: And then I can try to, like, form that into a hypothesis of, like, for you, yes, maybe it is, like, the quick… the quick win is, like, retail, or is wholesale, but then, like.
285 00:40:19.710 ⇒ 00:40:39.810 Shivani Amar: the, like, thing in… the next thing we’ll want is stitching these types of data together, right? Yeah. And so I can, like, form a perspective on that. And then, we’re also talking to, like, a couple other companies, and so I think, like, you and I have talked about costs and stuff, and again, I’m not trying to do a huge RFP process, but if you, like, can, like.
286 00:40:39.810 ⇒ 00:40:42.470 Shivani Amar: Synthesize this, kind of put, like, a…
287 00:40:42.630 ⇒ 00:40:43.170 Uttam Kumaran: Definitely.
288 00:40:43.170 ⇒ 00:40:58.560 Shivani Amar: sense of a proposal together for what that discovery and, like, implementation work would be. Just, like, I get that it’s ballpark because it’s, like, the tools that you decide to use will kind of determine, like, some of the cost pieces. I think that’ll just, like, help us compare the different.
289 00:40:58.560 ⇒ 00:40:59.330 Uttam Kumaran: Totally.
290 00:40:59.330 ⇒ 00:41:02.569 Shivani Amar: Cool. Yeah, we’ll put together sort of options, and then again, the…
291 00:41:02.570 ⇒ 00:41:08.469 Uttam Kumaran: Even if… even if it’s like, hey, we still don’t have enough to do, like, a core, hey, it’s gonna take 2 months, 3 months.
292 00:41:08.470 ⇒ 00:41:12.159 Shivani Amar: and you want us to come just help you with discovery, we can do that as well. Yeah.
293 00:41:12.410 ⇒ 00:41:26.219 Uttam Kumaran: we can… we can put that together, and then I think the biggest thing, I know there’s some deadlines on, like, NetSuite, so if there’s any other deadlines that you want us to sort of, like, focus on… yeah, I know Domo’s, there’s just a ton of tools across the board.
294 00:41:26.440 ⇒ 00:41:29.530 Uttam Kumaran: So I think that’s where, like, we can certainly help on, like.
295 00:41:29.530 ⇒ 00:41:32.059 Shivani Amar: Why pick one versus the other? Yeah.
296 00:41:32.060 ⇒ 00:41:35.080 Uttam Kumaran: Situation based on what budget and which users, you know?
297 00:41:35.080 ⇒ 00:41:42.330 Shivani Amar: Yeah, so I think that that’s the question, is like, similar to you, I’m not feeling like, just because one team is using Looker, we should use Looker, like, I think.
298 00:41:42.330 ⇒ 00:41:42.660 Uttam Kumaran: Yeah.
299 00:41:42.660 ⇒ 00:41:43.420 Shivani Amar: modern…
300 00:41:43.420 ⇒ 00:42:03.090 Uttam Kumaran: You’re gonna find a lot of consultancies that are… they just… that’s all they do. And we… we do do a lot of a certain amount of tools, but all of it is a decision. We also go in a lot of companies where they have a setup, and we can’t… it’s not that their… their dollars are not best spent us ripping that out. Instead, we just go, and so…
301 00:42:03.100 ⇒ 00:42:12.080 Uttam Kumaran: we’re not, like, hamstrung by that. In fact, most of it is, like, can we just make the right decision, and then drive towards the dashboard and the decision? Like…
302 00:42:12.120 ⇒ 00:42:16.200 Uttam Kumaran: The infrastructure is all, like, piping in the wall, you know? Yeah.
303 00:42:16.770 ⇒ 00:42:22.640 Shivani Amar: Makes sense. Cool. Dan and Jason, any other quick questions for the Brainforge folks?
304 00:42:24.480 ⇒ 00:42:32.489 Jason Wu: Nothing on my side. Siobhan, I think you and I can connect after this. I think your next steps are right.
305 00:42:33.340 ⇒ 00:42:34.080 Shivani Amar: Great.
306 00:42:34.360 ⇒ 00:42:40.219 Jason Wu: I’m trying to think… yeah, I don’t have any other questions on my side. And apologies, my… the internet here is just so crappy, that’s.
307 00:42:40.220 ⇒ 00:42:41.320 Uttam Kumaran: No problem.
308 00:42:41.320 ⇒ 00:42:41.950 Shivani Amar: All good.
309 00:42:41.950 ⇒ 00:42:42.919 Jason Wu: Damn it back.
310 00:42:43.290 ⇒ 00:42:45.810 Shivani Amar: Dan, any other questions on your end?
311 00:42:45.810 ⇒ 00:42:47.870 Dan O’Keefe: I think I’m good on my side.
312 00:42:47.870 ⇒ 00:42:58.469 Shivani Amar: Okay, cool. Okay, Utam, so I guess you’ll put some thoughts together while I’ll be… like, over the next, like, week and a half, I just have a lot of conversations slated, and then I can, like.
313 00:42:58.470 ⇒ 00:43:08.089 Shivani Amar: when we connect, maybe, like, let’s say, like, we connected, like, end of next week or something, I might have more of a perspective. If you’ve sent a proposal over, I’ve synced with the…
314 00:43:08.090 ⇒ 00:43:18.580 Uttam Kumaran: I’ll send some sample, like, proposals that we’ve written for some other folks that are really aligned with, like, we have a couple other clients that this is exactly what they’re doing, like omni-channel reporting, driving towards, like.
315 00:43:18.710 ⇒ 00:43:20.510 Uttam Kumaran: customer segmentation.
316 00:43:20.700 ⇒ 00:43:31.319 Uttam Kumaran: order segmentation, so we can show some of the scopes that we did for them, and then if you’re like, okay, that’s exactly what it is, then it at least helps us align. Yeah. Yeah, so we can do that.
317 00:43:31.320 ⇒ 00:43:38.139 Shivani Amar: Okay, that sounds great. Well, thank you for your time, and robert, hopefully we can chat more next time.
318 00:43:38.530 ⇒ 00:43:39.420 Uttam Kumaran: Cool.
319 00:43:39.420 ⇒ 00:43:40.260 Shivani Amar: Thank you.
320 00:43:40.410 ⇒ 00:43:40.950 Uttam Kumaran: Okay.
321 00:43:40.950 ⇒ 00:43:41.320 Dan O’Keefe: Yes.
322 00:43:41.810 ⇒ 00:43:42.650 Shivani Amar: Seeing you. Bye.
323 00:43:42.950 ⇒ 00:43:43.440 Robert Tseng: vinyl.
324 00:43:43.440 ⇒ 00:43:44.860 Uttam Kumaran: Thanks, everyone. Bye.