Meeting Title: Working session Date: 2025-06-13 Meeting participants: Caio Velasco, Emily Giant
WEBVTT
1 00:01:16.030 ⇒ 00:01:18.620 Emily Giant: Good morning or afternoon.
2 00:01:18.960 ⇒ 00:01:20.309 Emily Giant: What time is it there.
3 00:01:21.350 ⇒ 00:01:22.070 Caio Velasco: 3.
4 00:01:22.690 ⇒ 00:01:24.840 Emily Giant: Great. Okay? So you’re like.
5 00:01:25.110 ⇒ 00:01:29.929 Emily Giant: this is 3 o’clock when I start to get like real tired and kind of dumb.
6 00:01:30.080 ⇒ 00:01:31.530 Emily Giant: I think they really cool.
7 00:01:34.245 ⇒ 00:01:34.940 Caio Velasco: Exactly.
8 00:01:35.480 ⇒ 00:01:36.990 Caio Velasco: Yeah window.
9 00:01:38.150 ⇒ 00:01:38.859 Caio Velasco: I was. Gonna say.
10 00:01:39.690 ⇒ 00:01:54.829 Emily Giant: Oh, it’s it’s good! I could use another cup of coffee, but other than that, I’m I’m good I did finish up the the 2 like outstanding questions about quickbooks and changed that sink.
11 00:01:55.300 ⇒ 00:02:05.130 Emily Giant: So I don’t know if we need to update that ticket to done, or if we have to turn off the syncs to market is done. I forget. I need to pull that up.
12 00:02:05.740 ⇒ 00:02:14.830 Caio Velasco: So most of them are turned off already. And the email is ready. The only thing there is one missing, the one from evil. The.
13 00:02:16.070 ⇒ 00:02:17.589 Emily Giant: Oh, wow!
14 00:02:17.590 ⇒ 00:02:19.440 Caio Velasco: Yeah, yeah, that’s the only thing.
15 00:02:19.720 ⇒ 00:02:21.570 Emily Giant: Yep, that one will still be wrong.
16 00:02:21.860 ⇒ 00:02:25.239 Caio Velasco: Okay, cool, though. Yeah. So I think the email is.
17 00:02:26.040 ⇒ 00:02:28.060 Caio Velasco: And then we can turn that
18 00:02:28.310 ⇒ 00:02:32.900 Caio Velasco: ticket done. Yeah, let me put it like.
19 00:02:32.900 ⇒ 00:02:36.369 Emily Giant: Did work on the the dashboards
20 00:02:36.550 ⇒ 00:02:38.770 Emily Giant: for a while yesterday. There’s still like
21 00:02:39.500 ⇒ 00:02:44.972 Emily Giant: a handful of maybes that I need to reach out to the stakeholders. So I will do that
22 00:02:45.960 ⇒ 00:02:54.885 Emily Giant: like this morning, so that we can get that across the finish line. But I think it’s only let me see.
23 00:03:00.820 ⇒ 00:03:04.790 Emily Giant: it’s not many that I still have to to go through.
24 00:03:07.008 ⇒ 00:03:11.290 Caio Velasco: Will update the ticket now, and so your loyalty keeps.
25 00:03:11.480 ⇒ 00:03:12.360 Emily Giant: That stays.
26 00:03:12.740 ⇒ 00:03:13.549 Caio Velasco: Based on the tool
27 00:03:17.350 ⇒ 00:03:18.250 Caio Velasco: greeting.
28 00:03:18.860 ⇒ 00:03:19.539 Caio Velasco: Hold on.
29 00:03:22.430 ⇒ 00:03:23.790 Emily Giant: Like many of these.
30 00:03:24.060 ⇒ 00:03:27.970 Emily Giant: So 1, 2, 3, 4, 5, 6,
31 00:03:28.230 ⇒ 00:03:36.779 Emily Giant: 7, 8, 8 of them are like system dashboards in Looker. So I’m not sure like if we can delete those.
32 00:03:36.930 ⇒ 00:03:39.170 Emily Giant: That was the maybe on those.
33 00:03:40.730 ⇒ 00:03:43.509 Caio Velasco: The which ones where.
34 00:03:43.660 ⇒ 00:03:46.349 Emily Giant: I’m like just talking randomly without sharing my screen.
35 00:03:46.350 ⇒ 00:03:46.690 Caio Velasco: Here we go!
36 00:03:46.690 ⇒ 00:03:55.608 Emily Giant: My penalty is the not sharing screen. And then talking like what I’m talking about.
37 00:03:57.200 ⇒ 00:04:00.400 Emily Giant: Can you see my spreadsheet.
38 00:04:00.750 ⇒ 00:04:01.970 Caio Velasco: Yes, I did.
39 00:04:01.970 ⇒ 00:04:08.909 Emily Giant: It’s it’s these, the the usage reports that I’m not sure if we can delete those.
40 00:04:09.860 ⇒ 00:04:11.760 Emily Giant: Not that we need to, but.
41 00:04:11.760 ⇒ 00:04:12.690 Caio Velasco: Yeah.
42 00:04:12.690 ⇒ 00:04:18.199 Emily Giant: Maybe just because I needed to investigate, like what what those are.
43 00:04:19.459 ⇒ 00:04:21.169 Caio Velasco: They’re only gonna put a.
44 00:04:23.550 ⇒ 00:04:24.620 Emily Giant: Let me see.
45 00:04:35.040 ⇒ 00:04:40.510 Emily Giant: Oh, yeah, these are all like system dashboards that can stay. I actually do use these.
46 00:04:41.550 ⇒ 00:04:42.260 Caio Velasco: Okay.
47 00:04:43.190 ⇒ 00:04:46.809 Caio Velasco: So the looker use it once.
48 00:04:47.480 ⇒ 00:04:48.360 Emily Giant: Yes.
49 00:04:49.180 ⇒ 00:04:50.300 Caio Velasco: So.
50 00:04:50.600 ⇒ 00:04:54.090 Caio Velasco: So I’ll put a note index.
51 00:05:05.020 ⇒ 00:05:13.880 Caio Velasco: So, as I said, when I look at the deprecate column, there’s no blank, so everything has either yes no, or maybe already.
52 00:05:14.270 ⇒ 00:05:15.010 Emily Giant: Yeah.
53 00:05:15.260 ⇒ 00:05:22.180 Emily Giant: it should. There should only be those 3. So I’ll reach out. Jessica is the owner of most of the ones. I didn’t know
54 00:05:22.290 ⇒ 00:05:25.260 Emily Giant: if I could delete them.
55 00:05:25.810 ⇒ 00:05:30.809 Emily Giant: So here, let me let me just write to her.
56 00:05:33.400 ⇒ 00:05:35.800 Caio Velasco: I can see your screen just in case you you.
57 00:05:35.800 ⇒ 00:05:37.176 Emily Giant: Oh, no, it’s fine.
58 00:05:38.970 ⇒ 00:05:43.110 Emily Giant: I know people have gotten in trouble showing slack messages, but
59 00:05:44.110 ⇒ 00:05:46.539 Emily Giant: I keep it. I keep it pretty real on mine.
60 00:05:46.540 ⇒ 00:05:47.190 Caio Velasco: Obviously.
61 00:05:48.440 ⇒ 00:05:57.669 Emily Giant: Mostly, it’s just like a hundred unread messages. And people like, Are you okay, are you there? Because I’m I don’t know how people that work in data actually
62 00:05:58.040 ⇒ 00:06:02.540 Emily Giant: communicate with anyone else. We’re like, head down all the time.
63 00:06:03.000 ⇒ 00:06:03.750 Caio Velasco: Yeah.
64 00:06:32.970 ⇒ 00:06:40.610 Emily Giant: Great the rest I’ll reach out at another time. Most of them that I marked as maybe.
65 00:06:41.370 ⇒ 00:07:05.583 Emily Giant: Reports that, like former employees made, but people still seem to be looking at them. Now I think that they’re deprecated outright. If somebody who worked here 3 years ago made it. There’s no way that data pipeline is. Still, those models are like still intact. So I just more or less want to let them know that the plan is to rebuild these with the new data.
66 00:07:06.100 ⇒ 00:07:18.099 Emily Giant: but I want to get a timeline on like if there is anything in there that they’re relying on like. There might be one report in the dashboard that we can pull out as a look. I just need like better instruction there
67 00:07:18.510 ⇒ 00:07:25.510 Emily Giant: from the but I I think we will wind up deprecating all of them. But
68 00:07:25.820 ⇒ 00:07:30.369 Emily Giant: I just wanna like get the rundown before making the executive call.
69 00:07:33.230 ⇒ 00:07:34.200 Emily Giant: All right.
70 00:07:34.590 ⇒ 00:07:39.480 Emily Giant: I have written to Jess the links to all the reports.
71 00:07:43.490 ⇒ 00:07:46.210 Emily Giant: I know this one can’t be deleted.
72 00:07:48.030 ⇒ 00:07:56.980 Emily Giant: Okay, so did you want to go into the Dbt models, since, like these are all kind of in a holding pattern. I thought.
73 00:07:57.400 ⇒ 00:07:58.969 Caio Velasco: Yeah, yeah, we can definitely do that.
74 00:07:59.580 ⇒ 00:08:06.590 Caio Velasco: Just just a question, how? How many? Just by curiosity, how many are not to be deprecated? How many ends?
75 00:08:08.280 ⇒ 00:08:10.750 Caio Velasco: I mean, those are definitely the important ones.
76 00:08:11.670 ⇒ 00:08:13.580 Caio Velasco: Can I unfilter this or.
77 00:08:13.580 ⇒ 00:08:14.830 Emily Giant: Yeah, go for it.
78 00:08:14.830 ⇒ 00:08:15.390 Caio Velasco: Yep.
79 00:08:25.290 ⇒ 00:08:29.784 Caio Velasco: 55. 0, wow! Okay. So from 800 to 55, that’s progress.
80 00:08:33.409 ⇒ 00:08:33.869 Caio Velasco: Oh, shit.
81 00:08:33.870 ⇒ 00:08:34.750 Emily Giant: Fine.
82 00:08:36.080 ⇒ 00:08:38.489 Caio Velasco: Great. Great. Great. Okay. Cool.
83 00:08:38.919 ⇒ 00:08:43.079 Emily Giant: Yeah, that’s definitely gonna feel that’s gonna feel good.
84 00:08:43.440 ⇒ 00:08:49.420 Caio Velasco: Yep, and then I’ll also see that the Dbt. Part it’s on the Dbt. Tab.
85 00:08:51.180 ⇒ 00:08:54.250 Emily Giant: Yes, the right.
86 00:08:55.500 ⇒ 00:08:59.220 Caio Velasco: Models. Yeah, there was another one.
87 00:08:59.920 ⇒ 00:09:07.430 Emily Giant: Okay, yeah, this is the one where I was ranking them and giving a rundown of what they are.
88 00:09:08.661 ⇒ 00:09:28.338 Emily Giant: My thought was to go through, since revenue is the focus of the next sprint. I didn’t spend a lot of time with those models during this process so far, because I was thinking that we could go through them together, kind of show you in Dbt, where they’re breaking down and
89 00:09:29.720 ⇒ 00:09:35.409 Emily Giant: and they’re all 10 like they don’t work so like I don’t know what else to write, but, like.
90 00:09:35.810 ⇒ 00:09:59.329 Caio Velasco: So maybe here one thing that I was missing, since I didn’t see also layers that we are, gonna we’re gonna do like raw stage marks. I think that’s what I don’t see here, and maybe it’s helpful. When we focus in those, maybe we can add another column here to point the business unit or business domain, or marks.
91 00:09:59.510 ⇒ 00:10:04.880 Caio Velasco: just to have an idea, because well read by reading. I cannot say what is what.
92 00:10:05.770 ⇒ 00:10:11.724 Emily Giant: Oh, me, neither. And I work here. So not a good sign
93 00:10:13.000 ⇒ 00:10:17.480 Emily Giant: like, for the most part, I know that, like Google Ads is marketing
94 00:10:18.170 ⇒ 00:10:18.680 Caio Velasco: Exactly.
95 00:10:18.680 ⇒ 00:10:30.925 Emily Giant: Marketing. But then, when you get into some of the others, I mean, I’m kidding. I do know what most of them are, but like. What what I mean to say is, the names are not intuitive at all.
96 00:10:31.710 ⇒ 00:10:33.639 Caio Velasco: And what I see, my bad.
97 00:10:33.640 ⇒ 00:10:35.360 Emily Giant: Go ahead! Go ahead!
98 00:10:35.670 ⇒ 00:10:42.520 Caio Velasco: I was just gonna ask like, from what I see here this was. You were rating them from one to 1010 being
99 00:10:42.940 ⇒ 00:10:46.480 Caio Velasco: the most reliable, and one the least reliable.
100 00:10:46.880 ⇒ 00:10:48.460 Emily Giant: Oh! Opposite!
101 00:10:49.700 ⇒ 00:10:50.260 Caio Velasco: No problem.
102 00:10:50.260 ⇒ 00:10:57.870 Emily Giant: Color like the red. I was like red looks broken. So 10 is gonna be the the most broken.
103 00:10:58.430 ⇒ 00:11:01.880 Caio Velasco: Okay, no perfect. I just put a comment on the column.
104 00:11:05.030 ⇒ 00:11:10.370 Caio Velasco: Can I also add the type? Because I know that at the end you have green, and you had green and
105 00:11:10.880 ⇒ 00:11:15.209 Caio Velasco: yellow, but I just added, as a new column, like what is source and what is seat.
106 00:11:16.690 ⇒ 00:11:19.830 Emily Giant: Yeah, that’s a good call.
107 00:11:21.680 ⇒ 00:11:28.270 Emily Giant: There aren’t many seeds, but for example, all of these holiday.
108 00:11:29.130 ⇒ 00:11:36.167 Emily Giant: So all of these, like budget purchase feeds holiday component forecast any of these.
109 00:11:37.000 ⇒ 00:11:41.549 Emily Giant: like Sn of P forecast titled models.
110 00:11:42.070 ⇒ 00:11:57.630 Emily Giant: Every holiday the the planning team will start a new document like a new Google sheet for upload. And it’s just so that, like our holiday specific metrics are are
111 00:11:58.607 ⇒ 00:12:03.509 Emily Giant: compared to like more granular kpis.
112 00:12:03.710 ⇒ 00:12:09.349 Emily Giant: But there’s no real like process here, and I wind up
113 00:12:09.680 ⇒ 00:12:19.260 Emily Giant: like in the column, I said. Like lacks a consistent stakeholder submission process. While these are like reliable.
114 00:12:19.420 ⇒ 00:12:29.770 Emily Giant: the they’re not efficient. I fail to understand why we need a new document every holiday, when, like
115 00:12:29.910 ⇒ 00:12:35.660 Emily Giant: it should be the same columns, it should be the same as the previous one, just like
116 00:12:36.220 ⇒ 00:12:37.960 Emily Giant: marking the year.
117 00:12:38.140 ⇒ 00:12:41.280 Emily Giant: And was
118 00:12:42.200 ⇒ 00:12:48.579 Emily Giant: wanting to call that out in case like the move is to make a seed file, or something like that, where they submit
119 00:12:48.880 ⇒ 00:13:17.830 Emily Giant: whatever to me. But instead of me creating a new sink in stitch, I go in and change the seed file, or they do occasionally change the numbers throughout the holiday. But it still seems like there should be a more like this is the holiday forecast document. This is what we use all the time like, why do we have 15 different spreadsheets and 15 different models that I have to join or union? And it really clutters up.
120 00:13:17.980 ⇒ 00:13:19.000 Emily Giant: Dbt,
121 00:13:20.410 ⇒ 00:13:26.200 Emily Giant: Do you see that in a lot of businesses that, like they have 15 different documents for the same
122 00:13:27.630 ⇒ 00:13:28.180 Emily Giant: thing.
123 00:13:29.450 ⇒ 00:13:34.699 Caio Velasco: So yeah, from what you’re saying, it makes sense. That should be something close to to a seat.
124 00:13:34.940 ⇒ 00:13:36.819 Caio Velasco: It doesn’t make sense. Yeah.
125 00:13:37.760 ⇒ 00:13:42.830 Emily Giant: Yeah. And the only part that is like it feels like
126 00:13:42.950 ⇒ 00:13:55.049 Emily Giant: once the holidays over. Put that in a seed file and then use that same Google Sheet, or whatever it is, forever to be changing the dates. Not like
127 00:13:55.480 ⇒ 00:14:01.690 Emily Giant: I don’t know. It just gets very like messy and stitch half of what we turned off was
128 00:14:02.560 ⇒ 00:14:07.310 Emily Giant: planning document from Snop. And of course they like, don’t
129 00:14:08.570 ⇒ 00:14:27.050 Emily Giant: I like. My idea, I guess, is to create the document that they update so that the titles that there’s no like white space, they leave white space all the time. They leave, like titles that I have to like completely render for every single column, because they’re leaving spaces. They’re not. They’re not normalized.
130 00:14:27.050 ⇒ 00:14:38.990 Emily Giant: And so it takes a day of mine in order to get these forecast documents to work. And then I have to go into looker and like, create a new view and create a new because it’s a new document. So I can’t just like.
131 00:14:39.700 ⇒ 00:14:40.050 Caio Velasco: In this.
132 00:14:40.050 ⇒ 00:14:43.529 Emily Giant: Going to join it like they’ve gotten better
133 00:14:43.740 ⇒ 00:14:48.779 Emily Giant: like the since I’ve been in this seat like I gave them a document.
134 00:14:49.100 ⇒ 00:15:10.790 Emily Giant: and that then I didn’t have to create a new looker view. I could just like join it in Dbt, and it would pick through the same thing. But, like in in the past. A lot of what you’re going to see in Looker when you start auditing the views. Is this garbage of like old Holiday, Google sheet forecast documents. So these are all
135 00:15:11.320 ⇒ 00:15:12.580 Emily Giant: sources.
136 00:15:13.290 ⇒ 00:15:17.770 Caio Velasco: Not this one. This one is a but no issue.
137 00:15:17.770 ⇒ 00:15:22.319 Emily Giant: Let’s see. But you can just forecast I’m gonna filter for the word forecast.
138 00:15:22.500 ⇒ 00:15:27.429 Emily Giant: Just to give you an idea of how many of these models, our Google sheets.
139 00:15:31.950 ⇒ 00:15:34.980 Caio Velasco: I’ll also open the structure so that they can.
140 00:15:35.480 ⇒ 00:15:36.470 Caio Velasco: Okay.
141 00:15:37.660 ⇒ 00:15:49.230 Emily Giant: So, yeah, I mean, these are all like, they all work. They’re not malfunctioning. They’re just.
142 00:15:50.440 ⇒ 00:15:51.060 Caio Velasco: Hmm.
143 00:15:51.090 ⇒ 00:15:52.479 Emily Giant: Ideal. So I’m not.
144 00:15:52.480 ⇒ 00:15:55.919 Caio Velasco: We have this 2 way with it. Yeah, yeah, okay.
145 00:15:56.800 ⇒ 00:16:02.300 Caio Velasco: yeah, maybe just make a that comment over there. And we can take a look at it.
146 00:16:07.720 ⇒ 00:16:09.470 Emily Giant: Yeah, these are all.
147 00:16:14.020 ⇒ 00:16:15.310 Emily Giant: Hello.
148 00:16:16.280 ⇒ 00:16:25.310 Emily Giant: yeah, these are all the same thing. And then there’s a couple that like it looks like, maybe a new person took the seat.
149 00:16:25.410 ⇒ 00:16:26.365 Emily Giant: And
150 00:16:28.070 ⇒ 00:16:35.380 Emily Giant: So one of these forecast documents, for example, is like outbound forecast. And then there’s map type forecast.
151 00:16:38.120 ⇒ 00:16:41.800 Emily Giant: Here, this holiday service forecast. And then there’s
152 00:16:42.390 ⇒ 00:16:49.180 Emily Giant: the map type forecast. From what I can see, they’re the same thing. But 2 different people had the job. So
153 00:16:49.900 ⇒ 00:16:55.329 Emily Giant: I did leave notes in both of those columns. They just are separate models.
154 00:16:55.460 ⇒ 00:17:04.172 Emily Giant: And so like, we’re skipping years depending on which model the looker user is using
155 00:17:05.109 ⇒ 00:17:13.550 Emily Giant: because they’re not joined. So it’s just those are less reliable. The holiday service forecast, the holiday map type forecast
156 00:17:13.670 ⇒ 00:17:18.810 Emily Giant: than some of these other ones that are at the very least joined together.
157 00:17:19.624 ⇒ 00:17:25.280 Emily Giant: In looker, so that you can get a historical hindsight.
158 00:17:27.430 ⇒ 00:17:31.240 Emily Giant: But again, not sure how to rate them. I’ll give them a 3, because they’re like
159 00:17:31.620 ⇒ 00:17:37.549 Emily Giant: they’re kind of just janky, but they don’t work.
160 00:17:37.780 ⇒ 00:17:41.227 Emily Giant: But yeah, you’ll see in Looker. They’re a mess.
161 00:17:41.740 ⇒ 00:17:46.459 Emily Giant: alright. So the ones that are related to revenue.
162 00:17:50.570 ⇒ 00:17:58.780 Emily Giant: So this is some of the work I was doing with products. It starts with oms, suborders.
163 00:17:59.040 ⇒ 00:18:06.819 Emily Giant: The 1st calculation that is actually like logic based is in Oms suborders.
164 00:18:07.520 ⇒ 00:18:10.590 Emily Giant: And then it’s a good spot to start, because
165 00:18:11.400 ⇒ 00:18:25.889 Emily Giant: it will point you towards all of the other staging models that it uses in order to start calculating revenue, but, like from the very 1st place where the piece is calculated, it’s broken.
166 00:18:26.590 ⇒ 00:18:30.009 Emily Giant: So it starts with item total.
167 00:18:34.380 ⇒ 00:18:39.550 Emily Giant: So okay, so this is the logic. Starting with
168 00:18:40.250 ⇒ 00:18:50.080 Emily Giant: 4, 24 down to 4, 44. That calculates the 1st piece, and essentially saying that like
169 00:18:50.570 ⇒ 00:18:52.879 Emily Giant: don’t count if it’s a redelivery.
170 00:18:53.050 ⇒ 00:18:58.695 Emily Giant: if it’s a subscription, this piece actually is somewhat sound.
171 00:18:59.730 ⇒ 00:19:19.100 Emily Giant: But when you look back at the subscription model, which is called loop subscriptions, Xf. I did rate it as maybe like a 7 on the unreliable scale. Mostly because it’s all transformation. It’s all logic, and none of it comes from the source so ideally we would want
172 00:19:19.816 ⇒ 00:19:41.990 Emily Giant: we would want our subscription platform to be like feeding us that information, instead of taking orders from our order management system and then putting a price on top of them. So, for example, there’s like 3 types of subscriptions, and they’ll come in as 1,200 that needs to be amortized over
173 00:19:42.480 ⇒ 00:19:46.339 Emily Giant: 7 months. I have to say.
174 00:19:46.550 ⇒ 00:19:58.520 Emily Giant: okay, look at the kind of subscription it is. Remove the 50 on top of it every time this renews. So there’s just a lot of like room for error in how it’s built.
175 00:19:58.980 ⇒ 00:20:02.480 Emily Giant: And then, during the migration period, from like
176 00:20:02.720 ⇒ 00:20:13.250 Emily Giant: November 6th of 2024 to January 1st it was completely denormalized. It was that period of time is kind of the like.
177 00:20:14.180 ⇒ 00:20:26.449 Emily Giant: the snow globe of bad, where we’re gonna continually run into all of the pieces that deprecated after the migration, and things got a little bit better starting on the first, st because
178 00:20:26.660 ⇒ 00:20:35.050 Emily Giant: I had communicated enough with the Dev team that I could get those elements from the payload that I needed to get the data back on track.
179 00:20:35.160 ⇒ 00:20:41.860 Emily Giant: But part of me thinks we should just like forget the snow globe, because
180 00:20:42.030 ⇒ 00:20:46.530 Emily Giant: there is there are no answers there unless we do like.
181 00:20:47.322 ⇒ 00:20:49.510 Emily Giant: What’s the word when you go back
182 00:20:49.750 ⇒ 00:21:06.540 Emily Giant: backfill, unless we backfill every order that was created during that time. So anyhow, this this part says, if it’s a subscription, there’s a set cost, ignore what is actually in the total in the transaction line. Use this instead.
183 00:21:07.200 ⇒ 00:21:09.695 Emily Giant: Then we start to get crazy.
184 00:21:10.680 ⇒ 00:21:11.300 Emily Giant: So
185 00:21:12.460 ⇒ 00:21:24.450 Emily Giant: I mean, you can see it if we start to time gate things and then say, it’s not a b 2 b email. So it’s got to be a business to customer, not business to business.
186 00:21:24.940 ⇒ 00:21:53.539 Emily Giant: taking promo codes into account. And then this is the part that, like all of that is garbage. But what gets really crazy is this so the T dot subtotal that comes from a table called staging totals after the migration. It’s like all of the columns that used to be used to calculate. Item total stopped working. And now there’s ones called like new, subtotal new item total that are very all over the place.
187 00:21:53.650 ⇒ 00:21:59.770 Emily Giant: Staging totals has to go during this process. We need to like, take
188 00:22:00.586 ⇒ 00:22:10.840 Emily Giant: whatever was happening before. Put it in a model. Say, this used to be reliable and going forward, use, shopify, or something else, whatever the team determines. But
189 00:22:11.150 ⇒ 00:22:12.490 Emily Giant: that is
190 00:22:13.730 ⇒ 00:22:20.240 Emily Giant: staging totals is, there’s a lot that goes into this oms suborders. It’s 1 of like the hubs
191 00:22:20.360 ⇒ 00:22:26.240 Emily Giant: of every order, line, order
192 00:22:27.130 ⇒ 00:22:34.660 Emily Giant: subtotal. It’s where it all comes together the 1st time. So that one is.
193 00:22:36.310 ⇒ 00:22:40.780 Emily Giant: It’s 1 of these very early on staging totals, but
194 00:22:41.000 ⇒ 00:22:48.420 Emily Giant: very unreliable, very broken. So if I haven’t done that one yet, I should
195 00:22:53.010 ⇒ 00:22:54.300 Emily Giant: staging totals.
196 00:22:58.790 ⇒ 00:23:00.100 Emily Giant: Yep, okay.
197 00:23:01.640 ⇒ 00:23:08.312 Emily Giant: So yeah, this one I would recommend just pulling like a hundred rows to see how
198 00:23:09.520 ⇒ 00:23:19.059 Emily Giant: how we started adding new totals after the migration, and how they don’t really align to what’s in shopify.
199 00:23:20.480 ⇒ 00:23:21.080 Emily Giant: Most.
200 00:23:21.080 ⇒ 00:23:25.839 Caio Velasco: One quick. One quick question more over, more like a larger overview. It’s a
201 00:23:26.070 ⇒ 00:23:29.120 Caio Velasco: so I see that the rating it’s
202 00:23:29.350 ⇒ 00:23:34.080 Caio Velasco: to show what what is reliable, what is not reliable. But maybe.
203 00:23:34.690 ⇒ 00:23:49.199 Caio Velasco: as you just said, oms, supporters is the most important one. But we can kind of. We see a 7, which is like being unreliable. But we still want to do that. So I I’m kind of missing. Maybe we can also track the order of the importance.
204 00:23:49.550 ⇒ 00:23:55.090 Caio Velasco: Yeah, like, okay, that’s super important and super unreliable. Man definitely have to take a look at this.
205 00:23:55.650 ⇒ 00:24:00.790 Emily Giant: Yeah, a 1 on importance, meaning extremely important, and
206 00:24:01.020 ⇒ 00:24:15.529 Emily Giant: 7 on. There are many elements of it that actually are reliable, that are used like we pass through delivery details, customer details. We pass too many things through this model to be fair, that are then carried through
207 00:24:15.910 ⇒ 00:24:27.739 Emily Giant: 50 other ones before they hit the mart, and so the ones that are actually reliable in oms suborders by the time they get to the the mart.
208 00:24:28.090 ⇒ 00:24:30.380 Caio Velasco: Not I get it?
209 00:24:30.550 ⇒ 00:24:31.200 Emily Giant: Yeah.
210 00:24:32.250 ⇒ 00:24:36.420 Caio Velasco: Oh, okay. Still. Maybe time. Rating.
211 00:24:38.260 ⇒ 00:24:40.049 Emily Giant: So staging tools.
212 00:24:41.350 ⇒ 00:24:44.620 Caio Velasco: Of. Maybe here we can just do.
213 00:24:47.550 ⇒ 00:24:50.949 Caio Velasco: How do they call like importance, or something?
214 00:24:51.210 ⇒ 00:24:52.280 Emily Giant: Yes.
215 00:24:53.050 ⇒ 00:24:58.819 Caio Velasco: Then we can like, make, how many levels do you think? Like, just 3, like, nothing
216 00:24:59.820 ⇒ 00:25:02.390 Caio Velasco: very important. Yeah. Okay, so let me update this.
217 00:25:14.730 ⇒ 00:25:17.209 Caio Velasco: It may be just the zoom stuff.
218 00:25:22.920 ⇒ 00:25:28.720 Caio Velasco: Let me just unfilter everything and then come back to one sec.
219 00:25:28.720 ⇒ 00:25:30.430 Emily Giant: You do what you need to do.
220 00:25:44.170 ⇒ 00:25:47.410 Emily Giant: Yeah, you can even see, like in this staging totals.
221 00:25:47.660 ⇒ 00:25:59.729 Emily Giant: I went through and made some notes like during the migration of like, well, this part’s not gonna work anymore. And it doesn’t work in any of the other ones, either. And yeah, the one is
222 00:26:01.360 ⇒ 00:26:04.719 Emily Giant: bad revenue calculations from the source.
223 00:26:07.580 ⇒ 00:26:13.890 Emily Giant: All right. So staging totals is the 1st thing that goes sideways
224 00:26:14.150 ⇒ 00:26:17.039 Emily Giant: pretty badly in this revenue calculation.
225 00:26:18.540 ⇒ 00:26:34.729 Emily Giant: and then the subscription portion of it isn’t problematic as well because of how subscriptions change during the migration. It went from being a set number of things to promo codes are now eligible on subscriptions, the way that shopify
226 00:26:35.128 ⇒ 00:26:51.261 Emily Giant: discounts a prepaid subscription. It’s not an actual discount. It’s just their way of like not charging an order, and they a hundred percent discount it. But what that’s doing. In that subscription model that goes all the way downstream to revenue reporting is
227 00:26:51.860 ⇒ 00:27:02.150 Emily Giant: it makes it look like we’re giving a hundred percent discounts to customers. So it’s throwing off that kpi of how many discount dollars. So
228 00:27:02.620 ⇒ 00:27:06.409 Emily Giant: if somebody spends $1,200 on a subscription.
229 00:27:07.160 ⇒ 00:27:14.809 Emily Giant: even if the logic is overlaying and turning it to a hundred 50, it’s still saying like a hundred percent
230 00:27:15.560 ⇒ 00:27:18.730 Emily Giant: discount in the top line.
231 00:27:19.080 ⇒ 00:27:21.210 Emily Giant: So that’s 1 of the issues
232 00:27:21.320 ⇒ 00:27:29.540 Emily Giant: that’s interfering with revenue. All of these other totals are
233 00:27:29.960 ⇒ 00:27:34.039 Emily Giant: somewhat reliable. But I do want to call out a part that’s
234 00:27:34.880 ⇒ 00:27:43.479 Emily Giant: skewing revenue, and always has. This existed prior to the migration. And it’s a cte in here called line item split.
235 00:27:44.800 ⇒ 00:27:55.370 Emily Giant: And what it’s doing is assigning, like a percent of a suborder, that a line item is
236 00:27:55.480 ⇒ 00:27:58.309 Emily Giant: so that in the event of like
237 00:27:58.480 ⇒ 00:28:05.160 Emily Giant: somebody ordering a kit or an item where, like there’s a bouquet in a vase. But you’re only to have
238 00:28:05.510 ⇒ 00:28:13.889 Emily Giant: one line per order in this in this model. So sub orders can have, like
239 00:28:14.120 ⇒ 00:28:22.110 Emily Giant: 3 or 4 lines, because you can get like a candle, a bouquet, a vase. This needs to jam them all into one line.
240 00:28:22.340 ⇒ 00:28:39.739 Emily Giant: And it will say, Okay, based on this calculation, the bouquet is 70% of the order, the bases 10% and the candles 10%. And then, further down in item total, based on
241 00:28:40.530 ⇒ 00:28:41.590 Emily Giant: what
242 00:28:41.830 ⇒ 00:28:56.700 Emily Giant: what that line looks like. It is going to use that line item, split transformation to break it into parts. But it’s it’s never right. It’s just like a logical assumption that over time has really thrown off
243 00:28:57.000 ⇒ 00:28:58.890 Emily Giant: the reality of revenue.
244 00:28:59.130 ⇒ 00:28:59.935 Emily Giant: And
245 00:29:01.130 ⇒ 00:29:04.460 Emily Giant: I guess what my question. My, I question
246 00:29:05.990 ⇒ 00:29:13.079 Emily Giant: the utility of this model because of having to use that kind of calculation to calculate revenue like if we’re
247 00:29:13.530 ⇒ 00:29:20.434 Emily Giant: if we’re having to do that like. Is that how we want to calculate revenue in the 1st place, because that doesn’t make sense to me.
248 00:29:21.720 ⇒ 00:29:30.679 Emily Giant: so that, I think, is one of the more problematic areas of the whole Oms suborders. And you’ll see that like it’s used
249 00:29:31.340 ⇒ 00:29:32.810 Emily Giant: line item split.
250 00:29:37.310 ⇒ 00:29:46.490 Emily Giant: It’s used a lot in this in order to calculate tax total. Promo total.
251 00:29:49.470 ⇒ 00:29:58.990 Emily Giant: All all manner of financial metrics are using an assumption, and I don’t like it.
252 00:30:00.120 ⇒ 00:30:04.660 Emily Giant: So that was, this is like one of the bigger areas I wanted to call out.
253 00:30:05.000 ⇒ 00:30:09.367 Emily Giant: And then, if you look down the line,
254 00:30:10.850 ⇒ 00:30:19.830 Emily Giant: actually let me back it up. This is like the central hub. But where revenue is currently going very sideways
255 00:30:20.000 ⇒ 00:30:25.180 Emily Giant: is the staging line items and staging split line items.
256 00:30:25.410 ⇒ 00:30:28.850 Emily Giant: So this one, let me pull this up.
257 00:30:33.390 ⇒ 00:30:42.739 Emily Giant: So staging split line items I occasionally confuse what’s in each. But what happened was
258 00:30:43.960 ⇒ 00:30:45.640 Emily Giant: prior to the migration.
259 00:30:46.970 ⇒ 00:30:51.079 Emily Giant: Let me see which one this is because it it does matter before I start talking. I don’t wanna.
260 00:30:52.990 ⇒ 00:30:55.870 Emily Giant: I don’t want to make this more convoluted than it already is.
261 00:30:58.800 ⇒ 00:30:59.500 Emily Giant: Okay.
262 00:31:00.190 ⇒ 00:31:06.569 Emily Giant: So prior to the migration, if somebody bought a kit which is going to be like this, flrl, dash k
263 00:31:08.330 ⇒ 00:31:10.410 Emily Giant: a kit always contains multiple items.
264 00:31:11.170 ⇒ 00:31:13.629 Emily Giant: Only one of the models would get the kit.
265 00:31:14.130 ⇒ 00:31:23.749 Emily Giant: the other model would get the components of the kit, so both of these would have their own flow and would
266 00:31:24.630 ⇒ 00:31:32.239 Emily Giant: be used in order to like more accurately calculate that split line item, situation
267 00:31:32.410 ⇒ 00:31:34.469 Emily Giant: that we see in Oms suborders.
268 00:31:34.590 ⇒ 00:31:52.100 Emily Giant: After the migration these no longer went through their individual pipes like they both go every direction in in the dag, and it’s causing duplication it’s causing
269 00:31:54.620 ⇒ 00:32:01.400 Emily Giant: It’s causing like units delivered to be wrong, because it’s counting like a kit.
270 00:32:01.960 ⇒ 00:32:08.260 Emily Giant: and then again, the components of the kit in certain circumstances. But
271 00:32:08.610 ⇒ 00:32:13.534 Emily Giant: this lineage with the pieces of each order.
272 00:32:14.790 ⇒ 00:32:24.320 Emily Giant: This is heavily deprecated out of 10, and then where the 2 come together, this is one of the.
273 00:32:27.880 ⇒ 00:32:32.450 Emily Giant: So we’ve got this base of the split line items and the staging line items
274 00:32:33.140 ⇒ 00:32:36.850 Emily Giant: and strikethrough is if something in the kit gets canceled.
275 00:32:37.190 ⇒ 00:32:52.250 Emily Giant: So this whole model. The purpose is to say, Okay, we know that, like these 2 flows are the same orders, but represented in different ways. In one it will represent it as a bundle and one. It’s the pieces of the bundle
276 00:32:52.600 ⇒ 00:32:54.930 Emily Giant: if somebody
277 00:32:56.120 ⇒ 00:32:58.800 Emily Giant: Sorry. There’s a better way to say this, and I wrote it down.
278 00:32:59.120 ⇒ 00:33:03.560 Emily Giant: It’s that one of the models
279 00:33:03.680 ⇒ 00:33:12.100 Emily Giant: is the order the pieces that are fulfilled and actually sent to the customer, and the other are the pieces they saw online and ordered.
280 00:33:12.390 ⇒ 00:33:18.599 Emily Giant: And as you’ve learned at urban stems, not always the same thing, because flowers die so
281 00:33:18.800 ⇒ 00:33:26.080 Emily Giant: so often. What is ordered is not actually what’s fulfilled. And this is the model that determines what is fulfilled.
282 00:33:26.280 ⇒ 00:33:29.870 Emily Giant: This is the one that says, Okay, you’re looking at staging line items.
283 00:33:30.250 ⇒ 00:33:33.480 Emily Giant: This is what they ordered. Do we have it? Oh, we don’t
284 00:33:33.760 ⇒ 00:33:49.579 Emily Giant: cross out that line, but because there is no commonality between, there is no join between these tables anymore. It doesn’t know to delete lines. So anytime there’s a forced upgrade.
285 00:33:50.120 ⇒ 00:33:58.470 Emily Giant: It still looks in the data as though it was charged and sent because of a missing join. So
286 00:33:58.610 ⇒ 00:34:02.400 Emily Giant: this to me, is like red, alert.
287 00:34:03.940 ⇒ 00:34:10.260 Emily Giant: Misreporting revenue, misreporting units sold, and this is like the crux of it.
288 00:34:12.120 ⇒ 00:34:20.409 Emily Giant: So I’ll mark that on in the I cannot.
289 00:34:20.699 ⇒ 00:34:21.400 Emily Giant: We’ll wait.
290 00:34:21.400 ⇒ 00:34:33.149 Caio Velasco: The idea. The idea of flowers dying before being delivered it’s could be even interpret as like, Oh, someone stole it, you know, could be another use case.
291 00:34:33.520 ⇒ 00:34:33.860 Emily Giant: Yep.
292 00:34:34.199 ⇒ 00:34:35.400 Caio Velasco: Yeah, yeah.
293 00:34:35.409 ⇒ 00:34:36.159 Emily Giant: Totally.
294 00:34:36.429 ⇒ 00:34:37.799 Caio Velasco: Yeah, yeah, okay.
295 00:34:39.440 ⇒ 00:34:40.999 Emily Giant: I mean, there’s a yeah.
296 00:34:41.820 ⇒ 00:34:46.449 Emily Giant: Okay, what is this called strike through?
297 00:34:49.719 ⇒ 00:34:58.419 Emily Giant: Why am I saying 9? It’s a 10. Importance is 3 being wait. One is important type in source.
298 00:34:59.030 ⇒ 00:35:04.690 Emily Giant: Okay, and the note is after migration.
299 00:35:07.450 ⇒ 00:35:12.910 Emily Giant: There is no way to join these models.
300 00:35:13.450 ⇒ 00:35:18.830 Emily Giant: So if there is a 1st upgrade.
301 00:35:19.970 ⇒ 00:35:29.129 Emily Giant: There is no indication that a unit was canceled are not delivered.
302 00:35:32.040 ⇒ 00:35:39.169 Emily Giant: and like during a holiday, there will be 5,000 forced upgrades. So that’s like 5,000
303 00:35:40.000 ⇒ 00:35:41.369 Emily Giant: orders that are
304 00:35:41.690 ⇒ 00:35:47.190 Emily Giant: affecting top line revenue and saying we sent more or made more money than we actually did.
305 00:35:52.380 ⇒ 00:36:01.029 Emily Giant: Okay. So, lineage being. This goes to the dimension line. Item, they all eventually come to Oms suborders.
306 00:36:07.140 ⇒ 00:36:16.380 Emily Giant: I have a couple tickets, too. That might be good to look at just to like visualize
307 00:36:16.760 ⇒ 00:36:20.289 Emily Giant: what eventually happens with these like
308 00:36:20.450 ⇒ 00:36:25.350 Emily Giant: problematic line item tables. For some reason, by the time it gets to this, like
309 00:36:28.450 ⇒ 00:36:38.000 Emily Giant: the Oms items, table and Oms components. There will be like 4 lines for one item, and I.
310 00:36:38.800 ⇒ 00:36:45.259 Emily Giant: I don’t know why, but somewhere in the lineage, like a calculation fans out
311 00:36:45.380 ⇒ 00:36:51.600 Emily Giant: and is creating additional lines. So
312 00:36:53.650 ⇒ 00:36:58.780 Emily Giant: this is the mart for sales, data, tableau items. Xf.
313 00:36:58.960 ⇒ 00:37:14.940 Emily Giant: and what it is is like an aggregation of all of the problematic models. So this is where Oms suborders, lands. This is where the line items land. This is where the component data lands, and all of them
314 00:37:15.050 ⇒ 00:37:25.679 Emily Giant: are deprecated. So yeah, if you need to trace back
315 00:37:25.870 ⇒ 00:37:35.720 Emily Giant: the most problematic models, go to tableau items Xf, and every single one listed is an importance level one
316 00:37:36.120 ⇒ 00:37:44.010 Emily Giant: and deprecated 7 to 10, depending
317 00:37:44.390 ⇒ 00:37:50.729 Emily Giant: like, there are definitely elements of order, level build that are intact.
318 00:37:54.010 ⇒ 00:37:58.049 Emily Giant: But for the most part the other ones are all completely deprecated.
319 00:38:02.520 ⇒ 00:38:04.280 Emily Giant: And then there’s a lot of like
320 00:38:06.790 ⇒ 00:38:18.419 Emily Giant: in in this. This is like, of course, we’re revenue lands. But there’s additional calculations like through here. We already had an entire model upstream to determine.
321 00:38:18.730 ⇒ 00:38:23.459 Emily Giant: like overriding the cost of a subscription. And then it happens again.
322 00:38:23.570 ⇒ 00:38:33.630 Emily Giant: And there’s another model like an intermediate where it happens again. So you’re seeing all of this like
323 00:38:34.690 ⇒ 00:38:47.220 Emily Giant: continued logic being laid on top of an already problematic logic. So that’s really like affecting subscription forecasting sales. Just something to like.
324 00:38:47.390 ⇒ 00:38:52.810 Emily Giant: You’ll see it when you get into the models. But like calling out that we have
325 00:38:53.240 ⇒ 00:38:59.850 Emily Giant: laid 5 transformations on top of subscriptions. By the time they get to the mart.
326 00:39:02.500 ⇒ 00:39:06.180 Emily Giant: let me stop. I’ve been talking a lot. Do you have questions?
327 00:39:06.646 ⇒ 00:39:07.160 Emily Giant: Based on.
328 00:39:07.160 ⇒ 00:39:20.829 Caio Velasco: Like, that’s definitely a lot of information, and they’ll have to do it. And yeah, for sure. But I think that’s super helpful, especially because, since it’s recorded, I can always go back and also ask the agent to summarize something. So for sure, that would be helpful.
329 00:39:23.020 ⇒ 00:39:27.570 Caio Velasco: Yeah, maybe maybe what we could also do is
330 00:39:28.750 ⇒ 00:39:40.480 Caio Velasco: well given that we have the rating and the importance. How do you think we should approach like the most important ones and most unreliable ones to start with, and then maybe the business unit would be related to like
331 00:39:40.760 ⇒ 00:39:48.109 Caio Velasco: revenue inventory or something like that, so that at least we have like subsets to start working on or or learning.
332 00:39:48.775 ⇒ 00:39:50.840 Caio Velasco: Maybe we can also do that. But
333 00:39:51.000 ⇒ 00:39:56.059 Caio Velasco: yeah, that’s super important. And I see it’s like, it’s huge. It’s definitely huge.
334 00:39:56.840 ⇒ 00:39:57.440 Emily Giant: Yeah.
335 00:39:57.610 ⇒ 00:39:58.520 Caio Velasco: One other question.
336 00:39:58.520 ⇒ 00:40:06.400 Emily Giant: One is one person trying to fix it. I was like, I can’t. Either you get consultants or I walk. I can’t do this.
337 00:40:06.780 ⇒ 00:40:08.729 Caio Velasco: No, no, I can’t imagine, and.
338 00:40:08.730 ⇒ 00:40:09.100 Emily Giant: Yeah.
339 00:40:09.100 ⇒ 00:40:17.120 Caio Velasco: And also, for example, the did this, the definition for
340 00:40:17.520 ⇒ 00:40:20.599 Caio Velasco: how each thing works, for example, revenue, you know.
341 00:40:20.990 ⇒ 00:40:39.549 Caio Velasco: the simple thing is price times quantity. That’s it. So. But then, you know, it’s either gonna be something super logical on the side of the client or in the model itself. Right. But if you go in the model, it’s always extremely complicated to understand really what is happening. So for the 1st part.
342 00:40:40.190 ⇒ 00:40:48.020 Caio Velasco: I’m assuming that everything it’s either like in your head or someone else’s head, like everything that you’re explaining. There’s no
343 00:40:48.544 ⇒ 00:40:55.959 Caio Velasco: like knowledge base or something like that where you would like. Oh, how is revenue calculated? How is the use case? I don’t think there is right.
344 00:40:56.960 ⇒ 00:41:16.650 Emily Giant: No, I’ve tried. And it’s each business unit has a different need, so that that is expected. But we need a knowledge base, so that we know what we’re supposed to be building in the 1st place. And as you were recapping I’m thinking, like, would it be worth
345 00:41:16.900 ⇒ 00:41:31.249 Emily Giant: noting whether a model should exist at all. Because I do think that the way that some of these are built we shouldn’t. I don’t let the thinking around them is flawed from
346 00:41:31.960 ⇒ 00:41:39.920 Emily Giant: the genesis. From the beginning the thinking is flawed, and we need to like forget it and start over.
347 00:41:40.542 ⇒ 00:41:49.387 Emily Giant: I think that, like the staging line items, table versus the line item, split table, we need to forget about it entirely.
348 00:41:50.240 ⇒ 00:41:58.239 Emily Giant: those should not exist in the state that they are, because they will continue breaking. And
349 00:41:58.940 ⇒ 00:42:21.339 Emily Giant: there’s no sense in pretending we’re using the same software that we’re using now forever, because we won’t and we need to build it in a way that like could support any schema. And it was so specific to our schema that, like there’s no way that a seamless migration could have occurred.
350 00:42:21.680 ⇒ 00:42:27.340 Emily Giant: So I can leave notes to saying like.
351 00:42:27.520 ⇒ 00:42:33.470 Emily Giant: maybe we discuss getting rid of this entirely, time, gating it just
352 00:42:34.050 ⇒ 00:42:40.629 Emily Giant: like fashioning it after whatever we build, so that we can have that as like a legacy table that is referenced. But
353 00:42:41.020 ⇒ 00:42:42.460 Emily Giant: it’s just.
354 00:42:42.570 ⇒ 00:42:52.230 Emily Giant: It isn’t right, like none of it’s right. You’ll see, too, when you get into the component level
355 00:42:52.660 ⇒ 00:43:01.649 Emily Giant: items are. It’s so denormalized. One of the models takes better to show than to tell.
356 00:43:02.740 ⇒ 00:43:08.724 Emily Giant: Honestly, let’s not even talk about that one today unless you want to cause it’s so bad that
357 00:43:10.780 ⇒ 00:43:13.009 Emily Giant: think it’s it’s this one.
358 00:43:16.950 ⇒ 00:43:25.310 Emily Giant: But it’s no. The one I’m talking about is like 1 1 line of python, and all it does is like
359 00:43:25.930 ⇒ 00:43:33.510 Emily Giant: flip like 50 columns, 50 rows into columns, and it’s supposed to rank
360 00:43:33.630 ⇒ 00:43:49.470 Emily Giant: main product or the components of an order by like most expensive to least. And it just doesn’t do it. And so like. Sometimes you’re getting. It’s causing vases to be 20.
361 00:43:50.470 ⇒ 00:43:58.545 Emily Giant: I don’t know even really what it was supposed to do
362 00:44:03.170 ⇒ 00:44:03.960 Emily Giant: this.
363 00:44:08.180 ⇒ 00:44:15.460 Emily Giant: but I’ll send examples of all of these. I have a spreadsheet where, when I was teaching myself what was in each model like, I would
364 00:44:15.990 ⇒ 00:44:22.019 Emily Giant: take screenshots and like 10 lines and then add notes about what was broken.
365 00:44:22.710 ⇒ 00:44:24.830 Emily Giant: I can definitely share that with. You
366 00:44:29.100 ⇒ 00:44:39.750 Emily Giant: see, see if this spreadsheet still exists. This has some good examples of the line item issues.
367 00:44:49.980 ⇒ 00:44:56.780 Emily Giant: So this is, for example, staging line line items versus staging split line items.
368 00:44:57.551 ⇒ 00:45:04.070 Emily Giant: So the staging line items. This one is the individual elements.
369 00:45:04.690 ⇒ 00:45:07.239 Emily Giant: The staging split line items is the kit.
370 00:45:07.660 ⇒ 00:45:11.789 Emily Giant: This is the strikethrough portion, like, if you pull.
371 00:45:12.900 ⇒ 00:45:18.299 Emily Giant: If you pull in like an order number. So in all of these examples, I’m pulling the same order
372 00:45:19.290 ⇒ 00:45:21.189 Emily Giant: and then seeing what it does
373 00:45:21.935 ⇒ 00:45:24.809 Emily Giant: this order had deleted items.
374 00:45:25.070 ⇒ 00:45:30.880 Emily Giant: and they were marked in this table as deleted, but because
375 00:45:31.170 ⇒ 00:45:37.880 Emily Giant: the individual items are not pulling into this model where it marks
376 00:45:38.350 ⇒ 00:45:41.869 Emily Giant: that those elements have been deleted. Then
377 00:45:43.520 ⇒ 00:45:46.070 Emily Giant: They never get marked as deleted.
378 00:45:46.230 ⇒ 00:45:53.650 Emily Giant: So that’s just some notes here about how that went down. And then all of the like upstream.
379 00:45:53.800 ⇒ 00:45:56.169 Emily Giant: so this could be helpful. I’ll just.
380 00:45:56.170 ⇒ 00:45:57.279 Caio Velasco: Yeah, that’s correct.
381 00:45:57.280 ⇒ 00:46:01.409 Emily Giant: December and put it in a new spreadsheet. So you don’t have to look at all the junk
382 00:46:03.283 ⇒ 00:46:08.270 Emily Giant: but yeah, those are like the crux of the revenue issue.
383 00:46:10.710 ⇒ 00:46:13.598 Emily Giant: Revenues actually calculated inside.
384 00:46:14.560 ⇒ 00:46:16.749 Emily Giant: I believe it’s order level details.
385 00:46:18.180 ⇒ 00:46:26.240 Emily Giant: So not even like I don’t know. It just seems like hard to track down all of the different
386 00:46:26.830 ⇒ 00:46:30.609 Emily Giant: places and ways that revenue is calculated.
387 00:46:35.350 ⇒ 00:46:36.809 Emily Giant: okay, so not this.
388 00:46:48.190 ⇒ 00:46:59.690 Emily Giant: so any of these that start with oms are on that revenue flow. That’s 1 way. If, if, like me, you lose the thread of where revenue landed.
389 00:47:00.420 ⇒ 00:47:03.859 Emily Giant: you can just start with Oms suborders, and you’ll find it from there.
390 00:47:04.050 ⇒ 00:47:05.619 Emily Giant: So it must be nice.
391 00:47:15.740 ⇒ 00:47:21.100 Emily Giant: I think it’s this one, yes, and is revenue. So
392 00:47:22.630 ⇒ 00:47:26.070 Emily Giant: I did Update this recently, because revenue was getting like
393 00:47:26.720 ⇒ 00:47:40.939 Emily Giant: way blown up. We were showing a couple 1 million dollars over what we had made over mother’s day, because that split line item table was throwing duplicates. So right now it’s item total plus shipping cost.
394 00:47:41.300 ⇒ 00:47:45.029 Emily Giant: And then item, total being
395 00:47:45.590 ⇒ 00:47:48.049 Emily Giant: what was calculated into oms suborders.
396 00:47:48.570 ⇒ 00:48:00.079 Emily Giant: And then there’s some time gating about how to treat promo. So by the time you get here it’s like a relatively simple calculation that is time gated. But the issue is
397 00:48:00.820 ⇒ 00:48:03.605 Emily Giant: right here with the item total.
398 00:48:04.450 ⇒ 00:48:23.910 Emily Giant: I’d probably do well to change that to either or to subtotal, for now, because revenue is still broken and subtotal, is the most consistent revenue. Most consistent financial metric that we have. It seems to very much be one to one between our order management system and shopify.
399 00:48:26.540 ⇒ 00:48:31.160 Emily Giant: So I’ll think about that today. But yeah, okay.
400 00:48:31.160 ⇒ 00:48:33.761 Caio Velasco: You know, that’s that was great. That was great.
401 00:48:35.400 ⇒ 00:48:49.680 Caio Velasco: okay, cool. So yeah, that’s all about like the modeling part. And if we go back to the exactly the spreadsheet are we already able to look at it? And, for example, this would be more connected to the deprecation work of dashboard, for example, and also now rate.
402 00:48:49.680 ⇒ 00:48:50.020 Emily Giant: Thank you.
403 00:48:50.150 ⇒ 00:48:53.549 Caio Velasco: But I’m still looking at with the deprecation eyes.
404 00:48:55.100 ⇒ 00:49:01.500 Caio Velasco: will we be able to? Yeah, probably, if we have like rating important, I think we can kind of map
405 00:49:02.215 ⇒ 00:49:02.650 Caio Velasco: the
406 00:49:03.290 ⇒ 00:49:09.089 Caio Velasco: well, the best Dbt models to to their dashboards. That’s what I need to do next. And then.
407 00:49:09.090 ⇒ 00:49:09.780 Emily Giant: Yes.
408 00:49:09.780 ⇒ 00:49:12.710 Caio Velasco: Dive into exactly what you mentioned. Like all the
409 00:49:13.090 ⇒ 00:49:22.910 Caio Velasco: the modeling part. And then 1st thing in my mind is always the knowledge base we mentioned like, yeah, what is revenue? And I like to do top down. What is revenue
410 00:49:23.150 ⇒ 00:49:24.959 Caio Velasco: urban centers that work.
411 00:49:25.110 ⇒ 00:49:31.069 Caio Velasco: Then I’ll see if I if I re-watch this video. Start taking ready some notes
412 00:49:31.220 ⇒ 00:49:34.100 Caio Velasco: and then see if I can start. I don’t know.
413 00:49:34.340 ⇒ 00:49:40.740 Caio Velasco: building a little monster and then and then seeing like, if if each part of the revenue is, it’s well defined.
414 00:49:42.810 ⇒ 00:49:45.550 Caio Velasco: Yeah, yeah. Okay. I have some ideas.
415 00:49:46.160 ⇒ 00:49:55.310 Emily Giant: Okay, yeah. So with the like, what views are connected to what models?
416 00:49:57.740 ⇒ 00:50:03.029 Emily Giant: Is there a way to like export that? Or do you just like go one by one.
417 00:50:03.030 ⇒ 00:50:12.890 Caio Velasco: No, I’ll check. I’ll check. Because, yeah, before, as I mentioned, I didn’t even know how to use Looker, and it was able to get the content usage. So I’m assuming that
418 00:50:13.230 ⇒ 00:50:17.539 Caio Velasco: in my there there is a way, I assume there could be a way of
419 00:50:19.210 ⇒ 00:50:21.410 Caio Velasco: just getting the name of the models
420 00:50:21.510 ⇒ 00:50:28.240 Caio Velasco: somehow filtering them. There is maybe, like a a tag for that. I’ll I’ll check.
421 00:50:28.880 ⇒ 00:50:38.019 Emily Giant: Yeah, I’m happy to help but I’m not a manual girl like when I I had to.
422 00:50:38.020 ⇒ 00:50:41.786 Caio Velasco: Yeah way of automate that. Otherwise I already don’t like my day.
423 00:50:42.100 ⇒ 00:50:45.862 Emily Giant: Yeah, exactly. I’m like, Oh, this seems really boring.
424 00:50:46.280 ⇒ 00:50:59.349 Caio Velasco: Yeah, yeah, I get it. Yeah, it was the same with ingestion. First, st things like, Where are, where’s the Api keys? And then I saw that for stitch. I think there were no Api keys, and we have to upgrade some. It’s like, Oh, my God!
425 00:50:59.480 ⇒ 00:51:04.530 Caio Velasco: So then I was like, No, but maybe a scraper would work. So then it worked otherwise.
426 00:51:04.530 ⇒ 00:51:07.309 Emily Giant: Oh, my, gosh, that’s amazing. Yeah.
427 00:51:07.310 ⇒ 00:51:08.970 Caio Velasco: Oh, that would be! Oh, no!
428 00:51:09.200 ⇒ 00:51:09.810 Caio Velasco: But.
429 00:51:09.810 ⇒ 00:51:20.419 Emily Giant: Yeah. When when had me like documenting the every field in Dbt, this was like, at the beginning of the engagement a couple of months ago
430 00:51:20.530 ⇒ 00:51:36.319 Emily Giant: I spent all of my time just researching like scripts to scrape our Dbt. Instead of actually doing it. I probably could have done it and finished it in the amount of time I spent researching how to scrape it because I was like, I cannot do this manually. It will break me.
431 00:51:36.893 ⇒ 00:51:48.209 Caio Velasco: No, that, of course, in the beginning you always put this, maybe even the same amount of hours, or even more. Then, if you will do it manually, but the next one would be where you start to see the economies of scale.
432 00:51:49.020 ⇒ 00:51:58.769 Emily Giant: Exactly. I am of that mindset. Okay, so I think that like, we’re probably, is there anything else you want to do? Today? Because
433 00:51:59.640 ⇒ 00:52:13.199 Emily Giant: map. But just shoot me a DM. If you have any questions. I’m glad you recorded that could you send it to me so I can see how I can better tell people about our like.
434 00:52:13.510 ⇒ 00:52:34.240 Emily Giant: I need to make notes for myself on doing this chat because it’s not going to be the last time. That would be really helpful for me to just look at how I did it, and then like, better organize it, and have like a template in the future, for how to like, make it clearer and easier, and add examples and stuff.
435 00:52:34.880 ⇒ 00:52:37.170 Caio Velasco: Sure. Sure I’ll I’ll send you for sure.
436 00:52:38.090 ⇒ 00:52:48.950 Caio Velasco: Yeah. And then just make sure that that tab the Dbt models that it’s Updated with the 2 columns and the business business unit for the most important ones at least.
437 00:52:49.160 ⇒ 00:52:53.350 Caio Velasco: and the most important one being the combination of importance and rating
438 00:52:53.974 ⇒ 00:53:00.829 Caio Velasco: and then, if that’s done, then I think we are ready to go and map the dashboards, and then I’ll go into the modeling part.
439 00:53:01.610 ⇒ 00:53:09.468 Emily Giant: Alright cool. Well, that’s that’s what I have left to do today. So if I don’t do it, then I’m like playing hooky or something going wrong.
440 00:53:10.430 ⇒ 00:53:10.970 Caio Velasco: Correct.
441 00:53:10.970 ⇒ 00:53:13.960 Emily Giant: Cool. I will see you at the stand up. Thank you so much.
442 00:53:13.960 ⇒ 00:53:14.849 Caio Velasco: You’re welcome.
443 00:53:14.970 ⇒ 00:53:16.239 Caio Velasco: Thank you. Appreciate.