Meeting Title: LMNT QA Sync Date: 2026-05-07 Meeting participants: Greg Stoutenburg, Advait Nandakumar Menon, Jasmin Multani, Shivani Amar
WEBVTT
1 00:00:50.680 ⇒ 00:00:51.580 Greg Stoutenburg: Hey, team.
2 00:00:51.580 ⇒ 00:00:53.409 Advait Nandakumar Menon: Hey, Greg, hey, just…
3 00:00:53.670 ⇒ 00:00:54.590 Jasmin Multani: Hey!
4 00:00:54.750 ⇒ 00:00:55.520 Greg Stoutenburg: So…
5 00:01:18.200 ⇒ 00:01:22.870 Greg Stoutenburg: I need to just get an AI avatar to do these, demo videos.
6 00:01:22.870 ⇒ 00:01:23.760 Jasmin Multani: What was that?
7 00:01:23.760 ⇒ 00:01:24.790 Greg Stoutenburg: Hey, Shivani.
8 00:01:25.640 ⇒ 00:01:31.939 Shivani Amar: I did not review… I think you slacked me, Jasmine, I haven’t opened anything that you asked me to open, so…
9 00:01:31.940 ⇒ 00:01:32.580 Jasmin Multani: Yeah, that’s funny.
10 00:01:32.580 ⇒ 00:01:33.779 Shivani Amar: Do this on the fly.
11 00:01:35.250 ⇒ 00:01:40.170 Jasmin Multani: Yeah, yeah, yeah, I kind of, partially also assumed that, like.
12 00:01:40.490 ⇒ 00:01:44.920 Jasmin Multani: we’ll be doing some of these reviews on the fly, too. Yeah.
13 00:01:45.420 ⇒ 00:01:56.000 Jasmin Multani: But today, you have an option of agendas to go through. We can either start with the live reviews of the dashboards, there are three.
14 00:01:56.160 ⇒ 00:02:00.499 Jasmin Multani: Or, we can review, Odovitz…
15 00:02:00.880 ⇒ 00:02:04.320 Jasmin Multani: follow up on salt exposure. You want to go with that?
16 00:02:05.140 ⇒ 00:02:06.390 Jasmin Multani: Okay, cool, cool, cool.
17 00:02:06.440 ⇒ 00:02:20.299 Jasmin Multani: So, since Tuesday, Avid was able to compile a few things. A, he was able to map out which of the data table metrics are currently exposed to SALTI, so we have that handy.
18 00:02:20.300 ⇒ 00:02:31.660 Jasmin Multani: But before… but instead of just going row by row, I want Adva to lean on, reveal the recommendation he was able to click into, so that we can meet your…
19 00:02:31.980 ⇒ 00:02:34.799 Jasmin Multani: Final assessment of, like, let’s just turn everything on.
20 00:02:34.970 ⇒ 00:02:35.630 Shivani Amar: Okay.
21 00:02:35.630 ⇒ 00:02:36.950 Jasmin Multani: So I’ll let Avid lead.
22 00:02:37.720 ⇒ 00:02:38.160 Shivani Amar: Cool.
23 00:02:38.160 ⇒ 00:02:41.470 Advait Nandakumar Menon: I’ll share my screen real quick.
24 00:02:47.000 ⇒ 00:02:48.670 Advait Nandakumar Menon: Are you able to see it?
25 00:02:50.030 ⇒ 00:03:06.510 Jasmin Multani: Yeah, so I think it’d be helpful if we first went under, like, the trade-offs. What does it mean if we, turn everything on versus turn certain things off? So Avid is gonna briefly go over the underlying, trade-offs.
26 00:03:06.720 ⇒ 00:03:09.319 Shivani Amar: I get it, I’m reading it, and that makes sense.
27 00:03:09.480 ⇒ 00:03:10.320 Shivani Amar: Yeah.
28 00:03:11.520 ⇒ 00:03:13.600 Advait Nandakumar Menon: Okay. Yep, so…
29 00:03:13.860 ⇒ 00:03:23.939 Advait Nandakumar Menon: basically, there’s the option of exposing all the fields, which is what we recommend right now to Blobby. So, the thing here is that
30 00:03:24.230 ⇒ 00:03:36.570 Advait Nandakumar Menon: Then, we expose all the fields to Blobby. So, just for your context, for your understanding, there are two parameters within Omnitopics that is called the fields and the AI fields. So.
31 00:03:36.680 ⇒ 00:03:46.459 Advait Nandakumar Menon: the thing we really care about is AI fields, because that’s what Blobby is gonna see. So, Omni’s AI has a context window of around
32 00:03:46.460 ⇒ 00:03:57.370 Advait Nandakumar Menon: 200K characters, and across all the topics and tables and whatever you’re gonna include, it’s a maximum of 550 fields across all topics, so…
33 00:03:57.470 ⇒ 00:04:01.390 Advait Nandakumar Menon: In order to come under this minimum limit.
34 00:04:01.570 ⇒ 00:04:13.229 Advait Nandakumar Menon: I mean, the maximum limit, we need to keep in check which fields we expose to Omni, I mean blobby, and keep the token character count under check, basically, so that…
35 00:04:13.290 ⇒ 00:04:32.540 Advait Nandakumar Menon: If we go above the token character count or the field limit, blobby might not be able to see all the metadata description or whatever fields we are exposed to, and there can be truncation to be, to summarize it. So, there can be instances where, the quality of answers may be degraded.
36 00:04:32.730 ⇒ 00:04:39.770 Advait Nandakumar Menon: So, my recommendation is that we start exposing all the fields to SALTI for now.
37 00:04:39.770 ⇒ 00:04:54.229 Advait Nandakumar Menon: And then, during our QA for Blobby, during that sprint, we can iteratively restrict the fields we’re exposing it to it one by one, based on the questions we are going to ask Blobby, based on the type of
38 00:04:54.230 ⇒ 00:05:04.760 Advait Nandakumar Menon: topics we’re looking at, and also iteratively work on the metadata, the descriptions, the synonyms behind each field in a topic and table, so…
39 00:05:04.880 ⇒ 00:05:11.140 Advait Nandakumar Menon: This is the process we are thinking to go ahead with. What do you think about it, Shivani?
40 00:05:11.760 ⇒ 00:05:25.499 Shivani Amar: I mean, like, what I was hoping to see was actually just the list of fields. Like, I understand the context here, but I thought you were gonna show me, this is what we are currently suppressing, and then maybe I can say, yeah, keep that suppressed versus not. So, is that what we’re gonna do?
41 00:05:25.500 ⇒ 00:05:27.730 Greg Stoutenburg: We can, it’s very long.
42 00:05:28.190 ⇒ 00:05:44.739 Shivani Amar: So then I can do it… remember I said I can look at this async also? It’s like, if these are… if you’re just like, hey, this is a list of what’s being suppressed, I can, like, spend some time digesting this, if you orient me to it right now, and then I can comment where I think, hey, we should be including this.
43 00:05:46.610 ⇒ 00:05:54.610 Jasmin Multani: Sure, let’s do that. Aviv, can you, show Divani how to use this doc, and where she should explicitly say, keep it repressed, keep it going?
44 00:05:55.310 ⇒ 00:06:02.499 Greg Stoutenburg: And just to be, like, super-duper clear, is this the document that contains for every single field where it is exposed?
45 00:06:03.400 ⇒ 00:06:11.460 Advait Nandakumar Menon: Okay, so this document specifically, so the actual document, like, the list is very big, it goes on for, like.
46 00:06:11.460 ⇒ 00:06:13.300 Greg Stoutenburg: Yeah, 107 pages, this is about.
47 00:06:13.300 ⇒ 00:06:16.390 Advait Nandakumar Menon: 107 pages, so… Yeah, it’s…
48 00:06:16.390 ⇒ 00:06:17.169 Greg Stoutenburg: high level.
49 00:06:17.600 ⇒ 00:06:24.670 Advait Nandakumar Menon: Yeah, it’s all the topics, all the tables behind those topics, and all the fields in it, whether it’s exposed to blobby or not, so…
50 00:06:24.940 ⇒ 00:06:35.090 Advait Nandakumar Menon: It’s very detailed, so to just summarize it at a high level, I have come up with certain… not certain, all the topics, but…
51 00:06:35.180 ⇒ 00:06:52.489 Advait Nandakumar Menon: the business-facing relevant fields, and whatever fields you mentioned about, like, the fridges, the addresses should do, and the relevant, fields we talked about in the previous meeting. So, I have, come up with a table of sorts for each topic, and
52 00:06:52.630 ⇒ 00:07:02.230 Advait Nandakumar Menon: the fields that’s used in it, whether it’s exposed or not. So, in some situation, it’ll be exposed to the fields parameter, which is just
53 00:07:02.310 ⇒ 00:07:15.939 Advait Nandakumar Menon: the field used in the dashboard or the query, but it might not be, exposed to AI fields, which is, again, the field which Blobby cares about. So, there can be situations like that, there can be situations wherein the
54 00:07:16.420 ⇒ 00:07:19.709 Advait Nandakumar Menon: Whole field itself isn’t brought into the topic, so…
55 00:07:19.920 ⇒ 00:07:22.810 Advait Nandakumar Menon: this is just a high-level table I prepared for
56 00:07:23.480 ⇒ 00:07:27.209 Advait Nandakumar Menon: And a variety of topics in wholesale and retail.
57 00:07:27.210 ⇒ 00:07:29.970 Shivani Amar: So let’s just go up to that first table for a second.
58 00:07:30.120 ⇒ 00:07:32.519 Shivani Amar: So, already I’m kind of like…
59 00:07:33.140 ⇒ 00:07:35.610 Shivani Amar: If I’m in the seat of Madison.
60 00:07:35.770 ⇒ 00:07:38.940 Shivani Amar: A big thing she’s trying to understand is…
61 00:07:40.610 ⇒ 00:07:44.990 Shivani Amar: Okay, I have these wholesale customers that order to a certain address.
62 00:07:45.560 ⇒ 00:07:47.969 Shivani Amar: And then sometimes they order to another address.
63 00:07:48.240 ⇒ 00:07:55.140 Shivani Amar: And they have a default address tied to their customer profile, but then they also sometimes order to other addresses.
64 00:07:55.310 ⇒ 00:07:58.749 Shivani Amar: So… when I see that addresses
65 00:07:59.050 ⇒ 00:08:01.750 Shivani Amar: Are not currently brought into this topic.
66 00:08:01.920 ⇒ 00:08:13.570 Shivani Amar: And maybe that’s something else, like wholesale customers, wholesale DIM customers, default address, zip phone number, default address. Maybe I don’t need phone numbers, but I think I need the address to be able to say…
67 00:08:14.090 ⇒ 00:08:16.290 Shivani Amar: If I were to say, how many
68 00:08:16.470 ⇒ 00:08:19.500 Shivani Amar: Addresses has this gym shipped to?
69 00:08:20.960 ⇒ 00:08:21.860 Shivani Amar: Right?
70 00:08:22.040 ⇒ 00:08:27.720 Shivani Amar: So, or… Show me, like, what they often want to do is give sampling.
71 00:08:28.770 ⇒ 00:08:35.359 Shivani Amar: So they want to give samples of a new flavor to a certain list of wholesale, partners.
72 00:08:35.850 ⇒ 00:08:37.280 Shivani Amar: Now they could.
73 00:08:38.610 ⇒ 00:08:56.860 Shivani Amar: like, go into Shopify and pull some data, right? But, like, if we’re trying to say that Bobby’s the source of truth, or that Omni’s the source of truth, and people don’t need to really go into other systems, they might say, give me a list of active wholesale partners in California that I can send product to, and then we would want their address.
74 00:08:58.920 ⇒ 00:09:05.840 Shivani Amar: And by the way, give me their most recent shipping address as opposed to their default address, or show me where the delta is between those.
75 00:09:06.370 ⇒ 00:09:13.679 Shivani Amar: So, to me, like, yes, I want their default address, and I want their recent order address.
76 00:09:14.790 ⇒ 00:09:15.500 Advait Nandakumar Menon: Okay.
77 00:09:15.610 ⇒ 00:09:16.480 Advait Nandakumar Menon: Okay.
78 00:09:19.580 ⇒ 00:09:23.800 Advait Nandakumar Menon: Yeah, yeah, there is a shipping address, is that the shipping and billing address?
79 00:09:23.800 ⇒ 00:09:25.050 Shivani Amar: Exactly, so it’s like…
80 00:09:25.050 ⇒ 00:09:25.570 Advait Nandakumar Menon: Yeah.
81 00:09:25.570 ⇒ 00:09:40.430 Shivani Amar: Earlier, that was brought into Blobby, and I was actually able to pull this for Madison, so if you actually see my initial Blobby queries, I was saying, give me a list of wholesale partners, and both their default address and their shipping address. So, I don’t think that this should be something that’s, like, hidden.
82 00:09:41.340 ⇒ 00:09:42.000 Advait Nandakumar Menon: Okay.
83 00:09:42.750 ⇒ 00:09:43.450 Shivani Amar: Okay.
84 00:09:54.210 ⇒ 00:10:03.310 Greg Stoutenburg: Shivani, did Madison have a chance to weigh in on, the questions that we would use to train Blobby, Salty for this, for wholesale?
85 00:10:03.310 ⇒ 00:10:06.669 Shivani Amar: Yeah, all this is coming from conversations I’ve had with Madison.
86 00:10:06.920 ⇒ 00:10:07.400 Greg Stoutenburg: Okay.
87 00:10:07.400 ⇒ 00:10:08.669 Shivani Amar: Everything I’m saying.
88 00:10:08.890 ⇒ 00:10:16.049 Greg Stoutenburg: Oh, no, sorry, I mean, when we were initially saying, like, some of the questions that we would treat as success criteria for training up…
89 00:10:16.050 ⇒ 00:10:33.649 Shivani Amar: Actually, like, she actually said, like, is a fridge effective? Like, from the beginning, when we were talking to Brainforge, that’s why in the Google Sheet version, we have a field that says, do they have a fridge or not? Because she’s like, I want to know if they have a fridge. So this is, like, even if I didn’t put it in that document with the questions, it’s like, Brainforge folks who were…
90 00:10:33.740 ⇒ 00:10:47.529 Shivani Amar: institutionally knowledge, like, aware of, like, talking to Madison. They’re like, do they have a fridge, do they not have a fridge? I want to see their address, I want to see their zip code. Like, all of that has been abundantly clear that it’s, like, something that the team needs.
91 00:10:48.040 ⇒ 00:10:48.530 Greg Stoutenburg: Okay.
92 00:10:48.530 ⇒ 00:10:56.440 Shivani Amar: from the Google Sheet version through to Omni, I don’t think I needed… I didn’t think I needed to re-articulate it, but if that’s the level of, like.
93 00:10:57.110 ⇒ 00:11:16.509 Shivani Amar: if it’s, like, we need… there’s, like, context that gets said in calls, and then there’s context that’s, like, from a source of truth document that’s used really… you know, it’s like, that’s what I’m feeling about that set of questions. So I’m like, if that’s what it is, then I need to be super thorough about that set of questions, otherwise we’re gonna always have this back and forth.
94 00:11:17.370 ⇒ 00:11:24.920 Jasmin Multani: Yeah, so right now, this is what we have, across… those tabs for wholesale.
95 00:11:24.920 ⇒ 00:11:25.550 Shivani Amar: Yeah.
96 00:11:25.760 ⇒ 00:11:43.859 Jasmin Multani: And I see… I do see fridges in number 6, so that makes sense. And then, when I think of who are our biggest customers, I am thinking about addresses as well. So, because we’re using these questions as our success criteria to QA,
97 00:11:44.040 ⇒ 00:11:46.720 Jasmin Multani: The relationship between topic to blobby.
98 00:11:46.960 ⇒ 00:11:54.179 Shivani Amar: But, like, if I say, who are my biggest customers, and you just give me a list of names, inevitably, I’m gonna be like, where are they based?
99 00:11:55.070 ⇒ 00:12:02.849 Shivani Amar: So, I don’t know how to tell you… in every cut of every data that we can have, if you can cut it by zip code, I want it by zip code.
100 00:12:03.170 ⇒ 00:12:10.800 Shivani Amar: Like, that’s a grain that I care about, because eventually you want to say, how am I performing across zip codes? Not just across states.
101 00:12:11.070 ⇒ 00:12:23.930 Shivani Amar: So, I don’t want to have to write that in every question. That’s like, by the way, I’m going to want this cut by zip code, by the way, I’m going to want this cut by zip code. I want you guys to all just internalize that today. That’s like, if I see zip code data, I should make sure that it’s available.
102 00:12:24.650 ⇒ 00:12:25.180 Jasmin Multani: Okay.
103 00:12:25.180 ⇒ 00:12:34.389 Greg Stoutenburg: Yeah, and we can do that, and we can have something like a list of desired cuts across any possible tables, and make sure that that’s… that’s present wherever possible, so we can definitely do that. Cool.
104 00:12:34.550 ⇒ 00:12:53.209 Greg Stoutenburg: So, the reason that I wanted to bring up the question specifically is just because, you know, we want to make sure that we’re including all the right things so that when you do go in and ask questions, you know, as you suggested, Madison is going to treat the question about addresses as a very natural one. We want to make sure that we’re getting that right in the first place.
105 00:12:54.050 ⇒ 00:13:00.420 Shivani Amar: So, can you go to the Brain Forge tab, potential questions by team?
106 00:13:01.950 ⇒ 00:13:03.870 Shivani Amar: Like, the, the brain force…
107 00:13:04.020 ⇒ 00:13:05.649 Greg Stoutenburg: Data platform documentation.
108 00:13:05.650 ⇒ 00:13:09.519 Shivani Amar: Yeah, and go to the sheet that says potential questions by team.
109 00:13:12.020 ⇒ 00:13:16.790 Shivani Amar: I don’t know which one is the source of truth, but there’s one that says potential question. It’s to the right.
110 00:13:16.790 ⇒ 00:13:17.430 Jasmin Multani: Yeah.
111 00:13:17.430 ⇒ 00:13:19.100 Shivani Amar: Next to Dashboard Specs.
112 00:13:19.660 ⇒ 00:13:30.149 Shivani Amar: Yeah, you see that? Yep. What are the typical drop-offs between orders placed? Sorry, okay, no. What impact do merchandising assets, e.g. fridges, have on partner sales?
113 00:13:30.150 ⇒ 00:13:30.610 Jasmin Multani: Yes.
114 00:13:30.610 ⇒ 00:13:33.970 Greg Stoutenburg: Yeah, no, you’re right, Bridges is there. You’re right. No, you’re right about that.
115 00:13:33.970 ⇒ 00:13:45.889 Shivani Amar: So I don’t know which file you’re using, but I’m like, I’ve talked about fridges, I don’t know which file you guys are all referencing to say, like, we checked the box, we didn’t, but like, it’s not a new concept.
116 00:13:47.110 ⇒ 00:13:55.590 Shivani Amar: Which is fine, but you tell me which tab to pay attention to. Is dashboard specs the tab, or is potential questions by team?
117 00:13:55.800 ⇒ 00:14:04.009 Greg Stoutenburg: Yeah, this is… no, Shivani, you’re right. This is… 9… 9B, this is… this should have been accounted for. Fridges should have been accounted for.
118 00:14:04.010 ⇒ 00:14:10.129 Jasmin Multani: It’s… it’s, it’s when we… It is part of it.
119 00:14:10.510 ⇒ 00:14:18.080 Jasmin Multani: it’s just, I think we had to make the trade-off between what we want Bloppy to be exposed to as we’re developing out the dashboards.
120 00:14:18.210 ⇒ 00:14:26.829 Jasmin Multani: But this is still part of the QA crush, success question criterias.
121 00:14:28.050 ⇒ 00:14:29.089 Shivani Amar: Okay, so…
122 00:14:29.090 ⇒ 00:14:33.429 Jasmin Multani: This is… there’s… this is definitely part of the roadmap.
123 00:14:33.430 ⇒ 00:14:39.229 Shivani Amar: Well, I was like, if you guys have forgotten about this tab, I’m like, this tab exists, I could add.
124 00:14:39.230 ⇒ 00:14:39.830 Greg Stoutenburg: I love this.
125 00:14:39.830 ⇒ 00:14:40.850 Shivani Amar: Good job.
126 00:14:40.850 ⇒ 00:14:43.420 Greg Stoutenburg: I love this tab. I wanted this tab because
127 00:14:43.560 ⇒ 00:15:01.079 Greg Stoutenburg: well, you know, like I said the other day, in principle, you could ask Bobby a literally infinite number of different questions about even one and the same table. So, you know, that’s why we want to have it narrow and clean, and at the same time, as we’re working on QA, you know, if we want to have something like.
128 00:15:01.220 ⇒ 00:15:14.300 Greg Stoutenburg: a set of cuts that apply everywhere, you know, we can do that as well. Cool. But I’ll, yeah, I’ll back away from the bike, because Jasmine’s got some great ideas about how we can approach, QA for topics in general.
129 00:15:14.300 ⇒ 00:15:17.900 Shivani Amar: Okay, cool. Jasmine, I had to request access to whatever you just dropped in the chat.
130 00:15:19.300 ⇒ 00:15:20.260 Shivani Amar: Okay.
131 00:15:22.290 ⇒ 00:15:25.150 Jasmin Multani: Okay, editing. There you go.
132 00:15:29.990 ⇒ 00:15:31.879 Jasmin Multani: Yes, you should have access by now.
133 00:15:33.120 ⇒ 00:15:35.500 Jasmin Multani: But…
134 00:15:36.310 ⇒ 00:15:46.569 Jasmin Multani: This is just a primer. It’s just a Brainforge-facing primer of what our success… okay, what are our core questions we want to be answered by Blobby?
135 00:15:46.570 ⇒ 00:15:58.159 Jasmin Multani: Through, the retail topics and through the wholesale topics. And as you scroll down, there are gonna be break metrics that’s like, hey, if these aren’t answered.
136 00:15:58.900 ⇒ 00:16:05.310 Jasmin Multani: Like, we pull the dashboard immediately, and, there’s something broken, like, egregiously broken.
137 00:16:05.540 ⇒ 00:16:14.779 Jasmin Multani: And from there, success metrics go into, like, the quality of the answers, right? Like, how often is Salty actually answering?
138 00:16:15.260 ⇒ 00:16:28.979 Jasmin Multani: giving back a question rather than saying, hey, I don’t really know. What is the consistency? Like, right, if, 5 different stakeholders ask that fridge question different ways, is it giving the same answer?
139 00:16:29.220 ⇒ 00:16:43.530 Shivani Amar: And we’ll know so much more about the assessing adoption once we actually roll this out, so I like this framework a lot, and I’m looking forward to, like, when Omni’s actually in the fall in the hands of a lot of people in the business, then we can, like.
140 00:16:43.530 ⇒ 00:16:49.679 Shivani Amar: have weekly touchpoints where we’re looking at, like, okay, like, how did Blobby perform, right?
141 00:16:50.370 ⇒ 00:16:51.130 Shivani Amar: Okay.
142 00:16:51.130 ⇒ 00:17:06.009 Jasmin Multani: And the idea is, in the next two weeks, as we march towards May 22nd, we’re gonna… we have this template, and it’s gonna be under operationalizing evaluation. Whatever we do internally, that’s what you scale out in fall.
143 00:17:08.240 ⇒ 00:17:09.969 Shivani Amar: Okay, perfect.
144 00:17:10.200 ⇒ 00:17:23.090 Jasmin Multani: Right now, I just wanted… I was like, sir, this is a Brain Forge-facing doc right now, but I just wanted to show you that these things are being accounted for. Yeah. But I hear you, in what you said, hey.
145 00:17:23.109 ⇒ 00:17:32.480 Jasmin Multani: The level of granularity, that should be something that we take ownership on, and focus on building consistency, rather than having you, like.
146 00:17:32.820 ⇒ 00:17:40.099 Jasmin Multani: be so, so nitty-gritty. I think that’s something, we were, like, playing hot potato with, but… Yeah. Cool.
147 00:17:40.210 ⇒ 00:17:44.980 Jasmin Multani: But the core questions, like, we have visibility into. It’s just the grains,
148 00:17:45.570 ⇒ 00:17:47.460 Jasmin Multani: And the trade-offs of those grains.
149 00:17:47.460 ⇒ 00:17:48.150 Shivani Amar: Okay.
150 00:17:48.600 ⇒ 00:17:51.899 Jasmin Multani: Cool, cool, cool. And so…
151 00:17:52.350 ⇒ 00:18:05.700 Jasmin Multani: I was able to get Greg’s feedback, and I feel good about setting up the actual operations. And once we clear out these dashboards, these three dashboards, I can kickstart the QAing.
152 00:18:06.720 ⇒ 00:18:08.080 Jasmin Multani: For the answers.
153 00:18:09.130 ⇒ 00:18:10.730 Shivani Amar: Okay, that sounds great.
154 00:18:12.500 ⇒ 00:18:18.429 Jasmin Multani: Back to Advait’s exposure.
155 00:18:19.030 ⇒ 00:18:23.679 Jasmin Multani: not the 107-page document, but the Shivani-facing one.
156 00:18:23.930 ⇒ 00:18:32.729 Jasmin Multani: Shivani, do you still want to go row by row, or do you just want to look at this async and give us high-level direction on this?
157 00:18:33.740 ⇒ 00:18:36.360 Shivani Amar: Let’s see, so…
158 00:18:40.590 ⇒ 00:18:47.759 Shivani Amar: Can we just go down further? Because I just… I give you some… now, I’ve given you, like, high-level detail around what I’d want, but, like, I’m just seeing if anything else…
159 00:18:48.500 ⇒ 00:18:57.350 Shivani Amar: Catches my eye. So, default address we just talked about. Is there a difference between the default address here than what we saw up there?
160 00:18:57.750 ⇒ 00:19:01.260 Advait Nandakumar Menon: No, it’s the same table, it’s two different topics, yeah.
161 00:19:02.830 ⇒ 00:19:06.689 Shivani Amar: So… Does it need to be in both?
162 00:19:10.610 ⇒ 00:19:30.250 Advait Nandakumar Menon: So, when you are querying blobby, you will select a topic, so as long as you select the right topic, it need not be included in both places. So, if you’re expecting to see default address and the billing address or whatever in one topic, and in the other, it’s not
163 00:19:30.360 ⇒ 00:19:35.979 Advait Nandakumar Menon: expected to be seen, we can avoid exposing lobby to that field and that topic.
164 00:19:36.210 ⇒ 00:19:36.990 Shivani Amar: Hmm.
165 00:19:45.360 ⇒ 00:19:46.940 Shivani Amar: Ochre…
166 00:19:57.400 ⇒ 00:20:05.990 Shivani Amar: It’s like, they should know that the topic is more specific than wholesale, like, they should know that the topic is wholesale customer versus wholesale orders.
167 00:20:08.810 ⇒ 00:20:11.240 Advait Nandakumar Menon: Yeah, I would say that’s part of…
168 00:20:11.520 ⇒ 00:20:15.710 Advait Nandakumar Menon: the whole blobby and Omni experience, because,
169 00:20:16.200 ⇒ 00:20:24.400 Advait Nandakumar Menon: when you don’t specify a select topic, when you’re querying Blobby, it can lead to, like, non-deterministic answers, like.
170 00:20:24.560 ⇒ 00:20:31.500 Advait Nandakumar Menon: it can go sideways, so it’s always better… that’s what Omni has recommended as well, it’s always better to…
171 00:20:31.600 ⇒ 00:20:46.460 Advait Nandakumar Menon: Select a topic and be as specific as possible, but yes, the user must be given instructions or prior knowledge about this, like which topic and whatever should be selected, so that, can be, something that we offer as well.
172 00:20:47.280 ⇒ 00:20:47.920 Shivani Amar: Okay.
173 00:20:52.280 ⇒ 00:21:11.629 Shivani Amar: Yeah, I’m just trying to figure out, like, how we would train… maybe I’m getting ahead of myself, but, like, let’s say that, like, the address is in the customer topic, but not in the orders topic, or for some reason, or whatever, then, like, we don’t have to think about this too deeply right now, but I’m like, okay, like, then Laura and Madison would need to know to try a different angle to get the information they need, right?
174 00:21:12.320 ⇒ 00:21:14.230 Jasmin Multani: Okay. Yeah.
175 00:21:14.230 ⇒ 00:21:14.970 Shivani Amar: Okay.
176 00:21:16.180 ⇒ 00:21:23.110 Jasmin Multani: There should be, like, a troubleshooting FAQ, that we also provide for these dashboards and… Okay.
177 00:21:23.110 ⇒ 00:21:31.589 Shivani Amar: So, let’s keep scrolling, because we’re still on this address thing. So, not currently brought in Omni risk level, discount amount, that seems fine.
178 00:21:31.810 ⇒ 00:21:37.999 Shivani Amar: Okay, then shipping address, we just talked about, yes, we want shipping addresses and default addresses.
179 00:21:38.780 ⇒ 00:21:47.090 Shivani Amar: then… Full name, phone, Default address, phone addresses.
180 00:21:47.200 ⇒ 00:21:56.739 Shivani Amar: We’ve talked about this. See, like, to me, I’m like, I don’t know which one it should be in, but it should be, like, we should be feeding addresses and…
181 00:21:56.910 ⇒ 00:22:06.430 Shivani Amar: Of default address and shipping address in some capacity, I just don’t know which one, okay? So…
182 00:22:06.790 ⇒ 00:22:07.360 Advait Nandakumar Menon: Okay.
183 00:22:09.040 ⇒ 00:22:13.110 Shivani Amar: It feels like it’s not needed in the dashboard, in the executive pulse, I agree.
184 00:22:13.630 ⇒ 00:22:15.350 Shivani Amar: Right? That’s fine.
185 00:22:29.760 ⇒ 00:22:37.919 Shivani Amar: Like, okay, so let’s talk about the funnel for a second. So, let’s say I’m like, wait, of the people
186 00:22:38.130 ⇒ 00:22:39.840 Shivani Amar: Who churned?
187 00:22:41.620 ⇒ 00:22:47.360 Shivani Amar: last week or something, right? You’re like, okay, like, I had these 10 wholesale partners churn last week.
188 00:22:48.740 ⇒ 00:22:53.490 Shivani Amar: I want to… Send them samples.
189 00:22:53.800 ⇒ 00:22:57.210 Shivani Amar: Something about… then… then I’d want their address, so, like.
190 00:22:57.540 ⇒ 00:23:15.120 Shivani Amar: I don’t… we can, like, play through this later, I think. But, like, I think… I’m just trying to play… like, give you an example, like, okay, like, I might go from looking at a funnel to wanting to take action about the people who are at risk, and then I might want to double-click into who are the people that are at risk, how do I contact them?
191 00:23:23.070 ⇒ 00:23:29.030 Greg Stoutenburg: I think maybe here, the direction could be to just clarify what the purpose of each topic is.
192 00:23:29.330 ⇒ 00:23:41.929 Greg Stoutenburg: and which tables or dashboards they’re going to be used for, because that’s going to be, I think that’s gonna be just the… help us put in the right container, conceptually, where the questions are going to be going to.
193 00:23:42.340 ⇒ 00:23:43.570 Shivani Amar: It sounds good.
194 00:23:44.400 ⇒ 00:23:47.140 Greg Stoutenburg: Yeah, so why don’t we make that revision, team?
195 00:23:48.100 ⇒ 00:23:49.460 Greg Stoutenburg: And,
196 00:23:49.980 ⇒ 00:24:04.450 Greg Stoutenburg: Yeah, and then with that, I think maybe the next step can be, Shivani, you take a look at it and, you know, run that simulation of, like, which questions would I want to ask here? And we can make sure that we’re, we’re doing this in the right way, so it’ll answer their questions correctly.
197 00:24:05.930 ⇒ 00:24:06.500 Advait Nandakumar Menon: Yep.
198 00:24:06.740 ⇒ 00:24:07.480 Shivani Amar: Great.
199 00:24:17.050 ⇒ 00:24:25.399 Advait Nandakumar Menon: Yeah, for, retail, so we haven’t really hidden anything from Blobby, it’s all open.
200 00:24:25.620 ⇒ 00:24:38.189 Advait Nandakumar Menon: Today, so I want to understand from you, like, have you ever come across any situations where the data is modeled, but Blobby is not able to answer, like, how you came across wholesale for fridges, and…
201 00:24:38.600 ⇒ 00:24:39.120 Advait Nandakumar Menon: Oh, sweet.
202 00:24:39.120 ⇒ 00:24:46.469 Shivani Amar: I’ll play… I can play around with it more, but I, like, I probably haven’t… I probably could dedicate, like.
203 00:24:47.180 ⇒ 00:25:04.329 Shivani Amar: like, if this is what we want, is, like, for me to test case these things, then I can, like, dedicate a couple hours to, like, just wholesale topic next week in Blobby and stuff like that, which is fine. I think I gave you guys the example of what Madison would want to know. She’s like, okay.
204 00:25:04.780 ⇒ 00:25:09.980 Shivani Amar: These are some partners. I want to know how many places they’ve shipped to. I want to know if they’ve ever ordered
205 00:25:10.150 ⇒ 00:25:14.810 Shivani Amar: sparkling. I wanna know if they… Have fridges or not.
206 00:25:15.300 ⇒ 00:25:20.060 Shivani Amar: And I think that will give you, like, a very good test case to, like, play with it yourself, also.
207 00:25:21.410 ⇒ 00:25:21.910 Greg Stoutenburg: I like that.
208 00:25:21.910 ⇒ 00:25:22.540 Advait Nandakumar Menon: Yep.
209 00:25:23.470 ⇒ 00:25:36.189 Greg Stoutenburg: Yeah, and so, team, let’s provide that documentation around what the purpose of each topic is, so we have something nice and helpful to hand over to Shivani. Shivani, we don’t want you to have to do this entire lift of going through all this,
210 00:25:36.340 ⇒ 00:25:40.349 Greg Stoutenburg: And so much detail first, we want to at least set the stage in the way that we can.
211 00:25:41.340 ⇒ 00:25:43.100 Shivani Amar: Okay, that sounds good.
212 00:25:43.100 ⇒ 00:25:43.690 Advait Nandakumar Menon: Yep.
213 00:25:44.060 ⇒ 00:25:48.320 Shivani Amar: I know we only have a few minutes left. Are there any, like, dashboards you were hungry for me to look at?
214 00:25:49.230 ⇒ 00:25:51.920 Greg Stoutenburg: I’ll have to drop in 4 minutes, team, but you guys can carry on.
215 00:25:52.300 ⇒ 00:26:03.609 Jasmin Multani: I’d say the wholesale executive. I think that was the one where we wanted to have clarity on, granularity. Am I remembering this right?
216 00:26:04.880 ⇒ 00:26:05.580 Advait Nandakumar Menon: Yep.
217 00:26:14.520 ⇒ 00:26:30.450 Jasmin Multani: Yeah, so just as an FYI, when we tried migrating the Google Sheets over to OmniSheets, that file was so big that it ended up breaking, and it just… the migration just wasn’t happening. So, I wanted to ask, like.
218 00:26:30.680 ⇒ 00:26:40.719 Jasmin Multani: you know, we have, like, weekly, monthly, and quarterly grains. Which ones do you explicitly want visuals for? Which ones do you want explicitly, like.
219 00:26:41.240 ⇒ 00:26:43.240 Jasmin Multani: Dashboard Tables for.
220 00:26:44.410 ⇒ 00:26:45.220 Shivani Amar: Hmm.
221 00:26:46.230 ⇒ 00:26:49.800 Jasmin Multani: And you can go, like, in the Google Sheets and literally be, like, rows…
222 00:26:50.240 ⇒ 00:26:56.190 Jasmin Multani: 10 to 12 need to have a visual. Okay. Because, the lift also is…
223 00:26:56.310 ⇒ 00:27:05.369 Jasmin Multani: Some of this information is being sourced from different topics, so that’s gonna be a lift that we have to calculate, and we’re trying to figure out, like, what is…
224 00:27:06.420 ⇒ 00:27:08.270 Jasmin Multani: The best way to package those.
225 00:27:08.270 ⇒ 00:27:09.760 Shivani Amar: I see, okay.
226 00:27:10.260 ⇒ 00:27:12.880 Shivani Amar: Oh… can you scroll down?
227 00:27:15.180 ⇒ 00:27:24.960 Shivani Amar: Wherever you can go with, like, shades of element colors, I think that’s, like, better. I know you only have this, like, black and green, but if you’re confused by other shades, we can come up with, like, a palette.
228 00:27:25.350 ⇒ 00:27:27.799 Shivani Amar: Then the order funnel drop off.
229 00:27:27.800 ⇒ 00:27:30.420 Advait Nandakumar Menon: I did pull all of these from the…
230 00:27:30.890 ⇒ 00:27:36.090 Advait Nandakumar Menon: The PDF, the… you said a couple of weeks back, I pulled it from that.
231 00:27:36.090 ⇒ 00:27:44.450 Shivani Amar: Gotcha. Order funnel drop-off, that visual is very weird to me. But we can… we can come up with something else.
232 00:27:46.030 ⇒ 00:27:48.600 Shivani Amar: I was just imagining it more like.
233 00:27:49.150 ⇒ 00:28:01.020 Shivani Amar: first order, second order is a row below that third order. It doesn’t… it could just be a table. It could be a weekly table, like, a monthly table of, like, how many people are making their first, second, third, or it can be…
234 00:28:01.570 ⇒ 00:28:08.270 Shivani Amar: Or it can just be, like, accumu- if you want it to be a cumulative. Like, what is it right now? What is 2000? It’s all time.
235 00:28:09.410 ⇒ 00:28:10.120 Jasmin Multani: Given the.
236 00:28:10.120 ⇒ 00:28:13.589 Advait Nandakumar Menon: It’s for the past, yeah, 12 months.
237 00:28:14.700 ⇒ 00:28:15.580 Shivani Amar: Okay.
238 00:28:17.420 ⇒ 00:28:34.890 Jasmin Multani: And, yeah, the third order one was kind of confusing when we looked into it initially, but my gut tells me is that the way it’s calculated, it’s saying that, hey, let’s say it’s the past two months, that third order, they’re just in different cohorts.
239 00:28:34.980 ⇒ 00:28:39.640 Jasmin Multani: So, in the past month, it’s like, obviously your first order
240 00:28:39.870 ⇒ 00:28:42.550 Jasmin Multani: Is by someone who made that first order in the past month.
241 00:28:42.730 ⇒ 00:28:49.040 Jasmin Multani: this past month, someone who made a third order, their first order was probably in January.
242 00:28:49.040 ⇒ 00:28:49.989 Shivani Amar: That’s totally okay.
243 00:28:49.990 ⇒ 00:28:51.049 Jasmin Multani: Okay, cool, cool, cool.
244 00:28:51.050 ⇒ 00:28:57.339 Shivani Amar: It’s just like a… Over… in a given period, how many people are doing this activity?
245 00:28:57.620 ⇒ 00:29:16.160 Shivani Amar: Which just gives you a feel for, like, oh, what the drop-offs tend to be over time. Like, it doesn’t… nothing about this needs to be cohorted, in my opinion. And the VP of wholesale might disagree eventually, and be like, I want to look at it by cohort, but I’m like, I just want to know the shape of the thing. So, if what I see is right, we had…
246 00:29:16.870 ⇒ 00:29:19.690 Shivani Amar: A really high second order rate.
247 00:29:21.540 ⇒ 00:29:22.800 Jasmin Multani: 2,000, yeah.
248 00:29:23.020 ⇒ 00:29:31.430 Shivani Amar: And then I don’t even understand how the third order rate… yes, I get that it’s, like, more because it’s not cohorted, but I’m like, that’s kind of wild to me. I kind of don’t trust this data.
249 00:29:32.790 ⇒ 00:29:34.260 Jasmin Multani: Mmm, okay.
250 00:29:34.260 ⇒ 00:29:36.220 Greg Stoutenburg: I better drop. Thanks all. I’ll catch up after.
251 00:29:36.220 ⇒ 00:29:36.840 Shivani Amar: Bye.
252 00:29:36.840 ⇒ 00:29:37.159 Greg Stoutenburg: Have a good one.
253 00:29:38.510 ⇒ 00:29:48.190 Jasmin Multani: it looked… mmm… okay, we’ll cut it again. I think when I initially saw it and validated directionally, it was similar to what
254 00:29:48.530 ⇒ 00:29:55.179 Jasmin Multani: I had seen in the Google Sheets, but we’ll cut it again, and I’d say we’ll adapt it to this
255 00:29:55.880 ⇒ 00:29:56.500 Jasmin Multani: active.
256 00:29:56.500 ⇒ 00:30:01.539 Shivani Amar: I don’t even know if I trust the Google Sheet version, to be honest, because I’m like…
257 00:30:01.700 ⇒ 00:30:05.589 Shivani Amar: Okay, if I were to say… let me just, like, look at a week.
258 00:30:06.210 ⇒ 00:30:07.080 Shivani Amar: Okay.
259 00:30:07.210 ⇒ 00:30:09.670 Shivani Amar: Let’s just do this real quick.
260 00:30:10.050 ⇒ 00:30:21.249 Shivani Amar: So… My understanding is that If this is correct, 47 wholesale partners, let’s look at last…
261 00:30:21.390 ⇒ 00:30:24.580 Shivani Amar: Week, okay? The week ending 4-24.
262 00:30:24.750 ⇒ 00:30:28.819 Shivani Amar: Okay? 35 wholesale partners made a first order that week.
263 00:30:29.560 ⇒ 00:30:34.469 Shivani Amar: 37 wholesale partners made a second order. They placed an order for the second time that week.
264 00:30:35.720 ⇒ 00:30:40.010 Shivani Amar: 35 wholesale partners made a third order that week. I wanted to verify that.
265 00:30:40.830 ⇒ 00:30:41.530 Jasmin Multani: Okay.
266 00:30:41.530 ⇒ 00:30:51.049 Shivani Amar: And, like, see that in the data, and then I’m like, okay, are we def… are we defining these things correctly? Which is, if it’s correct, then that’s really cool. Like, we’ve got great retention.
267 00:30:52.360 ⇒ 00:30:53.550 Jasmin Multani: Okay, okay, okay.
268 00:30:54.480 ⇒ 00:30:55.880 Shivani Amar: But I just don’t trust it.
269 00:30:56.170 ⇒ 00:31:02.819 Jasmin Multani: Okay, we’ll do another validation check on this, through a different angle. We’ll probably go through the raw data.
270 00:31:03.820 ⇒ 00:31:08.969 Jasmin Multani: But okay, I hear you that the shape is not, up to standard?
271 00:31:09.430 ⇒ 00:31:11.169 Jasmin Multani: Let’s get the.
272 00:31:11.170 ⇒ 00:31:25.260 Shivani Amar: Let’s get the data right first, let’s get the is, and then I can tell you what I think the data should… the visual should be. Once… like, let’s actually aim to fix this in the Google Sheet, and then if you’re like, yes, this is 100% what it is, then I’ll… we can talk about what the visual should be.
273 00:31:25.750 ⇒ 00:31:27.260 Jasmin Multani: Okay, sounds good.
274 00:31:27.260 ⇒ 00:31:30.170 Shivani Amar: Would it be worth it of it? Like, do you, like…
275 00:31:30.320 ⇒ 00:31:32.039 Shivani Amar: I don’t know who would show me, like.
276 00:31:32.510 ⇒ 00:31:39.870 Shivani Amar: how can you show me all the orders that happened last week, and if it was their first, second, or third order? Like, can you actually show me that?
277 00:31:42.050 ⇒ 00:31:42.860 Advait Nandakumar Menon: So…
278 00:31:42.860 ⇒ 00:31:43.980 Shivani Amar: in Snowflake?
279 00:31:44.940 ⇒ 00:31:46.300 Advait Nandakumar Menon: Oh, and Snowflake?
280 00:32:18.030 ⇒ 00:32:18.740 Shivani Amar: Oh, God.
281 00:32:56.910 ⇒ 00:32:58.990 Advait Nandakumar Menon: What’s the week’s start date you mentioned?
282 00:33:01.080 ⇒ 00:33:05.320 Jasmin Multani: Let’s do 2026 minus 03.
283 00:33:06.200 ⇒ 00:33:07.690 Jasmin Multani: Minus 24.
284 00:33:07.970 ⇒ 00:33:09.070 Jasmin Multani: Is that one of them?
285 00:33:09.070 ⇒ 00:33:19.770 Shivani Amar: Oh, sorry, sorry, sorry. So, let’s see, you want to do the week, it should be week end date, so week ending 4-24.
286 00:33:20.810 ⇒ 00:33:21.370 Advait Nandakumar Menon: Okay.
287 00:33:43.840 ⇒ 00:33:45.550 Jasmin Multani: Okay, so this is saying…
288 00:33:46.670 ⇒ 00:33:47.690 Advait Nandakumar Menon: Yeah…
289 00:33:48.320 ⇒ 00:33:48.720 Jasmin Multani: what fifth.
290 00:33:48.720 ⇒ 00:33:55.570 Shivani Amar: Let’s just stay in the Snowflake, guys. Let’s stay in the Snowflake. There’s nothing else to do. Let’s just stay in Snowflake. Just do the sequel.
291 00:33:55.570 ⇒ 00:34:02.140 Advait Nandakumar Menon: I’m trying to figure out the week start date, because the week end date is not a field in Snowflake, it’s something that’s…
292 00:34:02.450 ⇒ 00:34:03.000 Advait Nandakumar Menon: Yeah.
293 00:34:03.000 ⇒ 00:34:08.310 Shivani Amar: If that’s the week end date, then the week start date is 4-18.
294 00:34:09.179 ⇒ 00:34:10.209 Advait Nandakumar Menon: Okay, okay.
295 00:34:10.370 ⇒ 00:34:11.130 Shivani Amar: Okay?
296 00:34:11.130 ⇒ 00:34:12.250 Advait Nandakumar Menon: That happens, yep.
297 00:34:30.420 ⇒ 00:34:37.860 Shivani Amar: So this is interesting, I actually want to see the sequel happen from the, like, orders table, and then understand how, like.
298 00:34:38.130 ⇒ 00:34:39.380 Advait Nandakumar Menon: You mean the raw data?
299 00:34:39.389 ⇒ 00:34:44.069 Shivani Amar: Yeah, because I’m like, how are you labeling if something was a second order? I don’t even know if I trust those labels.
300 00:34:57.499 ⇒ 00:35:01.259 Shivani Amar: And, like, to do that, you would have to know… like, you would have to have…
301 00:35:01.829 ⇒ 00:35:08.419 Shivani Amar: Your orders table would have to have when… like, This customer placed an order.
302 00:35:08.799 ⇒ 00:35:11.279 Shivani Amar: Their previous orders were these dates.
303 00:35:12.270 ⇒ 00:35:18.740 Jasmin Multani: It’ll rank it. There’s a command that’ll rank it. It’s literally, out of… The entire order history.
304 00:35:18.840 ⇒ 00:35:28.310 Jasmin Multani: who is the person, what was the time of their order creation, and then there’s a SQL command that literally ranks it 1234, or you can go in reverse.
305 00:35:28.990 ⇒ 00:35:33.770 Jasmin Multani: And you can do, like, moving averages as well.
306 00:35:34.820 ⇒ 00:35:40.719 Jasmin Multani: But I get that, like, we should double-check that the labeling and the ranking is being done accurately.
307 00:35:41.470 ⇒ 00:35:47.850 Advait Nandakumar Menon: Yeah, so behind that summary table you just saw, there are a couple of SQL steps and logics.
308 00:35:48.120 ⇒ 00:35:51.229 Advait Nandakumar Menon: That’s being implemented to come up with that value, so…
309 00:35:51.340 ⇒ 00:36:01.349 Advait Nandakumar Menon: Couple of the things Jasmine just mentioned, so we have to just backtrack and check whether it’s happening at the right, level, and whether it’s labeling it correctly.
310 00:36:04.150 ⇒ 00:36:09.290 Jasmin Multani: Can you, send this… Over to me, please. And then,
311 00:36:09.690 ⇒ 00:36:13.579 Jasmin Multani: We’ll go through together to see what the upstream calculations are.
312 00:36:14.950 ⇒ 00:36:15.590 Advait Nandakumar Menon: Sure.
313 00:36:20.680 ⇒ 00:36:22.979 Jasmin Multani: But then I also wanted to know, like,
314 00:36:23.100 ⇒ 00:36:27.180 Jasmin Multani: Advit, did you set up these calculations upstream, or was it always…
315 00:36:28.550 ⇒ 00:36:30.759 Advait Nandakumar Menon: This floor is set up by Avish.
316 00:36:31.900 ⇒ 00:36:32.900 Jasmin Multani: Okay.
317 00:36:33.910 ⇒ 00:36:38.869 Jasmin Multani: And we can make edits through the GitHub, or would the edits be done through Snowflake?
318 00:36:39.560 ⇒ 00:36:49.610 Advait Nandakumar Menon: So DBT is involved in this, so Avesh will have to make the edits in the dbt model, which will be pushed to GitHub and then Snowflake.
319 00:36:49.800 ⇒ 00:37:03.790 Shivani Amar: This is the level of weeds I don’t mind us going in together, so I hope you guys don’t feel like this is weird. Like, I’m used to sitting with people while they go into SQL, right? But, like, I’m like, I don’t trust that my funnel is, like, there’s no drop-off. I don’t trust that.
320 00:37:05.080 ⇒ 00:37:19.309 Shivani Amar: Because there’s been churn of customers, so it’s like, to see that, like, everybody who places a first order eventually places a third order, like, that just feels weird. I’m sure there are plenty of people that place one order and never order again.
321 00:37:19.820 ⇒ 00:37:22.910 Jasmin Multani: Oh, I think he’s doing it by count. Scroll up.
322 00:37:25.610 ⇒ 00:37:31.519 Jasmin Multani: count of… okay, count… third orders count… is it… oh, this is a summary, though.
323 00:37:33.970 ⇒ 00:37:42.329 Advait Nandakumar Menon: Yeah, so… The summary table SQL is where… it’s showing from where it’s pulling the data.
324 00:37:42.780 ⇒ 00:37:43.580 Jasmin Multani: Okay.
325 00:37:43.580 ⇒ 00:37:44.960 Advait Nandakumar Menon: These are the steps.
326 00:37:48.510 ⇒ 00:37:54.560 Jasmin Multani: You know, lower segment, inner joint, daily joins… can we scroll?
327 00:38:00.170 ⇒ 00:38:05.289 Jasmin Multani: At risk accounts created. Can you send this link over to my Slack, please?
328 00:38:26.460 ⇒ 00:38:27.810 Jasmin Multani: Just go.
329 00:38:37.870 ⇒ 00:38:42.129 Advait Nandakumar Menon: I think… It is at line 235, the logic.
330 00:38:42.740 ⇒ 00:38:44.080 Jasmin Multani: 2.30.
331 00:38:44.860 ⇒ 00:38:50.119 Advait Nandakumar Menon: It starts at 2.35, and once below it.
332 00:38:51.050 ⇒ 00:38:57.129 Jasmin Multani: 2, 3, orders with sequence. Yeah, he’s doing a row number, over partition.
333 00:38:57.130 ⇒ 00:38:57.480 Advait Nandakumar Menon: Yeah.
334 00:38:57.550 ⇒ 00:38:59.150 Jasmin Multani: Water created a…
335 00:39:04.530 ⇒ 00:39:07.330 Jasmin Multani: Or order created at is less than current day.
336 00:39:09.000 ⇒ 00:39:15.909 Jasmin Multani: And then, maybe it’s in the joins, where things get messy. Casement order sequence equals 1, 2, and 3.
337 00:39:31.000 ⇒ 00:39:36.930 Jasmin Multani: I think that’s the issue. My gut tells me there’s something going on with 255.
338 00:39:39.700 ⇒ 00:39:42.020 Jasmin Multani: But I might be wrong, yes.
339 00:39:42.440 ⇒ 00:39:44.330 Shivani Amar: I don’t trust that SQL at all.
340 00:39:44.580 ⇒ 00:39:47.440 Jasmin Multani: this is…
341 00:39:47.440 ⇒ 00:39:49.329 Shivani Amar: That’s why we do what we do, guys.
342 00:39:50.060 ⇒ 00:39:54.469 Jasmin Multani: Yeah, yeah, it’s just been so many handovers.
343 00:39:55.420 ⇒ 00:40:00.229 Jasmin Multani: Thank you for calling out. Yeah, something tells me…
344 00:40:01.630 ⇒ 00:40:09.109 Jasmin Multani: Maybe it’s double counting? If it’s doing, then we start to end.
345 00:40:10.610 ⇒ 00:40:24.739 Jasmin Multani: I would wanna… my method is always, like, running things CTE by CTE, but I think there’s something happening at that labeling section, because the initial rank and row number makes sense to me.
346 00:40:24.840 ⇒ 00:40:28.750 Jasmin Multani: But I feel like the joining is where it’s… Creating a double count.
347 00:40:31.620 ⇒ 00:40:33.435 Jasmin Multani: like… Mmm…
348 00:40:42.390 ⇒ 00:40:45.020 Jasmin Multani: Calculate order milestones per…
349 00:40:50.230 ⇒ 00:40:51.699 Advait Nandakumar Menon: Are you talking about this joint?
350 00:40:53.940 ⇒ 00:41:01.120 Jasmin Multani: It might be that join, and there’s, like, a series of joins at the very end, too, the union alls, like…
351 00:41:01.390 ⇒ 00:41:03.119 Jasmin Multani: Very, very, very end.
352 00:41:04.890 ⇒ 00:41:05.990 Jasmin Multani: Yeah.
353 00:41:07.450 ⇒ 00:41:12.389 Advait Nandakumar Menon: Yeah, this… this one is to get all the metrics in, because since this is a summary table.
354 00:41:12.390 ⇒ 00:41:13.070 Jasmin Multani: Hmm.
355 00:41:13.070 ⇒ 00:41:26.909 Advait Nandakumar Menon: we are computing each of the individual metrics up here in the different CTs, and then basically unioning it all together in this final CT, and from there, we are doing a select statement, and just.
356 00:41:27.140 ⇒ 00:41:27.690 Shivani Amar: Sure.
357 00:41:28.060 ⇒ 00:41:30.889 Shivani Amar: I mean, that’s how you’re getting it into the dashboard, I get it, but, like.
358 00:41:31.160 ⇒ 00:41:37.800 Shivani Amar: like, how are they defining what a first order or second order is? Can you find that, if you go up?
359 00:41:39.810 ⇒ 00:41:43.089 Jasmin Multani: Yeah, it’s in the two… line two… yeah, I…
360 00:41:43.090 ⇒ 00:41:43.890 Shivani Amar: 55.
361 00:41:43.890 ⇒ 00:41:44.480 Jasmin Multani: 240…
362 00:41:44.480 ⇒ 00:41:51.680 Shivani Amar: Order milestone as customer ID segment max case when order sequence. So how do they define order sequence?
363 00:41:51.780 ⇒ 00:41:52.880 Shivani Amar: Let’s go there.
364 00:42:00.020 ⇒ 00:42:03.050 Jasmin Multani: Row number over partition, customer order by.
365 00:42:03.200 ⇒ 00:42:08.700 Jasmin Multani: I wonder if this is a… if we need to hardcode this as ascending.
366 00:42:09.970 ⇒ 00:42:12.379 Jasmin Multani: Because I don’t remember what the default would be.
367 00:42:17.120 ⇒ 00:42:17.760 Shivani Amar: Like…
368 00:42:18.640 ⇒ 00:42:24.309 Jasmin Multani: Because this could be either descending or ascending, so that’s something I would want to follow up on.
369 00:42:24.670 ⇒ 00:42:25.720 Jasmin Multani: But I feel like this…
370 00:42:25.720 ⇒ 00:42:29.369 Advait Nandakumar Menon: Since we don’t specify it, it defaults to as something.
371 00:42:29.370 ⇒ 00:42:31.810 Jasmin Multani: Yeah, I feel like this is…
372 00:42:32.930 ⇒ 00:42:35.580 Jasmin Multani: I feel like that row number makes sense.
373 00:42:38.450 ⇒ 00:42:46.480 Jasmin Multani: I wonder if what would happen if, like, there were multiple orders on the same day. Does it create, like, a…
374 00:42:47.280 ⇒ 00:42:48.500 Jasmin Multani: double ring.
375 00:42:48.890 ⇒ 00:42:53.570 Jasmin Multani: So instead of, like, having explicit 123, it says 122.
376 00:42:53.920 ⇒ 00:42:59.539 Jasmin Multani: that’s the only time where I… can see something being equated.
377 00:43:00.800 ⇒ 00:43:06.890 Advait Nandakumar Menon: I think… Then we should avoid using row number, and then use something like dense rank, or something.
378 00:43:06.890 ⇒ 00:43:07.810 Jasmin Multani: Something, but…
379 00:43:08.400 ⇒ 00:43:14.950 Jasmin Multani: But we need to double check, so my gut tells me is that we need to, get the raw output of orders with sequence.
380 00:43:15.280 ⇒ 00:43:19.379 Jasmin Multani: In the past 2 months, just so we have something easy to check through.
381 00:43:19.690 ⇒ 00:43:23.449 Jasmin Multani: Or just, like, actually…
382 00:43:24.290 ⇒ 00:43:31.989 Jasmin Multani: You do that, and, like, or, like, we work with a high… highly high retention, customer, and troubleshoot there, but we’re gonna have to go, like.
383 00:43:32.410 ⇒ 00:43:34.760 Jasmin Multani: CTE by CTE, and…
384 00:43:34.760 ⇒ 00:43:36.509 Advait Nandakumar Menon: We have to backtrack it all the way, yeah.
385 00:43:36.510 ⇒ 00:43:37.730 Jasmin Multani: Yeah, yeah, yeah.
386 00:43:38.330 ⇒ 00:43:45.810 Jasmin Multani: My gut tells me, like, that row number is fine. It’s the join, or this… 255 lined up.
387 00:43:45.940 ⇒ 00:43:47.760 Jasmin Multani: Might be kind of weird.
388 00:43:48.700 ⇒ 00:43:50.760 Jasmin Multani: And possibly inflating.
389 00:43:50.760 ⇒ 00:43:59.459 Shivani Amar: sequence approach is the wrong tool entirely. I’ll give you a misleading result, it’ll give you misleading results, because the customer’s second order in the window might actually be their 15th order.
390 00:43:59.650 ⇒ 00:44:00.940 Shivani Amar: Lifetime.
391 00:44:01.470 ⇒ 00:44:09.330 Shivani Amar: What you actually want to do is rank orders within the window only. The fix is straightforward. You should say…
392 00:44:09.470 ⇒ 00:44:12.539 Shivani Amar: I mean, Claude can just, like, help you figure this out, guys.
393 00:44:12.820 ⇒ 00:44:15.880 Shivani Amar: Like, was it their first order in the window?
394 00:44:16.150 ⇒ 00:44:23.990 Shivani Amar: not ev- like… not ever, right? Like, do you get what I’m… sorry. Oho.
395 00:44:25.580 ⇒ 00:44:27.629 Jasmin Multani: I kind of disagree, though.
396 00:44:27.630 ⇒ 00:44:31.540 Shivani Amar: No, that was wrong, hold on. Behavior within time when… hold on.
397 00:44:36.110 ⇒ 00:44:37.330 Shivani Amar: One second…
398 00:44:46.370 ⇒ 00:44:47.130 Shivani Amar: Okay.
399 00:44:47.490 ⇒ 00:44:48.370 Shivani Amar: One second.
400 00:44:48.600 ⇒ 00:44:49.170 Shivani Amar: That was…
401 00:44:49.170 ⇒ 00:44:50.720 Jasmin Multani: Yeah, yeah, no, no worries, no worries.
402 00:44:51.950 ⇒ 00:44:59.940 Shivani Amar: Was this the order… order… was this order the customer’s first ever, and did it happen to fall in this time period? That’s what I want to know. Okay.
403 00:45:00.450 ⇒ 00:45:04.399 Shivani Amar: You just need to know the customer’s first order date, then check if it falls in the window.
404 00:45:04.760 ⇒ 00:45:11.210 Shivani Amar: Great. So it’s like, that’s what I’m, like, kind of confused. I’m like, this should be, like, going by…
405 00:45:30.650 ⇒ 00:45:35.119 Jasmin Multani: Can we try running order milestone CTE in Snowflake?
406 00:45:37.170 ⇒ 00:45:38.720 Jasmin Multani: I just wanna see every…
407 00:45:38.900 ⇒ 00:45:45.520 Jasmin Multani: like, no, no, no, like, the… I want to compare the orders with sequence output with the order milestone.
408 00:45:47.300 ⇒ 00:45:47.850 Advait Nandakumar Menon: Okay.
409 00:45:50.340 ⇒ 00:45:57.250 Jasmin Multani: And then, like, truncate it so that we’re only looking at, like, the last 3 months of data.
410 00:46:00.150 ⇒ 00:46:01.900 Jasmin Multani: Horrible time, just…
411 00:46:29.910 ⇒ 00:46:32.370 Advait Nandakumar Menon: Did you say last 3 months?
412 00:46:33.330 ⇒ 00:46:38.330 Jasmin Multani: Yeah, for the last 3 months. Or just do it of all time, actually, I just want to see the raw.
413 00:46:44.410 ⇒ 00:46:46.370 Jasmin Multani: And… no.
414 00:46:49.100 ⇒ 00:46:50.860 Jasmin Multani: It might take a few minutes.
415 00:46:58.080 ⇒ 00:47:05.639 Jasmin Multani: And then once it gives us an answer, let’s hand over the, the subsequent CTE.
416 00:47:05.680 ⇒ 00:47:18.160 Jasmin Multani: And be like, hey, can you tell us… can you tell us if this is an accurate labeling? And give it an example, and say, hey, if this customer made their first order last week.
417 00:47:18.680 ⇒ 00:47:30.239 Jasmin Multani: then I want the labeling to come out as first. Yeah. But if there… if there was an order that was made last week, but it’s that customer’s 10th order, we do not want that to be surfaced.
418 00:47:32.170 ⇒ 00:47:35.909 Advait Nandakumar Menon: You just want the first, second, or third, or within that week, or within that time period.
419 00:47:35.910 ⇒ 00:47:42.549 Shivani Amar: Sorry, sorry, sorry. Like, you’re like… Last week, How many customers?
420 00:47:43.670 ⇒ 00:47:45.780 Shivani Amar: Ever place their first order.
421 00:47:46.060 ⇒ 00:47:46.930 Jasmin Multani: Yes.
422 00:47:46.930 ⇒ 00:47:48.550 Advait Nandakumar Menon: Second order of all time.
423 00:47:49.220 ⇒ 00:47:50.920 Shivani Amar: Second order of all time.
424 00:47:51.060 ⇒ 00:47:52.739 Shivani Amar: Third order of all time.
425 00:47:53.880 ⇒ 00:47:54.510 Advait Nandakumar Menon: Okay.
426 00:47:54.750 ⇒ 00:47:55.510 Shivani Amar: Okay?
427 00:47:55.690 ⇒ 00:47:58.730 Shivani Amar: Not in the last 12 months, this was their third order.
428 00:47:58.840 ⇒ 00:48:00.120 Shivani Amar: Like, no.
429 00:48:00.120 ⇒ 00:48:00.610 Advait Nandakumar Menon: Perfect.
430 00:48:00.840 ⇒ 00:48:10.009 Shivani Amar: Of all our customers, how many placed their first order last week? How many placed their second order last week? How many placed their third order last week? I’m not cohorting it.
431 00:48:10.710 ⇒ 00:48:17.029 Shivani Amar: I just want to know what happened. How many people placed their first order? How many people placed their second order? How many people placed their third order?
432 00:48:21.530 ⇒ 00:48:25.320 Shivani Amar: And then the way that my brain works is that you see that weekly.
433 00:48:25.670 ⇒ 00:48:39.909 Shivani Amar: And then you can see, like, oh, I had a weird week where, like, not a lot of second orders happened, right? Like, you can, like, kind of, like, react to that. But then you can also see your quarterly averages over time, and you can, like, be like, hmm, okay, like…
434 00:48:40.060 ⇒ 00:48:49.170 Shivani Amar: the shape of my funnel seems to be changing. I’m… my… my… I have fewer people placing second orders, or have more people placing second orders than I used to.
435 00:48:51.420 ⇒ 00:48:55.595 Jasmin Multani: Yeah. About…
436 00:48:57.200 ⇒ 00:49:02.819 Shivani Amar: And, like, who knows? Maybe your numbers are completely right, but it just feels so weird that there’s no drop-off to me.
437 00:49:05.580 ⇒ 00:49:13.429 Jasmin Multani: So, 5, 6… so, why do we see repeating week start date, even though the order sequence is different?
438 00:49:14.280 ⇒ 00:49:21.009 Jasmin Multani: That’s really interesting. Are we saying that, like, Trusted Health made 10 orders.
439 00:49:21.520 ⇒ 00:49:25.000 Shivani Amar: I mean, that’s… Trusted Health is a segment, not a customer.
440 00:49:25.390 ⇒ 00:49:30.680 Shivani Amar: So your custom ID is ending in 418, right?
441 00:49:30.680 ⇒ 00:49:33.139 Jasmin Multani: It’s still highly repetitive, so that’s.
442 00:49:33.140 ⇒ 00:49:40.079 Shivani Amar: So then I’m like, okay, like, did… did that customer just place… who is customer… let’s go to who is that customer?
443 00:49:42.060 ⇒ 00:49:44.379 Jasmin Multani: Maybe we can ask Salty.
444 00:49:47.110 ⇒ 00:49:48.220 Advait Nandakumar Menon: Oh, you wanna tech…
445 00:49:48.220 ⇒ 00:49:51.620 Shivani Amar: Or no, go to Shop… Shopify is great, let’s go to the source system, that’s fine.
446 00:49:56.090 ⇒ 00:50:01.990 Jasmin Multani: This, I think the fix could also be, like, creating columns that’s literally, like.
447 00:50:02.110 ⇒ 00:50:07.579 Jasmin Multani: Case when order sequence is 1, then give me the order date.
448 00:50:07.820 ⇒ 00:50:11.079 Jasmin Multani: And then the next column, case, when blah blah blah blah blah.
449 00:50:11.260 ⇒ 00:50:20.249 Jasmin Multani: Does that make sense? We just overwrite the actual date, and then we create filters for those dates. Or, like, the summation based off of those dates.
450 00:50:37.210 ⇒ 00:50:40.709 Jasmin Multani: So, we’re looking for… can we filter by customer?
451 00:50:42.040 ⇒ 00:50:45.730 Advait Nandakumar Menon: I unfortunately don’t have access to customers, just the order.
452 00:50:49.310 ⇒ 00:50:54.730 Jasmin Multani: Shivani, can you… can we borrow… can we use your eyes on this? Or…
453 00:50:54.730 ⇒ 00:51:00.139 Shivani Amar: Yeah, I don’t even know what I have access to, so… Login…
454 00:51:00.300 ⇒ 00:51:02.839 Shivani Amar: I don’t know if I have access to anything.
455 00:51:03.270 ⇒ 00:51:05.350 Shivani Amar: Let’s see.
456 00:51:13.450 ⇒ 00:51:15.260 Shivani Amar: Looks like I’m logging in now…
457 00:51:22.230 ⇒ 00:51:29.299 Jasmin Multani: At least the good thing is that, like, the row number is able to distinguish between the ties, so there’s a distinct.
458 00:51:30.380 ⇒ 00:51:33.130 Advait Nandakumar Menon: Yeah, it’s not returning the same number.
459 00:51:35.110 ⇒ 00:51:39.740 Shivani Amar: Okay, so… Let’s see, customers, right?
460 00:51:40.440 ⇒ 00:51:41.220 Jasmin Multani: Yeah.
461 00:51:41.220 ⇒ 00:51:41.540 Shivani Amar: Okay.
462 00:51:41.540 ⇒ 00:51:43.080 Jasmin Multani: I’ll send you the custom ID.
463 00:51:45.550 ⇒ 00:51:46.160 Shivani Amar: Let’s see.
464 00:51:47.370 ⇒ 00:51:52.090 Shivani Amar: You wanna just put that… you put it in the chat here, so then do I put that in…
465 00:51:52.640 ⇒ 00:51:58.370 Shivani Amar: Okay, so it says Jan Keith, okay? 34 orders. Let’s see.
466 00:51:58.370 ⇒ 00:51:59.390 Jasmin Multani: In one day.
467 00:51:59.390 ⇒ 00:52:08.580 Shivani Amar: Let’s see… no, let me, let me just check, because maybe that’s what’s happening, right? So, no.
468 00:52:08.790 ⇒ 00:52:10.899 Shivani Amar: What date were you looking at, Avid?
469 00:52:12.650 ⇒ 00:52:13.630 Advait Nandakumar Menon: Oh, you said?
470 00:52:13.630 ⇒ 00:52:14.730 Shivani Amar: like, repeating.
471 00:52:15.860 ⇒ 00:52:17.760 Advait Nandakumar Menon: Yeah, let me run that again, I…
472 00:52:18.290 ⇒ 00:52:20.690 Jasmin Multani: It was, like, 20 orders in one day.
473 00:52:21.950 ⇒ 00:52:22.390 Shivani Amar: 7…
474 00:52:22.390 ⇒ 00:52:23.390 Advait Nandakumar Menon: 29, 29.
475 00:52:23.390 ⇒ 00:52:23.900 Shivani Amar: 19.
476 00:52:23.900 ⇒ 00:52:24.610 Advait Nandakumar Menon: Yeah.
477 00:52:24.930 ⇒ 00:52:25.840 Advait Nandakumar Menon: Yeah.
478 00:52:31.160 ⇒ 00:52:39.309 Jasmin Multani: And this is going… okay, so now the order… after, like, 30, row 30, I’m seeing difference in order created at.
479 00:52:40.270 ⇒ 00:52:42.810 Jasmin Multani: So maybe he did just create 3 visits.
480 00:52:42.810 ⇒ 00:52:45.690 Shivani Amar: Like, it’s weird, like, look, let me show you my screen for a second.
481 00:52:46.650 ⇒ 00:52:47.270 Advait Nandakumar Menon: Sure.
482 00:52:47.610 ⇒ 00:52:55.130 Shivani Amar: This is… so I just typed that number in, and it shows me Jan Keith, and then it shows me, theoretically, first order was…
483 00:52:55.940 ⇒ 00:52:59.949 Shivani Amar: I don’t even see anything for July. Wait, hold on one second.
484 00:53:00.880 ⇒ 00:53:01.900 Shivani Amar: Hey, girl!
485 00:53:04.300 ⇒ 00:53:10.379 Jasmin Multani: Abbott, maybe we can… Look up a different num- a different date?
486 00:53:10.500 ⇒ 00:53:16.140 Jasmin Multani: I wonder if that’s, he created that order, but then canceled it.
487 00:53:18.450 ⇒ 00:53:19.180 Advait Nandakumar Menon: Okay.
488 00:53:25.900 ⇒ 00:53:30.199 Jasmin Multani: These all look like they’re either fulfilled, not fulfilled, voided.
489 00:53:30.820 ⇒ 00:53:34.279 Advait Nandakumar Menon: There is one on 8th May 2023.
490 00:53:35.560 ⇒ 00:53:36.420 Jasmin Multani: Yeah.
491 00:53:37.180 ⇒ 00:53:41.049 Jasmin Multani: I see one order for May 8th, 2023.
492 00:53:41.870 ⇒ 00:53:47.020 Shivani Amar: Okay, sorry, so… This number was, in fact, Jan Keith.
493 00:53:47.870 ⇒ 00:53:53.849 Shivani Amar: Right? Then, if that’s what you see on your end, then I’m like, I’m not even seeing that date, which is common.
494 00:53:56.100 ⇒ 00:53:58.510 Jasmin Multani: That would be… Sorry, go on.
495 00:53:58.510 ⇒ 00:54:05.730 Advait Nandakumar Menon: So, yeah, the next is May 8th, then 7th June, then 18… October…
496 00:54:06.260 ⇒ 00:54:10.269 Advait Nandakumar Menon: All that is lining up. I’m not sure what’s happening before May 8th.
497 00:54:11.530 ⇒ 00:54:12.600 Shivani Amar: Hmm…
498 00:54:13.230 ⇒ 00:54:17.389 Jasmin Multani: Do we think that there was, like, some sort of data error? Yeah, and…
499 00:54:17.770 ⇒ 00:54:22.779 Jasmin Multani: Shivani, do you remember anything crazy that happened between… in 2019?
500 00:54:22.780 ⇒ 00:54:24.030 Shivani Amar: I wasn’t… I didn’t work here.
501 00:54:24.030 ⇒ 00:54:24.980 Jasmin Multani: Yeah, yeah, yeah.
502 00:54:26.000 ⇒ 00:54:26.660 Jasmin Multani: So…
503 00:54:26.660 ⇒ 00:54:27.480 Shivani Amar: I don’t know.
504 00:54:28.420 ⇒ 00:54:34.519 Jasmin Multani: When do we think that… Do we… do we have an opinion about, like, when data fidelity.
505 00:54:34.710 ⇒ 00:54:40.179 Shivani Amar: But, like, the thing is, it’s like, I’m not even seeing it in Shopify, so it’s like, why is it showing up like that in Snowflake?
506 00:54:40.350 ⇒ 00:54:44.979 Shivani Amar: This is just a question that we can do a little deep dive with Amanda Wesh.
507 00:54:45.940 ⇒ 00:54:46.730 Shivani Amar: Right?
508 00:54:47.020 ⇒ 00:54:48.610 Shivani Amar: Like, what’s going on?
509 00:54:48.610 ⇒ 00:54:49.860 Jasmin Multani: Just get a screenshot.
510 00:54:50.320 ⇒ 00:54:53.669 Shivani Amar: Yeah, take the screenshot, we can look at this again together with them.
511 00:54:55.990 ⇒ 00:55:00.000 Shivani Amar: Okay, give me another one.
512 00:55:00.110 ⇒ 00:55:04.779 Shivani Amar: This is fun. And then I have to go. But give me another one that’s weird.
513 00:55:05.750 ⇒ 00:55:06.370 Jasmin Multani: Okay.
514 00:55:06.370 ⇒ 00:55:08.170 Shivani Amar: Or, like, give me a vid…
515 00:55:08.590 ⇒ 00:55:12.230 Shivani Amar: Show me all the people that you think had their third order last week.
516 00:55:14.840 ⇒ 00:55:17.369 Jasmin Multani: That’s a good practice.
517 00:55:17.370 ⇒ 00:55:17.760 Shivani Amar: Yeah.
518 00:55:18.070 ⇒ 00:55:20.720 Jasmin Multani: We can… you don’t have to… we…
519 00:55:21.910 ⇒ 00:55:28.450 Jasmin Multani: yeah, just write the sequel the way you would feel like writing, or you can ask AI, but I would say let’s,
520 00:55:29.480 ⇒ 00:55:33.670 Jasmin Multani: add a CTE on top of the raw data that we just
521 00:55:33.920 ⇒ 00:55:38.080 Jasmin Multani: Got out. And, just build it the way you would think.
522 00:55:39.040 ⇒ 00:55:43.660 Jasmin Multani: There are a few different ways to build it. Sorry, this is, like, an interview question now.
523 00:55:44.750 ⇒ 00:55:50.170 Jasmin Multani: But, yeah, I can use AI, but, like, we also have to make sure we’re being apples to apples.
524 00:55:50.900 ⇒ 00:55:51.510 Advait Nandakumar Menon: Sure.
525 00:56:00.700 ⇒ 00:56:02.800 Jasmin Multani: How is the donor show up all the money?
526 00:56:02.800 ⇒ 00:56:05.600 Shivani Amar: It was, like, the best weekend of my life, it was so fun.
527 00:56:06.250 ⇒ 00:56:10.750 Shivani Amar: But I’m, like, I had a cold before going, and then today I’m still pretty congested, so I’m like…
528 00:56:11.000 ⇒ 00:56:13.149 Shivani Amar: Okay, we gotta get this…
529 00:56:13.310 ⇒ 00:56:13.949 Jasmin Multani: body and mind.
530 00:56:13.950 ⇒ 00:56:15.859 Shivani Amar: order before the wedding.
531 00:56:20.000 ⇒ 00:56:20.670 Shivani Amar: Shit.
532 00:56:37.760 ⇒ 00:56:47.070 Shivani Amar: Yeah, okay, that’ll be fun. We’ll just go through a few example customer IDs that you drop in of people who did their third orders, and be like, that was actually their 12th order, or…
533 00:56:47.640 ⇒ 00:56:50.750 Shivani Amar: maybe they’re all third orders. I’m so excited.
534 00:56:51.090 ⇒ 00:56:52.290 Jasmin Multani: That’s a good sprint.
535 00:56:53.450 ⇒ 00:56:59.560 Jasmin Multani: with it. I don’t miss being an analyst doing queries live.
536 00:57:28.240 ⇒ 00:57:33.200 Advait Nandakumar Menon: So, this one is on 23rd April, this custom ID.
537 00:57:50.590 ⇒ 00:57:56.770 Jasmin Multani: So, we’re saying that this customer ID has only placed 3 orders so far, and the.
538 00:57:56.770 ⇒ 00:57:57.100 Shivani Amar: Okay.
539 00:57:57.100 ⇒ 00:58:00.610 Jasmin Multani: Water is, happened last week.
540 00:58:00.940 ⇒ 00:58:02.340 Jasmin Multani: Floods verify.
541 00:58:03.160 ⇒ 00:58:04.750 Shivani Amar: Okay, 3 orders.
542 00:58:06.150 ⇒ 00:58:07.000 Shivani Amar: Woop boop!
543 00:58:10.000 ⇒ 00:58:11.340 Shivani Amar: Three orders!
544 00:58:12.290 ⇒ 00:58:13.070 Shivani Amar: Guys.
545 00:58:13.410 ⇒ 00:58:18.079 Shivani Amar: Maybe your funnel is perfectly right. 3 orders! So far, 3 for 3.
546 00:58:18.830 ⇒ 00:58:22.539 Jasmin Multani: Cool. Do you think there… was there a difference in the way you wrote it?
547 00:58:22.540 ⇒ 00:58:29.939 Shivani Amar: Oh, it’s interesting, the Chris Spencer, okay, so the third one that you sent me, looks like she placed… oh, check this out.
548 00:58:30.760 ⇒ 00:58:33.060 Shivani Amar: Well, if we’re learning.
549 00:58:33.170 ⇒ 00:58:34.110 Shivani Amar: Cheers.
550 00:58:34.890 ⇒ 00:58:37.360 Shivani Amar: 8.32, and then 0.
551 00:58:40.040 ⇒ 00:58:42.350 Jasmin Multani: Yeah, what does that mean? Undelivered.
552 00:58:42.510 ⇒ 00:58:43.560 Jasmin Multani: Is that it?
553 00:58:43.740 ⇒ 00:58:44.990 Jasmin Multani: a sampling?
554 00:58:47.270 ⇒ 00:58:47.980 Shivani Amar: Boom.
555 00:58:49.900 ⇒ 00:58:52.380 Shivani Amar: Because sometimes we give them displays.
556 00:58:52.500 ⇒ 00:58:56.150 Shivani Amar: But that’s $0, so let’s look at the other one that you sent.
557 00:58:59.270 ⇒ 00:59:01.239 Advait Nandakumar Menon: Wait, is this, again, Jhank?
558 00:59:01.660 ⇒ 00:59:04.589 Advait Nandakumar Menon: I thought the first one we… no, the one before this.
559 00:59:04.590 ⇒ 00:59:05.559 Shivani Amar: Hold on.
560 00:59:07.750 ⇒ 00:59:09.160 Shivani Amar: The one before this?
561 00:59:10.630 ⇒ 00:59:12.560 Advait Nandakumar Menon: Okay, maybe I got confused, like…
562 00:59:12.700 ⇒ 00:59:16.290 Advait Nandakumar Menon: Near a search bar, it’s still Jan Kids, so I thought it’s…
563 00:59:16.290 ⇒ 00:59:20.830 Shivani Amar: No, that’s fine. This one’s interesting, right? Because this one, there’s a $0 order.
564 00:59:21.590 ⇒ 00:59:23.660 Shivani Amar: And this one was a failed delivery.
565 00:59:24.890 ⇒ 00:59:28.720 Jasmin Multani: And I wonder if we still… charged them.
566 00:59:29.090 ⇒ 00:59:37.629 Shivani Amar: I don’t know. But, like, I think that the… it’s interesting, because I would say a $0 order, the same date, we’re just learning the same date as…
567 00:59:37.880 ⇒ 00:59:42.570 Shivani Amar: Another order might not be, like, considered part of the funnel, which is why you’re seeing
568 00:59:43.160 ⇒ 00:59:46.299 Shivani Amar: The third order actually be more than the second order.
569 00:59:46.300 ⇒ 00:59:48.009 Jasmin Multani: That’s the inflation. Okay.
570 00:59:48.010 ⇒ 00:59:53.899 Shivani Amar: Okay, okay, so can you give me another couple, can you give me another couple IDs?
571 00:59:54.630 ⇒ 00:59:55.979 Advait Nandakumar Menon: I have a bunch of them here.
572 01:00:07.850 ⇒ 01:00:18.030 Shivani Amar: okay, this one looks like it was, in fact, their third order, right? So I think… I think we just solved, maybe, why the number is higher, and it’s probably because of the…
573 01:00:18.200 ⇒ 01:00:19.400 Shivani Amar: the,
574 01:00:19.670 ⇒ 01:00:29.020 Shivani Amar: I don’t know how to validate this, right? Like, a lot of the ones you’re giving me, I’m like, yep, this is exactly what I would want to see. That this was their third order, which is great.
575 01:00:30.210 ⇒ 01:00:34.850 Jasmin Multani: Yeah. I have a quick question, how different…
576 01:00:35.660 ⇒ 01:00:41.689 Jasmin Multani: I want to see how you, created that SQL for that CTE, and how different the.
577 01:00:41.690 ⇒ 01:00:42.050 Advait Nandakumar Menon: No problem.
578 01:00:42.050 ⇒ 01:00:45.100 Jasmin Multani: are compared to what Oasis Original is.
579 01:00:45.100 ⇒ 01:00:49.020 Shivani Amar: And maybe it’s just a matter of filtering out $0 orders.
580 01:00:49.700 ⇒ 01:00:50.410 Jasmin Multani: Yeah.
581 01:00:50.900 ⇒ 01:00:51.500 Shivani Amar: Right?
582 01:00:51.500 ⇒ 01:00:52.220 Jasmin Multani: Yep.
583 01:00:52.220 ⇒ 01:00:54.100 Shivani Amar: Can you give me a few more?
584 01:00:55.220 ⇒ 01:01:01.860 Shivani Amar: While we’re on this… Because otherwise, it’s good. I’m like, it’s working, right? I just,
585 01:01:02.150 ⇒ 01:01:04.740 Shivani Amar: Like, seeing April 23rd.
586 01:01:06.350 ⇒ 01:01:09.449 Shivani Amar: An order was placed, and it was their third order. That’s great.
587 01:01:11.470 ⇒ 01:01:14.030 Jasmin Multani: Can we get access to the full Shopify view?
588 01:01:14.030 ⇒ 01:01:15.499 Shivani Amar: Yeah, yeah, for sure.
589 01:01:15.500 ⇒ 01:01:18.390 Jasmin Multani: It just, like, makes validation…
590 01:01:18.390 ⇒ 01:01:19.419 Shivani Amar: So much easier.
591 01:01:19.420 ⇒ 01:01:20.240 Jasmin Multani: Yeah, yeah.
592 01:01:21.500 ⇒ 01:01:24.680 Shivani Amar: This was also good. These are small orders, Dustin.
593 01:01:27.920 ⇒ 01:01:29.710 Advait Nandakumar Menon: Yeah, this is it.
594 01:01:33.490 ⇒ 01:01:38.519 Jasmin Multani: And even if they make 3 orders in the same day, Shivani, you still want that.
595 01:01:38.660 ⇒ 01:01:39.989 Jasmin Multani: Counted as 3.
596 01:01:41.500 ⇒ 01:01:43.049 Shivani Amar: Actually, I don’t know.
597 01:01:43.200 ⇒ 01:01:44.390 Shivani Amar: What do you think?
598 01:01:44.890 ⇒ 01:01:46.130 Shivani Amar: should be…
599 01:01:52.160 ⇒ 01:01:52.700 Jasmin Multani: I think…
600 01:01:52.700 ⇒ 01:01:57.389 Advait Nandakumar Menon: If a net new customer Place 3 orders on the same day.
601 01:01:58.260 ⇒ 01:02:03.560 Jasmin Multani: Yeah, because it’s like, why did they order 3 things on the same day? Like…
602 01:02:03.560 ⇒ 01:02:10.800 Shivani Amar: Yes, they’re ordering to 3 different locations, because they’re at a gym, and they’re like, okay, I’m gonna order to gym 1, 2, and 3.
603 01:02:10.800 ⇒ 01:02:11.670 Jasmin Multani: Yeah.
604 01:02:12.570 ⇒ 01:02:16.579 Jasmin Multani: I would say we should still keep that ranking, so it’s distinct.
605 01:02:16.580 ⇒ 01:02:17.070 Shivani Amar: Yeah.
606 01:02:17.070 ⇒ 01:02:25.300 Jasmin Multani: We have to have, like, a separate column that accounts for Rank the order sequence.
607 01:02:25.550 ⇒ 01:02:27.460 Jasmin Multani: And be distinct by day.
608 01:02:28.350 ⇒ 01:02:32.250 Jasmin Multani: So… because that’ll… I think that’s a truer form of retention.
609 01:02:32.390 ⇒ 01:02:35.250 Jasmin Multani: Given how long it takes for a delivery to happen.
610 01:02:35.930 ⇒ 01:02:49.859 Jasmin Multani: But, that’s where we start getting into cohorting. Cohorting is very messy. I would love if your analytics engineer wanted to own that part. But that’s gonna be more brainstorming.
611 01:02:50.140 ⇒ 01:02:50.920 Shivani Amar: Gotcha.
612 01:02:51.870 ⇒ 01:02:57.930 Shivani Amar: Okay, I just asked in the chat if we can get you access to what I’m seeing.
613 01:02:58.850 ⇒ 01:03:00.730 Shivani Amar: Yvette, wanna send me any more?
614 01:03:02.220 ⇒ 01:03:05.230 Advait Nandakumar Menon: That is it during last week, I mean, the week before.
615 01:03:05.590 ⇒ 01:03:06.510 Shivani Amar: That’s it.
616 01:03:07.160 ⇒ 01:03:10.490 Advait Nandakumar Menon: Yeah, these are the customers who placed there.
617 01:03:11.260 ⇒ 01:03:12.399 Advait Nandakumar Menon: Third order.
618 01:03:12.640 ⇒ 01:03:13.709 Advait Nandakumar Menon: Then this is…
619 01:03:13.710 ⇒ 01:03:15.520 Shivani Amar: This is 35 third orders.
620 01:03:19.770 ⇒ 01:03:21.690 Jasmin Multani: Are the DC 35 distinct customers.
621 01:03:21.690 ⇒ 01:03:27.839 Shivani Amar: Were you on a specific date of it, or were you on the week, the week?
622 01:03:29.610 ⇒ 01:03:30.740 Advait Nandakumar Menon: that week.
623 01:03:31.640 ⇒ 01:03:32.869 Advait Nandakumar Menon: Where’s that go in?
624 01:03:44.480 ⇒ 01:03:52.050 Jasmin Multani: So good, that could infer that, like, okay, yes, what, the original sequel was is very… is a little bit different than what
625 01:03:52.170 ⇒ 01:03:54.240 Jasmin Multani: This output is.
626 01:03:57.670 ⇒ 01:03:59.660 Shivani Amar: Because Advance output seems good.
627 01:04:02.030 ⇒ 01:04:04.970 Jasmin Multani: But we also want to make sure that, like, if it’s truly 35.
628 01:04:04.970 ⇒ 01:04:06.930 Shivani Amar: What is this one that you’re sending me, Ovid?
629 01:04:07.480 ⇒ 01:04:08.949 Advait Nandakumar Menon: Yeah, I just…
630 01:04:09.350 ⇒ 01:04:14.029 Advait Nandakumar Menon: looked for more, like, I removed the… I had a limit of just 10 rows, so I removed the.
631 01:04:14.030 ⇒ 01:04:15.249 Shivani Amar: That was white, okay.
632 01:04:15.530 ⇒ 01:04:16.480 Advait Nandakumar Menon: Yeah, yeah.
633 01:04:16.760 ⇒ 01:04:17.610 Jasmin Multani: Hmm.
634 01:04:17.800 ⇒ 01:04:20.340 Jasmin Multani: So do you see all 35 customers?
635 01:04:25.320 ⇒ 01:04:29.020 Shivani Amar: No, this one looks like it’s their… Fourth order.
636 01:04:29.020 ⇒ 01:04:29.570 Advait Nandakumar Menon: Yeah.
637 01:04:30.820 ⇒ 01:04:31.270 Jasmin Multani: Mmm…
638 01:04:31.270 ⇒ 01:04:35.819 Shivani Amar: Oh, because it says Vicki Sanchez, but the customer ID is the same.
639 01:04:39.640 ⇒ 01:04:42.589 Shivani Amar: So is it going based off the name of the person?
640 01:04:42.760 ⇒ 01:04:44.330 Shivani Amar: Or the customer ID.
641 01:04:45.410 ⇒ 01:04:56.709 Jasmin Multani: I think it depends on how the data tables are set up, because if it’s… if a column says name, and then customer ID, then it’s gonna be distinct by name and customer ID.
642 01:04:56.710 ⇒ 01:05:03.760 Shivani Amar: Okay, so this is an example, let’s just screenshot this one, where it’s saying third order, but it’s actually the wholesale partner’s
643 01:05:03.930 ⇒ 01:05:04.870 Shivani Amar: 4th?
644 01:05:05.180 ⇒ 01:05:08.509 Shivani Amar: Okay, I’ll just send it to you guys in the chat.
645 01:05:09.940 ⇒ 01:05:10.620 Shivani Amar: And then…
646 01:05:10.620 ⇒ 01:05:11.090 Jasmin Multani: Give me some…
647 01:05:11.090 ⇒ 01:05:13.599 Shivani Amar: Or, Advait, let’s do this thing.
648 01:05:20.810 ⇒ 01:05:21.570 Shivani Amar: Fantastic.
649 01:05:47.320 ⇒ 01:05:49.780 Shivani Amar: That was actually the third, that’s great.
650 01:05:55.960 ⇒ 01:05:56.820 Advait Nandakumar Menon: So, right now.
651 01:05:56.820 ⇒ 01:06:00.989 Shivani Amar: I’m off, right? It’s like, are we going off the name? Is it a zero?
652 01:06:01.420 ⇒ 01:06:05.659 Shivani Amar: Otherwise, the way that you’re pulling it… are you getting 35? I forget what you said.
653 01:06:06.300 ⇒ 01:06:09.949 Advait Nandakumar Menon: Yeah, I was saying I didn’t look it by the date anymore, it’s like…
654 01:06:10.500 ⇒ 01:06:15.889 Advait Nandakumar Menon: I’m looking at all the rows, so… Since we wanted more samples.
655 01:06:17.000 ⇒ 01:06:27.309 Shivani Amar: This one isn’t from last week, so… sorry. So, what I’m saying is, if you do your sequel to say, how many third orders happened last week, are you also getting 35?
656 01:06:30.010 ⇒ 01:06:32.710 Advait Nandakumar Menon: Let me see that.
657 01:06:33.250 ⇒ 01:06:33.870 Jasmin Multani: Yep.
658 01:06:40.630 ⇒ 01:06:41.750 Jasmin Multani: Like, whatever your answer is.
659 01:06:41.750 ⇒ 01:06:45.280 Shivani Amar: Or, like, 32, I guess, would be last week. Yeah.
660 01:06:45.280 ⇒ 01:06:51.290 Jasmin Multani: Even if your answer is different, that doesn’t mean we’re gonna, like, completely override OASIS. We…
661 01:06:51.290 ⇒ 01:06:51.629 Advait Nandakumar Menon: one of the.
662 01:06:51.630 ⇒ 01:06:54.250 Jasmin Multani: See, like, where else… the…
663 01:06:54.250 ⇒ 01:06:54.580 Advait Nandakumar Menon: Yeah.
664 01:06:55.440 ⇒ 01:06:58.230 Jasmin Multani: We want to just check in with Awash, like, was this intended or not?
665 01:06:58.230 ⇒ 01:07:00.750 Shivani Amar: Yeah, yeah, yeah, there’s… and this is not, like…
666 01:07:01.530 ⇒ 01:07:03.560 Shivani Amar: We’re all just figuring it out together.
667 01:07:06.320 ⇒ 01:07:07.010 Jasmin Multani: Okay.
668 01:07:12.150 ⇒ 01:07:16.170 Shivani Amar: But yeah, I’ll give you… Quiet time to do your sequel.
669 01:07:16.290 ⇒ 01:07:18.600 Shivani Amar: And I’m just curious how many you get.
670 01:08:12.760 ⇒ 01:08:13.750 Shivani Amar: Okay.
671 01:08:30.630 ⇒ 01:08:33.709 Shivani Amar: And this is, like, such a small example, like, I’m like…
672 01:08:34.160 ⇒ 01:08:38.169 Shivani Amar: this is not sales, but I’m like, it’s just a… it’s a good…
673 01:08:38.319 ⇒ 01:08:43.099 Shivani Amar: It’s… I’m so glad we looked at the funnel visual for us to, like, see that it was a little weird.
674 01:08:45.260 ⇒ 01:08:46.270 Jasmin Multani: Good job, Athi.
675 01:08:47.270 ⇒ 01:08:48.450 Shivani Amar: Oh, God.
676 01:09:05.060 ⇒ 01:09:07.910 Advait Nandakumar Menon: So I don’t know, it’s acting weird, it’s…
677 01:09:08.180 ⇒ 01:09:14.340 Advait Nandakumar Menon: Saying there’s nothing for last week, but from the week of April 19th to 25th?
678 01:09:14.560 ⇒ 01:09:19.370 Advait Nandakumar Menon: It says there is, there are 29 customers.
679 01:09:33.240 ⇒ 01:09:40.769 Jasmin Multani: Yeah, because I remember we… we hard-coded it as 20260418.
680 01:09:41.370 ⇒ 01:09:42.290 Jasmin Multani: Right?
681 01:09:43.210 ⇒ 01:09:44.140 Jasmin Multani: as the.
682 01:09:44.140 ⇒ 01:09:44.560 Advait Nandakumar Menon: Yeah, but.
683 01:09:44.560 ⇒ 01:09:45.060 Jasmin Multani: that.
684 01:09:45.069 ⇒ 01:09:50.569 Advait Nandakumar Menon: Because before I just handed over the query to AI, asking it to look at the third orders ever.
685 01:09:51.050 ⇒ 01:09:51.460 Jasmin Multani: God.
686 01:09:51.460 ⇒ 01:09:52.450 Advait Nandakumar Menon: Golden Bird.
687 01:10:04.310 ⇒ 01:10:06.600 Jasmin Multani: Can you share a screen, just out of curiosity?
688 01:10:07.460 ⇒ 01:10:08.110 Advait Nandakumar Menon: Sure.
689 01:10:28.200 ⇒ 01:10:33.310 Jasmin Multani: Okay, so… okay, where order… order above?
690 01:10:33.310 ⇒ 01:10:36.920 Advait Nandakumar Menon: And this one, over here. This is where I got the IDs from.
691 01:10:37.620 ⇒ 01:10:38.450 Jasmin Multani: Okay.
692 01:10:39.230 ⇒ 01:10:45.580 Jasmin Multani: So… For this, I don’t think,
693 01:10:47.610 ⇒ 01:10:58.849 Jasmin Multani: Yeah, I don’t think it’s actually filtering by order created at. I think it’s just making it a descending, and then we happen to be looking at the last week.
694 01:10:59.590 ⇒ 01:11:02.770 Jasmin Multani: So, I would ask AI…
695 01:11:05.010 ⇒ 01:11:06.940 Advait Nandakumar Menon: I can just filter it to…
696 01:11:10.040 ⇒ 01:11:16.389 Jasmin Multani: Let’s write it in a way that actually, like, Label the date.
697 01:11:16.780 ⇒ 01:11:19.660 Jasmin Multani: I, I, I would say, like, doing a case one.
698 01:11:20.040 ⇒ 01:11:25.289 Jasmin Multani: And saying, hey, case when order underscore sequels 1.
699 01:11:25.450 ⇒ 01:11:29.599 Jasmin Multani: then… Whatever the week.
700 01:11:29.900 ⇒ 01:11:31.789 Jasmin Multani: of order created at was.
701 01:11:48.400 ⇒ 01:11:53.669 Advait Nandakumar Menon: So this is the one which gave me… so you… I can just remove the limit, right?
702 01:11:55.050 ⇒ 01:11:57.050 Jasmin Multani: But even if you remove the limit.
703 01:11:58.390 ⇒ 01:12:04.010 Jasmin Multani: We’re not adding a systemic way to filter the weak.
704 01:12:05.260 ⇒ 01:12:09.809 Jasmin Multani: So, if I wanted to ask about what’s the first week of…
705 01:12:11.060 ⇒ 01:12:14.650 Jasmin Multani: Give me the output for the first week of March.
706 01:12:14.860 ⇒ 01:12:17.810 Jasmin Multani: Where should we plug that in?
707 01:12:23.640 ⇒ 01:12:29.699 Jasmin Multani: be, like, adjust this so, we get the… we also receive the week.
708 01:12:30.310 ⇒ 01:12:33.259 Jasmin Multani: Of the third date, in the third sequence.
709 01:12:43.390 ⇒ 01:12:49.510 Jasmin Multani: And then tell it that week’s start date. It’s supposed to be… Sunday?
710 01:12:49.920 ⇒ 01:12:53.100 Jasmin Multani: There’s something random then. Saturday. Saturday, yeah.
711 01:12:53.330 ⇒ 01:12:59.649 Jasmin Multani: tell it… tell AI, like, the week’s start date should start on the Saturday, and end on the Friday.
712 01:13:00.060 ⇒ 01:13:02.079 Jasmin Multani: I think that’s where it’s getting…
713 01:13:02.360 ⇒ 01:13:05.150 Jasmin Multani: weird, because I think SQL’s default is Sunday.
714 01:13:05.150 ⇒ 01:13:05.720 Advait Nandakumar Menon: Yeah.
715 01:13:05.720 ⇒ 01:13:06.580 Jasmin Multani: Saturday.
716 01:13:43.330 ⇒ 01:13:45.269 Jasmin Multani: Okay, let’s see.
717 01:13:56.850 ⇒ 01:14:01.330 Jasmin Multani: Can you send that code over to me? I want to play with it on Snowflake as well.
718 01:14:03.000 ⇒ 01:14:03.530 Advait Nandakumar Menon: Yep.
719 01:14:03.900 ⇒ 01:14:04.930 Jasmin Multani: Thank you.
720 01:15:35.600 ⇒ 01:15:43.890 Advait Nandakumar Menon: Yeah, it’s… It’s giving me the latest week started as 18th, and… I’m getting the same 29.
721 01:15:45.940 ⇒ 01:15:50.129 Jasmin Multani: Oh, 29, you’re getting 29 order? Or 29 customers?
722 01:15:50.130 ⇒ 01:15:52.220 Advait Nandakumar Menon: Customers with their third orders, yeah.
723 01:15:52.460 ⇒ 01:15:57.340 Jasmin Multani: Oh, perfect, so that is different, that is a little bit different than… 32.
724 01:15:57.730 ⇒ 01:15:58.959 Shivani Amar: What is his number?
725 01:15:59.390 ⇒ 01:16:01.600 Jasmin Multani: This number is 29.
726 01:16:02.080 ⇒ 01:16:04.620 Shivani Amar: Oh, but that’s not that different. Are your weeks the same?
727 01:16:07.210 ⇒ 01:16:09.440 Jasmin Multani: Week start date is 4-18.
728 01:16:09.440 ⇒ 01:16:11.449 Shivani Amar: No, I think that’s a different.
729 01:16:11.450 ⇒ 01:16:12.220 Jasmin Multani: Oh, man.
730 01:16:12.580 ⇒ 01:16:14.230 Shivani Amar: Yeah, so it’s fine.
731 01:16:14.330 ⇒ 01:16:20.460 Shivani Amar: So you’re getting close. Guys, we’re just checking, I hope you’re not like, this is a waste of time. We’re just trying to learn together, but…
732 01:16:23.000 ⇒ 01:16:31.639 Shivani Amar: The general… like, if we zoom out again, the general, like, takeaway is that, like, it’s a funny-shaped funnel if we have
733 01:16:32.890 ⇒ 01:16:36.999 Shivani Amar: more and more orders happening, so I’ve just been, like, curious about it.
734 01:16:37.110 ⇒ 01:16:41.439 Shivani Amar: And we can continue looking at it together with a wish. Does that sound good?
735 01:16:42.070 ⇒ 01:16:44.500 Jasmin Multani: Yeah, and then I’ll just be playing around.
736 01:16:44.730 ⇒ 01:16:46.619 Jasmin Multani: the CTE.
737 01:16:49.080 ⇒ 01:16:49.649 Shivani Amar: That’s why I…
738 01:16:49.650 ⇒ 01:16:51.410 Jasmin Multani: I can break this apart as well.
739 01:16:51.530 ⇒ 01:16:52.290 Shivani Amar: Yeah.
740 01:16:54.160 ⇒ 01:16:57.349 Shivani Amar: Because you did April 18th to the 23rd, right, of it?
741 01:16:58.020 ⇒ 01:16:58.990 Advait Nandakumar Menon: Yep, yep.
742 01:16:58.990 ⇒ 01:17:06.830 Shivani Amar: What we’re talking about is it would be the 18th to the 24th.
743 01:17:07.630 ⇒ 01:17:10.600 Shivani Amar: That’s probably why you’re under a little bit.
744 01:17:12.250 ⇒ 01:17:14.230 Shivani Amar: You’re just not including the 24th.
745 01:17:15.890 ⇒ 01:17:16.600 Advait Nandakumar Menon: Okay.
746 01:17:21.850 ⇒ 01:17:25.480 Shivani Amar: I think that’s what you’re… missing.
747 01:17:26.510 ⇒ 01:17:27.590 Shivani Amar: That’s interesting.
748 01:17:29.590 ⇒ 01:17:32.500 Shivani Amar: 8 days sounds weird, but I don’t… I don’t know.
749 01:17:34.760 ⇒ 01:17:46.809 Shivani Amar: But, whatever. We’re getting close. I think there’s some cleanup we can do around, like, $0 orders in general, not being an order number. And then I’m curious, like, what that does.
750 01:17:48.260 ⇒ 01:17:48.920 Shivani Amar: Okay?
751 01:17:49.580 ⇒ 01:18:04.670 Shivani Amar: Cool. And then, like, I’m imagining a funnel is usually like this, where it’s, like, first order, second, third. In this case, it’s, like, actually the other way, where it’s, like, first order, or not that big, but, like, first order, and then, like, second order is a little bit smaller, then third order is bigger than both of them.
752 01:18:05.120 ⇒ 01:18:07.229 Shivani Amar: And I’m like, that’s crazy.
753 01:18:07.350 ⇒ 01:18:11.930 Shivani Amar: But maybe that’s because… The third order has so many of these $0 orders.
754 01:18:11.930 ⇒ 01:18:14.480 Jasmin Multani: And it’s not the same cohort.
755 01:18:14.480 ⇒ 01:18:26.650 Shivani Amar: Yeah. I don’t… I don’t care… you’re gonna hear that I, like, never really care about cohort. I’m just kind of like, what happened in a time period? Eventually, if a VP wants a cohort analysis, we will give it to them, but I’m just like, what happened?
756 01:18:28.150 ⇒ 01:18:28.950 Shivani Amar: Okay?
757 01:18:29.150 ⇒ 01:18:29.980 Jasmin Multani: Sounds good.
758 01:18:29.980 ⇒ 01:18:33.950 Shivani Amar: Okay, everybody go take naps. This is during a sequel, okay?
759 01:18:33.950 ⇒ 01:18:34.720 Jasmin Multani: Good.
760 01:18:34.720 ⇒ 01:18:35.300 Shivani Amar: Okay.
761 01:18:35.300 ⇒ 01:18:35.860 Jasmin Multani: Thank you.
762 01:18:35.860 ⇒ 01:18:37.060 Shivani Amar: vibe Bye.
763 01:18:37.060 ⇒ 01:18:37.380 Jasmin Multani: here.
764 01:18:37.380 ⇒ 01:18:37.760 Advait Nandakumar Menon: game.
765 01:18:37.760 ⇒ 01:18:38.180 Jasmin Multani: often.
766 01:18:38.510 ⇒ 01:18:39.320 Jasmin Multani: Bye.