Meeting Title: LMNT x Brainforge: Data Initiatives Date: 2025-11-14 Meeting participants: Robert Tseng, Uttam Kumaran, philipmckeating, Shivani Amar
WEBVTT
1 00:00:34.720 ⇒ 00:00:35.050 philipmckeating: Hey, guys.
2 00:00:35.050 ⇒ 00:00:36.450 Uttam Kumaran: Hey! Hey, good morning.
3 00:00:36.450 ⇒ 00:00:37.619 Robert Tseng: Hey, mornin’!
4 00:00:37.620 ⇒ 00:00:38.220 Shivani Amar: Morning!
5 00:00:38.950 ⇒ 00:00:39.890 Uttam Kumaran: Good, how’s everything?
6 00:00:40.920 ⇒ 00:00:42.769 philipmckeating: Very good, very good. How about you?
7 00:00:43.210 ⇒ 00:00:44.879 Uttam Kumaran: Good. Busy week.
8 00:00:45.240 ⇒ 00:00:57.140 Uttam Kumaran: I feel like it’s… things haven’t been stopping for us, so… I don’t know, I thought holidays were gonna be slow. I mean, I’m talking to you guys, but yeah, it’s, for us and a lot of our clients, we’re in, like, peak
9 00:00:58.180 ⇒ 00:01:01.069 Uttam Kumaran: Black Friday, Cyber Monday stuff, so, yeah.
10 00:01:01.070 ⇒ 00:01:03.459 philipmckeating: Totally hear that, yeah.
11 00:01:03.460 ⇒ 00:01:04.099 Shivani Amar: Oh, God.
12 00:01:06.550 ⇒ 00:01:08.819 philipmckeating: You got the black sweatshirt memo today?
13 00:01:09.080 ⇒ 00:01:10.240 philipmckeating: Yeah, 100%.
14 00:01:11.660 ⇒ 00:01:13.370 Shivani Amar: We’re aligned.
15 00:01:13.880 ⇒ 00:01:14.579 philipmckeating: It was…
16 00:01:14.580 ⇒ 00:01:20.159 Uttam Kumaran: It’s funny, I was at… I was at a get-together on Wednesday, and they had a whole bowl of Element out, like…
17 00:01:20.160 ⇒ 00:01:20.930 philipmckeating: Is that right?
18 00:01:20.930 ⇒ 00:01:33.409 Uttam Kumaran: It was funny, and then it was weird, because they had, like, that, they had Breeze, they had all these, like, CPG, and I was like, maybe they got… it was, like, some meetup, they got sponsored by CPG, but it was, like, they had a… like, a candy bowl of elephant.
19 00:01:33.410 ⇒ 00:01:36.140 philipmckeating: That’s great.
20 00:01:36.860 ⇒ 00:01:37.510 Uttam Kumaran: Yeah.
21 00:01:37.510 ⇒ 00:01:49.250 Shivani Amar: So, Phil Utham is Austin-based, Robert is New York City-based, and then Phil is based in Boston, and he’s… he joined Element Phil in July.
22 00:01:49.560 ⇒ 00:02:00.489 Shivani Amar: But has worked in CPG, and, you know, I’m the one who’s, like, the CPG newbie. So, so I feel like he has more of a… more experience in, like, the CPG data realm.
23 00:02:00.490 ⇒ 00:02:11.049 Shivani Amar: And so, wanted to, like… I put his proposal in front of you, in front of him. I put your proposal in front of him. Sorry, I had a coffee, and I’d never drink coffee, so my brain was, like, a little black.
24 00:02:11.050 ⇒ 00:02:12.440 Uttam Kumaran: Never drink coffee, wow.
25 00:02:12.440 ⇒ 00:02:16.389 Shivani Amar: I don’t drink coffee, and I had a latte today, so I might be talking
26 00:02:16.940 ⇒ 00:02:22.370 Shivani Amar: I was like, oh my god, I’m, like, now just jumping into anxiety.
27 00:02:22.440 ⇒ 00:02:23.670 Uttam Kumaran: So…
28 00:02:23.670 ⇒ 00:02:29.069 Shivani Amar: I put the proposal in front of Phil this week, and he had some questions, so I thought we could just go through it together today.
29 00:02:29.070 ⇒ 00:02:39.449 Uttam Kumaran: Sure, definitely, yeah, and yeah, Robert from my team, both of us lead Brainforge together, we work across all of our clients, so yeah, I’m happy to jump in if you guys want to take the lead.
30 00:02:41.310 ⇒ 00:02:41.950 philipmckeating: Great.
31 00:02:46.010 ⇒ 00:02:52.619 philipmckeating: have, shivani, what we talked about the other day, we haven’t shared any of that yet, right?
32 00:02:52.620 ⇒ 00:02:53.710 Shivani Amar: No, no, no.
33 00:02:53.710 ⇒ 00:02:58.920 philipmckeating: Cool, okay. And do you guys have that propo… it might be helpful to toss the…
34 00:02:59.490 ⇒ 00:03:03.359 philipmckeating: document up that we walked through, and just kind of reference a little bit.
35 00:03:03.530 ⇒ 00:03:05.110 Robert Tseng: You’re gonna share? Yeah. Okay.
36 00:03:05.360 ⇒ 00:03:09.109 Uttam Kumaran: Yeah, and you can share it, you can run through it.
37 00:03:09.290 ⇒ 00:03:09.960 Robert Tseng: Alright.
38 00:03:11.650 ⇒ 00:03:13.209 Robert Tseng: This is the deck.
39 00:03:14.170 ⇒ 00:03:14.830 Uttam Kumaran: Yeah.
40 00:03:15.080 ⇒ 00:03:18.430 Robert Tseng: Alright, cool. If I need to blow it up, let me know.
41 00:03:18.550 ⇒ 00:03:20.329 philipmckeating: No, that’s perfect, yeah.
42 00:03:20.330 ⇒ 00:03:21.200 Robert Tseng: Yeah, okay.
43 00:03:21.200 ⇒ 00:03:24.460 philipmckeating: Great. We don’t need to watch it, we don’t actually Maybe.
44 00:03:24.460 ⇒ 00:03:26.489 Shivani Amar: A little bit of a timeline slide, Phil?
45 00:03:26.490 ⇒ 00:03:28.630 Robert Tseng: Yep, let’s go, let’s just jump to a slide, yeah.
46 00:03:28.630 ⇒ 00:03:29.000 philipmckeating: Okay.
47 00:03:29.000 ⇒ 00:03:30.040 Robert Tseng: This one? Yep.
48 00:03:32.010 ⇒ 00:03:38.310 philipmckeating: Yeah, so, I’m gonna say two things that are…
49 00:03:40.800 ⇒ 00:03:56.910 philipmckeating: you know, have some, dissonance between them, but that’s okay, we can all live in that. I love that you guys want to deliver business value quickly, and I like that that’s how you’re wired, that’s the right way to be, and we appreciate that, and that’s how we think about things.
50 00:04:03.300 ⇒ 00:04:13.789 philipmckeating: this… My goal here is to build, essentially, an enterprise-ready solution.
51 00:04:13.930 ⇒ 00:04:17.650 philipmckeating: And take the time to do that 100% right.
52 00:04:17.870 ⇒ 00:04:21.289 philipmckeating: I’m not… I don’t need to see business value
53 00:04:22.089 ⇒ 00:04:26.390 philipmckeating: day 14, to trust this process. I need to see…
54 00:04:26.880 ⇒ 00:04:30.689 philipmckeating: The right process, the right discipline and structure.
55 00:04:30.810 ⇒ 00:04:34.500 philipmckeating: And a pathway to get where we want to be.
56 00:04:34.640 ⇒ 00:04:35.710 philipmckeating: Quickly?
57 00:04:36.000 ⇒ 00:04:42.470 philipmckeating: And I, I, so, And…
58 00:04:42.910 ⇒ 00:04:51.310 philipmckeating: I think the reality is that our organization is… Going to be slower.
59 00:04:51.310 ⇒ 00:05:10.640 philipmckeating: than what you guys are setting up here. And so, like, I think that we need a significantly more robust, like, exploratory and information gathering phase, where we get, like, all the information together all at once, and then have you guys, like, really document and map it.
60 00:05:10.920 ⇒ 00:05:14.399 philipmckeating: And then go into implementation. And…
61 00:05:16.590 ⇒ 00:05:22.350 philipmckeating: I recognize I’ve ran a small consultancy at one point in my career, and like.
62 00:05:22.650 ⇒ 00:05:27.239 philipmckeating: It’s important to show business value, because that’s how you, like, keep the customers engaged and whatever, like.
63 00:05:27.840 ⇒ 00:05:28.830 philipmckeating: I’m okay.
64 00:05:29.210 ⇒ 00:05:44.850 philipmckeating: we’ll pay your retainer through that period of time. Like, I’m good with that. I trust, you know, as long as I see… I want to see a disciplined process, I want to see the projects being managed properly, I want to see the right… you guys asking the right questions. Based on everything I’ve seen here, that’s not going to be an issue at all. So,
65 00:05:45.120 ⇒ 00:06:00.429 philipmckeating: And ideally, I think it’s… that stage is going to take somewhere between 4 and 8 weeks. Ideally, we can get it done in 3 weeks, and I’ll be delighted, and, you know, whatever, you’ll be our partners for the next 10 years kind of thing. Like, that’s great.
66 00:06:00.730 ⇒ 00:06:10.150 philipmckeating: But documentation, to me, on all of this, is really, really critical, because I think one of the pieces that
67 00:06:10.380 ⇒ 00:06:24.469 philipmckeating: We’ve struggled with, or we struggle with right now, is we’re adding tons of people to the organization, and there’s lots of, like, pockets of knowledge and information that… it’s not only that data is siloed, it’s that
68 00:06:24.470 ⇒ 00:06:32.170 philipmckeating: knowledge about that data is siloed, right? And so, like, actually, I think documentation of this process is…
69 00:06:32.450 ⇒ 00:06:37.900 philipmckeating: equally as important as actually consolidating the data.
70 00:06:38.270 ⇒ 00:06:48.010 philipmckeating: So, just… I’ll stop there for a second, and just, like, does that framework make sense? I know it’s a little bit different than what you guys kind of, like, the approach you proposed, but,
71 00:06:48.700 ⇒ 00:06:51.710 philipmckeating: Does that feel like something that’s comfortable to you?
72 00:06:52.800 ⇒ 00:06:59.140 Uttam Kumaran: Yeah, maybe, Robert, I can let you sort of talk through, you know, our process a bit, given those points, yeah.
73 00:06:59.140 ⇒ 00:07:05.880 Robert Tseng: Yeah, yeah, thanks for clearing that, clearing that up for us. Yeah, I think obviously this 90-day plan is
74 00:07:05.880 ⇒ 00:07:20.369 Robert Tseng: typically just, like, our starting point for, you know, just showcasing our capabilities, how we think about things from, like, a waterfall progression, like, the different steps, but as far as, like, how long each of those takes, like, I think, obviously, it’s suggested to the organization.
75 00:07:20.960 ⇒ 00:07:34.480 Robert Tseng: Yeah, we recognize that, like, you know, Utah and I were kind of chatting before this call. We were thinking, you know, maybe this is a time, an opportunity for us to kind of slow down before we go fast. And so, yeah, I think, you know, if you want more details around
76 00:07:34.500 ⇒ 00:07:49.779 Robert Tseng: you know, our discovery process, what that first 30 days really looks like. You know, I… you know, we’ve, we’ve definitely worked, I mean, I worked, I worked at Ruggle Bowl previously, and so, I, I kind of felt like there were some similar
77 00:07:50.200 ⇒ 00:08:14.039 Robert Tseng: flavors here were a lot of influx of new staff, a lot of really competent former ex-consultants and bankers, not, like, a true data function that’s, like, kind of there yet, but a lot of people that can kind of triangulate their own insights that might be butting their heads, and so really keeping them all on the same page is really important. And so, we were kind of just, like, kind of talking through some of, like, the cultural dynamics there that maybe we didn’t capture in the initial
78 00:08:14.040 ⇒ 00:08:22.279 Robert Tseng: initial deck, and we won’t really know until we’re actually kind of inside the organization, kind of figuring it out. But what I will say is that,
79 00:08:22.280 ⇒ 00:08:44.409 Robert Tseng: Yeah, if you want us to kind of assess, like, the diligence that we’re… that we’d be taking as we’re kind of better framing the problem with your team, obviously, like, we didn’t… you know, these are just initial slides for us, and we tried to kind of write out those objectives based on some of these early conversations. But I think this would definitely involve us kind of
80 00:08:44.590 ⇒ 00:08:55.560 Robert Tseng: yeah, running… running… meeting every… all the folks in the room that are going to be important for helping us to describe, to frame the problem in the right way. And what we found is,
81 00:08:55.800 ⇒ 00:08:59.390 Robert Tseng: You know, especially… a lot of our clients have a…
82 00:08:59.710 ⇒ 00:09:24.630 Robert Tseng: they’re able to… to call… to describe the pain that they’re feeling, but they may not be able to… they may not actually know how to describe the problem that they’re having. Maybe, you know, Phil, because you’ve been in CBG, you kind of have seen… you’ve seen, like, kind of the whole timeline, and really what you need is just to kind of build advocacy and ownership amongst your… the folks on your team to kind of see things the same way. And so, we’re used to running those type of workshop sessions where it’s a bottom-up.
83 00:09:24.630 ⇒ 00:09:47.880 Robert Tseng: bottoms-up approach, we’re going to do the whole whiteboarding thing, get… sourcing, like, kind of what people are seeing. We’re also going to be seeding it with our own ideas as well, and, you know, kind of guiding, like, hey, these are some of the problems that are common, like, are you able to match the pain to the problem? And, like, once we’re able to find alignment there, then I think we’ll have the… we’ll have the problem framework that you want to see kind of moving forward.
84 00:09:47.880 ⇒ 00:10:00.669 Robert Tseng: You know, I think that’s… that’s kind of, like, one… one initial thing that I would… I would say, kind of to… back to you, just to kind of keep this com… this part of the conversation going. Yeah, I guess, kind of, what do you… what do you think about that?
85 00:10:02.130 ⇒ 00:10:06.539 philipmckeating: Yeah, I think that… that totally makes sense.
86 00:10:07.900 ⇒ 00:10:12.969 Shivani Amar: One thing I’ll note is that, like, when we look at this workstream approach slide,
87 00:10:13.180 ⇒ 00:10:17.960 Shivani Amar: Phil, like, I think this was very much, like, I had a few discovery calls, and so this is, like.
88 00:10:18.340 ⇒ 00:10:36.110 Shivani Amar: They’re… I talk to e-commerce, I talk to wholesale, and then I’m starting to form a perspective on where people need value added. But, like, I think to your point, I’m doing the discovery side with… I’m starting the discovery side with the commercial side of the business, but when Brainforge comes in, like.
89 00:10:36.110 ⇒ 00:10:47.450 Shivani Amar: We could have them… it would be nice to have them do discovery, like, across the board. Does that resonate with you? Or do you still feel like, starting with commercial is the way to go?
90 00:10:47.870 ⇒ 00:10:49.840 Shivani Amar: Starting with commercial is the way to go.
91 00:10:49.840 ⇒ 00:10:50.280 Robert Tseng: Okay.
92 00:10:50.280 ⇒ 00:10:52.520 philipmckeating: I, I think, I think the…
93 00:10:52.980 ⇒ 00:10:58.949 philipmckeating: Shivani, you didn’t happen to take a picture of that whiteboard that we did, just like the brief one that I did in the…
94 00:10:58.950 ⇒ 00:11:00.280 Shivani Amar: Oh, no, I didn’t.
95 00:11:00.280 ⇒ 00:11:01.819 philipmckeating: That’s okay.
96 00:11:01.820 ⇒ 00:11:02.800 Shivani Amar: Yeah, the, the, like.
97 00:11:02.800 ⇒ 00:11:05.479 philipmckeating: The way I… the way I was thinking about this, guys, is like…
98 00:11:05.690 ⇒ 00:11:06.280 Shivani Amar: Gantt.
99 00:11:06.490 ⇒ 00:11:19.710 philipmckeating: essentially, like, I’m gonna describe a Gantt chart, let’s see if I can do this with words, but it’s basically, like, a one-month Gantt chart, or the first row is… is call it, like, Shopify.
100 00:11:20.200 ⇒ 00:11:20.840 Robert Tseng: Yep.
101 00:11:20.840 ⇒ 00:11:40.309 philipmckeating: that’s a one-month Gantt chart, where we’re… week one is explore the data, week two is define and map the data, week three and four is import the data, and then the second row, there’s another one-week project in line with the fourth week, which is, okay, let’s use that data for some project, right? Yep.
102 00:11:40.680 ⇒ 00:11:42.320 philipmckeating: Probably starting…
103 00:11:42.490 ⇒ 00:11:53.659 philipmckeating: in week 2 or 3, we start with row 3, which is Amazon, right? And then you have a 4-week… you should have that same 4-week cycle, and then the last week or two, you’re, like, delivering.
104 00:11:53.660 ⇒ 00:11:55.060 Uttam Kumaran: Channel by channel, yeah.
105 00:11:55.060 ⇒ 00:12:00.350 philipmckeating: kind of like channel by channel, data set by dataset, because there’s…
106 00:12:01.140 ⇒ 00:12:04.880 philipmckeating: there’s a couple other things to note. One is that, I mean.
107 00:12:05.070 ⇒ 00:12:14.100 philipmckeating: our Shopify database is, like, really exten… or our Shopify volume is really extensive. I don’t know what you’ve seen, Robert, at Ruggable, but, like.
108 00:12:14.570 ⇒ 00:12:18.930 philipmckeating: at Circle, pulling out a similar history of Shopify, like.
109 00:12:19.220 ⇒ 00:12:21.609 philipmckeating: with, 5chan was, like, gonna take…
110 00:12:21.770 ⇒ 00:12:30.700 philipmckeating: I think it took, like, months. And so, like, with Shopify’s API limit, and maybe you have got… you guys have better ways around that, but, like.
111 00:12:31.100 ⇒ 00:12:35.799 philipmckeating: I don’t even think we could conceivably get that data out in, like, a week.
112 00:12:36.240 ⇒ 00:12:54.549 philipmckeating: so, like, I think we’re… we are gonna need… and Amazon’s probably gonna be similar. Our wholesale business is all on Shopify, so, like, it’s gonna be a smaller order volume with tags, so maybe we can pull… just, like, pull that out, or may just need to pull all of Shopify out and then separate it by tags, kind of thing. So…
113 00:12:55.750 ⇒ 00:12:57.179 philipmckeating: Well, all that to say.
114 00:12:57.690 ⇒ 00:13:03.520 philipmckeating: that’s fine. Like, the business is operating, it’s growing, it’s… the business is, like, doing great.
115 00:13:04.280 ⇒ 00:13:22.339 philipmckeating: Shivani’s done a great job of identifying, like, some clear pain points that we can rectify early, but when I got here, no one even had the idea of a data and analytics team. So, like, I built… just to say this, I will learn a lot from you guys, I don’t have all the answers here, but…
116 00:13:22.500 ⇒ 00:13:34.649 philipmckeating: at Circle, which is a similar scale business, I built our whole data and analytics team, and so we had a 13-person data and analytics team, build it from scratch, had, you know, Snowflake, DBT, Fivetran.
117 00:13:34.650 ⇒ 00:13:35.340 Uttam Kumaran: Great.
118 00:13:35.410 ⇒ 00:13:38.340 philipmckeating: Models on top of them, media mix model.
119 00:13:38.720 ⇒ 00:13:46.260 philipmckeating: forecasting models. So, like, a lot of what we’re gonna do here, not only do I have the CPG context, but I actually have specific.
120 00:13:46.260 ⇒ 00:13:46.780 Uttam Kumaran: Cool.
121 00:13:46.780 ⇒ 00:13:53.090 philipmckeating: on data and analytics context for this. So… Yeah.
122 00:13:53.090 ⇒ 00:14:01.220 Shivani Amar: So, Utam and Robert, what are your reflections on that, like, those timelines, now that you’re thinking about how large the data set is? Where does that,
123 00:14:01.360 ⇒ 00:14:04.780 Shivani Amar: Where have you seen, like, lengthy timelines to get.
124 00:14:04.780 ⇒ 00:14:06.300 Uttam Kumaran: Yeah, I mean, it…
125 00:14:06.560 ⇒ 00:14:21.830 Uttam Kumaran: It actually, you know, I would say the pain in discovery aligns with how manual or how complicated, like, the channel is. Like, stuff that’s non-digital is where there’s the most exposure to, like, okay, we have to go figure out how to get these sources wrangled.
126 00:14:21.830 ⇒ 00:14:32.099 Uttam Kumaran: On day one, you’re right, like, we should start to at least just kick off syncs and arrive at somewhere to land some amount of data, but we have to go channel by channel.
127 00:14:32.100 ⇒ 00:14:47.330 Uttam Kumaran: And so, I think exactly right, we’ve got to meet the people that are managing each channel, understand, like, what KPIs they’re going after, and then waterfall that. I mean, putting together, like, you know, we, as I mentioned, we work for a… we work for a pretty large
128 00:14:47.330 ⇒ 00:14:54.620 Uttam Kumaran: flower e-commerce business, and for them, we did, like, a two-month discovery. It was a huge business, and we walked into, like.
129 00:14:55.020 ⇒ 00:15:02.699 Uttam Kumaran: probably the opposite of you guys, which someone… they just had 8 years of, like, dbt code and Looker dashboards, and it was a complete mess.
130 00:15:02.720 ⇒ 00:15:13.160 Uttam Kumaran: And so, honestly, like, we basically had to both wrangle all of that and start something new on the side, but it was, like, a two-month discovery that… that…
131 00:15:13.160 ⇒ 00:15:26.230 Uttam Kumaran: we went and met every single person, and we now support the entire, like, 6 or 7-person analyst team that builds on top of all the marts that we develop. And so, I think we’re on the same page in the pacing.
132 00:15:26.230 ⇒ 00:15:44.910 Uttam Kumaran: For sure. And making sure that we go channel by channel and do that sort of Gantt-style, approach. And of course, I think one of the big things that, you know, wasn’t highlighted here is, and often is not asked by a lot of clients, is, like, documentation. So I’m, like, really happy to hear two things. One, that you’re familiar with.
133 00:15:44.910 ⇒ 00:16:00.869 Uttam Kumaran: what it is to set up a scope like this, and that you’re pressing for documentation. Usually, neither of those are the case when we walk into situations. And so, yeah, we do… we are a big, like, writing culture, even internally and externally. So, we do a lot of writing, we do a lot of…
134 00:16:00.870 ⇒ 00:16:06.010 Uttam Kumaran: presentation building, we record a lot of looms, and so we build, and within
135 00:16:06.010 ⇒ 00:16:16.199 Uttam Kumaran: you know, GitHub and in the data platform itself, we document a ton. So you’ll have all access to all of that, for sure. And so, again, as we are building that context.
136 00:16:16.200 ⇒ 00:16:32.960 Uttam Kumaran: we’re codifying that. And second, this is where, for us, the advantage of, you know, AI that’s available now, that context building is what can go power those insights. And so, for a lot of the solutions that we’re developing for clients, clients are now asking us to build chat with data type situations to
137 00:16:32.990 ⇒ 00:16:45.949 Uttam Kumaran: expand beyond just people having access to a dashboard, which maybe they’re trained to access or not, but that context gathering is really the make or break of that AI system, you know, and so that’s where, actually.
138 00:16:46.140 ⇒ 00:16:54.259 Uttam Kumaran: It’s actually benefiting a lot of our clients that we’ve done this much writing, because that immediately just goes to power that system, and it’s working really, really well.
139 00:16:54.470 ⇒ 00:17:04.489 philipmckeating: And that’s where, like, I think just… I’d rather take an extra week for every channel as part of the exploratory phase to be like, how should we name each.
140 00:17:05.020 ⇒ 00:17:05.450 Uttam Kumaran: Yes.
141 00:17:05.579 ⇒ 00:17:06.440 philipmckeating: variable.
142 00:17:06.440 ⇒ 00:17:07.730 Uttam Kumaran: All the naming convection…
143 00:17:07.730 ⇒ 00:17:27.349 philipmckeating: The naming conventions and stuff like that, like, because it’s… again, we have this across the business right now, where, you know, our sales team talks about revenue, and they’re talking about sales, and I’m like, guys, revenue is a GAAP-defined term. If you say revenue, and you’re not talking about things that are recognized as revenue by GAAP,
144 00:17:27.619 ⇒ 00:17:29.670 philipmckeating: I don’t trust you anymore, right?
145 00:17:29.670 ⇒ 00:17:30.050 Uttam Kumaran: I’m like.
146 00:17:30.170 ⇒ 00:17:32.969 philipmckeating: we’re gonna… I want everything fucking defined, and I want.
147 00:17:32.970 ⇒ 00:17:33.959 Uttam Kumaran: Yeah, yeah, yeah.
148 00:17:34.460 ⇒ 00:17:41.750 philipmckeating: in writing, right? Like, not like, oh, generally we all think it’s this way. It’s like, no, no, no. Let’s have a glossary. That’s…
149 00:17:41.750 ⇒ 00:17:42.200 Uttam Kumaran: Sweet.
150 00:17:42.290 ⇒ 00:17:53.390 philipmckeating: That… what we sell on a daily basis, that’s sales, not revenue, right? What the… what the retailer sells is point-of-sale sales, right? Yes.
151 00:17:53.390 ⇒ 00:18:07.489 Uttam Kumaran: And, you know, like, kind of two examples there, like, Shopify and Amazon recognize returns very differently, like, timing-wise, you know, and so these are all nuances that the channel people will not recognize the cross-channel differences until you reconcile them, right?
152 00:18:07.490 ⇒ 00:18:24.529 Uttam Kumaran: kind of another thing on sort of KPI standardization, I worked on WeWork sort of early on into the IPO, and KPI standardization going to the IPO was one of the biggest projects I worked on, where, yeah, we had basically the heads of state from every division in, like, a physical room in New York.
153 00:18:24.540 ⇒ 00:18:28.040 Uttam Kumaran: agreeing on, like, a Google Doc, and like, how do we define
154 00:18:28.120 ⇒ 00:18:45.599 Uttam Kumaran: gross margin. Like, how do we define cost of goods? How do we define contribution mar… like, all those things, and it’s… it is… it was literally written down and then codified in SQL. And then there is a committee to sort of define alternatives or variances that can be approved. Otherwise, it doesn’t end up
155 00:18:45.600 ⇒ 00:18:56.030 Uttam Kumaran: you can’t report on it, and it’s not like, yeah, and so… but again, you know it takes muscle to do that, and it takes, like, a vision to do that, and again, I think that’s what’s exciting.
156 00:18:56.080 ⇒ 00:19:10.330 Uttam Kumaran: for us, in that I think a lot of our… our clients, like, they’re not… they’re thinking about it, like, very broadly, and instead, I think thinking about it like this platform where, okay, there is governance over these metrics is important. So, yeah, I hear you on that.
157 00:19:10.770 ⇒ 00:19:16.659 philipmckeating: Yeah, so sounds great. I mean, I think that… that’s, like, that’s really what I wanted to talk about, is just, like, that’s the…
158 00:19:17.720 ⇒ 00:19:20.060 philipmckeating: If… if we’re doing that.
159 00:19:20.420 ⇒ 00:19:27.319 philipmckeating: don’t worry about proving to me that this adds value in one month, because I know that adds value, right? Like, I can…
160 00:19:27.320 ⇒ 00:19:27.850 Uttam Kumaran: Okay.
161 00:19:27.850 ⇒ 00:19:37.280 philipmckeating: And I have budget, so, you know, I can make that decision, and I can win that argument if anyone ever debates it with me. So…
162 00:19:38.570 ⇒ 00:19:45.249 philipmckeating: That’s all, really, I’m saying. So, Shivani, I think what would be really helpful is maybe if we…
163 00:19:46.110 ⇒ 00:19:49.939 philipmckeating: take that idea of a Gantt chart, could you guys all work together?
164 00:19:49.940 ⇒ 00:19:55.100 Robert Tseng: Yeah, let’s blow this up and really just, like, co-author kind of that workflow mapping.
165 00:19:55.100 ⇒ 00:20:10.809 philipmckeating: by channel, and, like, let’s define… so, like, when we’re like, okay, here’s what we’re doing from Shopify, when we get to, like, here’s the project that we do in week four, or, like, here’s how we deliver value, let’s have, like, three or four examples, and then we’ll decide as we get there, like, which ones to prioritize, right?
166 00:20:10.810 ⇒ 00:20:11.410 Shivani Amar: Great, yeah.
167 00:20:11.410 ⇒ 00:20:13.360 philipmckeating: Go, like, channel by channel, like.
168 00:20:13.750 ⇒ 00:20:21.379 philipmckeating: and again, I’d frame it as discovery, documentation, implementation, value is, like, my four phases.
169 00:20:21.560 ⇒ 00:20:24.060 philipmckeating: Channel by channel, and .
170 00:20:25.490 ⇒ 00:20:26.850 Uttam Kumaran: We can drag it around, yeah.
171 00:20:26.850 ⇒ 00:20:35.189 philipmckeating: Yeah, exactly. We’ll do it in something… I mean, you guys are obviously doing a good job of the planning, and you said the writing culture, so that all… it’s gonna vibe well, so.
172 00:20:37.130 ⇒ 00:20:42.389 Uttam Kumaran: Can we do that? And, and, I think maybe let’s get, like, a roughed-out version of that. Yeah.
173 00:20:42.390 ⇒ 00:20:48.840 philipmckeating: As part of the proposal, and then let’s get rolling here relatively quickly, and
174 00:20:49.490 ⇒ 00:20:57.219 philipmckeating: you know, that can be… the… that can be, like, that document should expand as we go through the discovery phase, right? Like, let’s have…
175 00:20:57.220 ⇒ 00:20:57.620 Uttam Kumaran: Yeah.
176 00:20:57.620 ⇒ 00:21:02.199 philipmckeating: The scaffolding, and then we kind of put the detail on it as we go.
177 00:21:02.400 ⇒ 00:21:03.439 Shivani Amar: That sounds great.
178 00:21:03.900 ⇒ 00:21:16.640 Shivani Amar: So, would you guys, like, do you feel like you have a sense with them from our previous conversations of getting started on that, and then blowing… Totally. We can talk about, like, you want to take a beat with that, or do you want to workshop some stuff live? What are you feeling?
179 00:21:16.990 ⇒ 00:21:34.250 Uttam Kumaran: No, let’s… let me take what we know, put it together, and I’ll put in, really, where the question marks are, and I’ll just make sure the scaffolding for every channel we’ve heard of is there. We’ll send it over to you today, and then we can just kind of collaborate on something, and then we can hop on as soon as we feel like, okay, we’re at a little bit of an inflection point.
180 00:21:34.800 ⇒ 00:21:35.809 Shivani Amar: That sounds great.
181 00:21:36.290 ⇒ 00:21:36.860 Uttam Kumaran: Okay.
182 00:21:37.230 ⇒ 00:21:38.360 Uttam Kumaran: Sweet. Perfect.
183 00:21:38.360 ⇒ 00:21:39.100 Shivani Amar: Cool.
184 00:21:39.100 ⇒ 00:21:41.460 Uttam Kumaran: Anything else, like, we can answer, or…
185 00:21:41.730 ⇒ 00:21:57.069 philipmckeating: Look, your pricing sounds reasonable. I wanted to hear how you guys react to this conversation. You guys… you know, I trust from this conversation that you know your shit, and if you don’t know your shit in 3 weeks, I’m gonna know that, and we’ll talk about it then. I don’t think it’s gonna happen, right?
186 00:21:57.070 ⇒ 00:21:57.780 Uttam Kumaran: I appreciate it.
187 00:21:57.780 ⇒ 00:22:02.170 philipmckeating: Yeah, feel good about this, just, I think, to me.
188 00:22:03.110 ⇒ 00:22:12.679 philipmckeating: your… your process rigor is what’s gonna earn my trust more than a proposal, right? So, I want to get rolling, and
189 00:22:13.280 ⇒ 00:22:19.970 philipmckeating: And, you know, we’ll just give you really specific feedback if we’re seeing any gaps. But, you know, I think…
190 00:22:19.970 ⇒ 00:22:20.520 Uttam Kumaran: Perfect.
191 00:22:20.520 ⇒ 00:22:24.080 philipmckeating: I like the proposal, I like how you’re thinking about it, we’ll figure out…
192 00:22:24.600 ⇒ 00:22:31.450 philipmckeating: stack and that type of stuff, I don’t think that, honestly, there’s that much controversial stuff that we’ll need to discuss there. I think there’s different…
193 00:22:32.290 ⇒ 00:22:39.569 philipmckeating: options and possibilities, we’ll just figure out what works. I think the only other thing that is worth noting, like, in those
194 00:22:40.020 ⇒ 00:22:47.330 philipmckeating: In those channels, and especially when it comes to this, kind of, like.
195 00:22:47.820 ⇒ 00:22:53.760 philipmckeating: you know, where I want to be a year from now in terms of documentation and data tying together is…
196 00:22:56.290 ⇒ 00:23:15.239 philipmckeating: I don’t know if Shivani gave you full visibility, this has been moving fast, but we were implementing an ERP, we were implementing Sage. We stopped that project, because I think Sage is the wrong ERP. We’re getting started on NetSuite. I expect NetSuite to be implemented probably by the beginning of the fourth quarter of next year, but we’re going to be moving relatively quickly on that in the background.
197 00:23:15.450 ⇒ 00:23:25.199 philipmckeating: And I want all of my NetSuite data in my data warehouse. I don’t like reporting out of NetSuite, and, I want… I want my data to tie.
198 00:23:25.370 ⇒ 00:23:32.750 philipmckeating: between my data sources. So, like, you know, again, in the data warehouse, I want NetSuite RevRec.
199 00:23:33.310 ⇒ 00:23:33.720 Uttam Kumaran: Yeah.
200 00:23:33.720 ⇒ 00:23:40.470 philipmckeating: and be able to tie back and forth, because I’ve seen that that can be really challenging, actually, with returns, with…
201 00:23:40.810 ⇒ 00:23:41.210 Uttam Kumaran: Yes.
202 00:23:41.210 ⇒ 00:23:43.870 philipmckeating: timing differences and that sort of thing. So just, like.
203 00:23:45.250 ⇒ 00:23:57.790 philipmckeating: as you think about, you know, as simple things as, like, okay, what is my, what attributes does my order, data element have in it? Like, I want a…
204 00:23:58.140 ⇒ 00:24:07.140 philipmckeating: order date, a process date, a ship date, a delivery date, a rev rec date, like, let’s have all of those elements, so then we can say, okay.
205 00:24:08.160 ⇒ 00:24:31.729 philipmckeating: when we pull RevRec, we’re pulling this subtotal, whatever that subtotal is, you know, minus this field for return flag or something like that, and use RevRec date, or however we end up defining that. But just know, like, I want that full kind of, like, traceability, auditability for how we’re kind of defining and organizing this data long-term.
206 00:24:31.730 ⇒ 00:24:46.959 Uttam Kumaran: Yeah, so it’s a lot of naming conventions and taxonomy in there, too, and yeah, we have a few clients that are using NetSuite or similar-style ERP, so… totally. And then, yeah, as soon as you guys get access to that, making sure we can set up the connection and land that. Yeah. Yeah, that’s… that’s perfect, great.
207 00:24:46.960 ⇒ 00:24:58.760 philipmckeating: We’ll be working with Myers Holm, I think, which is a big implementation partner, so, like, we’ll get you looped into that as soon as we’re, like, in an implementation phase, where we can actually start having data flowing, at least. Right.
208 00:24:58.760 ⇒ 00:25:09.330 Shivani Amar: And, like, the thing that I had shared with them, Phil, was more around once we have ERP and better inventory management, that’s when we can start, like, thinking about the supply side of the equation, with the, like.
209 00:25:09.590 ⇒ 00:25:17.379 Shivani Amar: with the eventual goal of having more sophisticated supply-demand modeling. Like, that, like, being a goal for, like, maybe a year or so from now.
210 00:25:17.380 ⇒ 00:25:18.240 philipmckeating: Yeah.
211 00:25:18.240 ⇒ 00:25:23.009 Shivani Amar: That’s kind of what we were talking about, but I think I hadn’t touched on the REVERIC piece, so that’s helpful.
212 00:25:23.010 ⇒ 00:25:33.559 philipmckeating: Yeah, well, and the other thing is… the other thing is, I think we’re gonna start implementing a tool called Atomic, which is, like, an AI supply-demand forecasting tool.
213 00:25:33.560 ⇒ 00:25:34.390 Uttam Kumaran: Awesome.
214 00:25:34.390 ⇒ 00:25:51.629 philipmckeating: But, the thing that, like, makes that work is good data in, right? Like, any forecasting tool is, like, you know, like any model, it’s if you put in good data, you typically get good outcomes, and if you can’t do that, then, you know, it’s a waste of time, so…
215 00:25:52.010 ⇒ 00:25:54.879 philipmckeating: Yeah, if we… if we have all the data organized, then…
216 00:25:55.080 ⇒ 00:26:12.830 philipmckeating: could be Atomic, it could be 5 different tools, we’ll figure out which tool works, right? Which… who has the best model, but as long as we have a clean data set, I know that there’s a model that can just sit on top of this, and we have, like, we have this easiest supply chain in the world that I… like, for a business like ours, it’s so freaking easy. Like, it’s…
217 00:26:13.580 ⇒ 00:26:14.170 Shivani Amar: It’s very…
218 00:26:14.170 ⇒ 00:26:25.989 Uttam Kumaran: Yeah, that’s like… that’s like Robert’s whole world, and we were just… we have two clients that are now at the point where we’re starting… we matured their data stack over the last 6 months to a year, where now we’re basically now able to actually
219 00:26:26.390 ⇒ 00:26:35.020 Uttam Kumaran: like, tried to attempt to replace whatever, like, remediary forecasting they were doing, like, through Google Sheets, and, like, actually set some outcomes.
220 00:26:35.020 ⇒ 00:26:35.640 philipmckeating: Yeah.
221 00:26:35.800 ⇒ 00:26:36.420 Uttam Kumaran: you know.
222 00:26:37.400 ⇒ 00:26:39.339 philipmckeating: 100%. Yeah, which is…
223 00:26:39.550 ⇒ 00:26:46.390 philipmckeating: we’re really excited for that. But, you know, again, my pretty firm view on that is.
224 00:26:47.180 ⇒ 00:26:56.279 philipmckeating: people try to do that too early in the process. It’s like, you have to have everything else set up well, and then that’s an easy exercise. If you try to do it too early.
225 00:26:56.550 ⇒ 00:27:03.589 philipmckeating: it just… it ends up not really working that well. Like, if you’re kind of hacking that together.
226 00:27:03.590 ⇒ 00:27:04.519 Uttam Kumaran: Strained, yeah, you’re.
227 00:27:04.520 ⇒ 00:27:08.270 philipmckeating: Yeah, the juice isn’t, like, the squeeze. You’re like, there’s too many manual, like, fixes.
228 00:27:08.270 ⇒ 00:27:08.940 Uttam Kumaran: Yes.
229 00:27:08.940 ⇒ 00:27:12.690 philipmckeating: Versus if you just have the infrastructure set up to begin with.
230 00:27:12.810 ⇒ 00:27:15.529 philipmckeating: It actually gets to be pretty turnkey, and…
231 00:27:15.710 ⇒ 00:27:20.269 philipmckeating: kind of scalable. Like, you can actually automate it once you have good data.
232 00:27:20.570 ⇒ 00:27:32.539 Uttam Kumaran: And we’re working with clients on, like, basically a gap between what you forecast and actuals, and, like, that’s what we attack, but for many people, they re-forecast every, like, few months. It’s like, what’s the point?
233 00:27:32.540 ⇒ 00:27:33.600 philipmckeating: Completely.
234 00:27:33.600 ⇒ 00:27:34.200 Uttam Kumaran: Yeah.
235 00:27:34.500 ⇒ 00:27:50.159 Shivani Amar: One little thing I’ll sprinkle in, a little bit of a side note, is you and I have spoken about the wholesale team, and how they would love more visibility into, like, when a customer is about to churn, or something like that. Phil had a nice idea. Phil, can you remind me what that acronym was that was, like, to…
236 00:27:50.160 ⇒ 00:27:55.050 Shivani Amar: to rank wholesale partners by their order size and frequency. What was that? .
237 00:27:55.050 ⇒ 00:27:55.910 Robert Tseng: RFM?
238 00:27:55.910 ⇒ 00:27:57.870 philipmckeating: I do RFM segments?
239 00:27:57.870 ⇒ 00:28:00.279 Shivani Amar: You guys know.
240 00:28:00.280 ⇒ 00:28:05.030 philipmckeating: That’s what I was saying, I was like, once we have the data, like, these guys have done this.
241 00:28:05.030 ⇒ 00:28:10.110 Shivani Amar: Yeah, okay, you guys got it. So yeah, that was… that was one, and I had this big conversation with, like, our
242 00:28:10.110 ⇒ 00:28:28.169 Shivani Amar: chief kind of commercial officer, I think is his title, yesterday about pricing, and you had kind of mentioned, like, at some point, are you going to want to do some sort of customer segmentation? And I was like, that doesn’t feel pressing to me right now. But in talking to him yesterday, I was like, at some point, that will also come into play. So just, like, I think you’re.
243 00:28:28.170 ⇒ 00:28:28.520 Uttam Kumaran: Yeah.
244 00:28:28.520 ⇒ 00:28:42.119 Shivani Amar: was right about the kinds of things we would eventually need, so, like, stay with that. As you’re, like, drafting the projects, you’ll kind of have a sense, even maybe better than I do, being new to this industry, of, like, what has typically helped customers when these… Exactly.
245 00:28:42.310 ⇒ 00:28:46.950 Shivani Amar: funnel, and so, like, feel free to put that in versus just a reflection of, like, what I’ve been sharing with you.
246 00:28:47.460 ⇒ 00:29:02.379 Uttam Kumaran: Totally, and it’s exactly what Phil mentioned, that as we’re going, we will attempt to do what exists or not. Like, if there’s someone at some point created a segmentation, we’ll at least have it there. Then it’s… then we can at least debate it at that point. Right now, there’s no columns that indicate, or no logic.
247 00:29:02.380 ⇒ 00:29:12.300 Uttam Kumaran: then there’s nothing to debate, right? So, totally, yeah. So, that’s sort of how we operate. We’ll do what’s best in class, and then we sort of align. So, perfect.
248 00:29:12.560 ⇒ 00:29:12.900 Shivani Amar: That’s.
249 00:29:12.900 ⇒ 00:29:16.850 philipmckeating: Sounds great. Awesome. We’ll look out for that later today. Thanks, guys.
250 00:29:16.850 ⇒ 00:29:19.950 Uttam Kumaran: Perfect. Thanks, everyone. Alright. Talk to you soon.