Meeting Title: High-Intensity Shopping Days Playbook Date: 2025-09-10 Meeting participants: Demilade Agboola, Jake Nathan
WEBVTT
1 00:00:24.570 ⇒ 00:00:25.890 Jake Nathan: Hey, Damon Lotta.
2 00:00:26.640 ⇒ 00:00:27.950 Demilade Agboola: How’d you call you?
3 00:00:28.570 ⇒ 00:00:29.780 Jake Nathan: Good, how are you doing?
4 00:00:30.200 ⇒ 00:00:34.850 Demilade Agboola: I’m pretty alright. I might need to…
5 00:00:36.190 ⇒ 00:00:41.130 Demilade Agboola: I mean, I need to make some coffee in, like, 10 minutes’ time, but… I’m fine.
6 00:00:42.260 ⇒ 00:00:49.019 Jake Nathan: Cool, do you, I mean, if you want to, you wanna go ahead and make it now, and I can… I can wait till you come back?
7 00:00:49.500 ⇒ 00:00:50.850 Demilade Agboola: Alright, sounds good.
8 00:00:51.060 ⇒ 00:00:51.650 Jake Nathan: Okay.
9 00:00:52.860 ⇒ 00:00:53.739 Demilade Agboola: I’ll be right back.
10 00:00:54.210 ⇒ 00:00:55.420 Jake Nathan: Okay, sounds good.
11 00:04:27.120 ⇒ 00:04:28.659 Demilade Agboola: Hi, Jake, so I’m back.
12 00:04:30.750 ⇒ 00:04:34.860 Jake Nathan: Awesome, awesome, sounds good. Are you a big coffee drinker?
13 00:04:35.220 ⇒ 00:04:45.839 Demilade Agboola: Yes, yes, unfortunately. And, like, I struggle to start my days without coffee, and usually I need, like, a midday coffee, because right here, it’s, like, 4…
14 00:04:46.310 ⇒ 00:04:48.010 Demilade Agboola: 4.36 PM.
15 00:04:48.370 ⇒ 00:04:48.780 Jake Nathan: Gotcha.
16 00:04:48.780 ⇒ 00:04:52.380 Demilade Agboola: So I need that second wave to get through the regular day.
17 00:04:52.770 ⇒ 00:05:06.169 Jake Nathan: Oh, I totally feel you. I… coffee is one of those things, like, if… if I couldn’t have coffee, I would be… I would be a very upset person. I would… like, I… I need it. So, that’s awesome. Yeah, where… where are you based out of?
18 00:05:06.950 ⇒ 00:05:15.719 Demilade Agboola: So I’m based out of Malta, but I’m currently, because I’m Nigerian, I’m currently visiting Nigeria, my best friend, one of my best friends gets married.
19 00:05:15.900 ⇒ 00:05:18.570 Demilade Agboola: So I’m his best man, and so…
20 00:05:18.570 ⇒ 00:05:19.410 Jake Nathan: Oh, nice!
21 00:05:19.760 ⇒ 00:05:20.440 Demilade Agboola: I’m here.
22 00:05:20.440 ⇒ 00:05:23.180 Jake Nathan: That would be fun. Are you giving a speech for the wedding?
23 00:05:23.720 ⇒ 00:05:25.799 Demilade Agboola: Yeah,
24 00:05:26.520 ⇒ 00:05:36.420 Demilade Agboola: I haven’t… I haven’t… I have what I… I have my template in my head, say, but I need to put it down, and just, like, you know, properly…
25 00:05:36.710 ⇒ 00:05:39.550 Demilade Agboola: Put down the speech, but there’s a lot of running around to do.
26 00:05:40.690 ⇒ 00:05:41.080 Jake Nathan: Absolutely.
27 00:05:41.310 ⇒ 00:05:44.420 Demilade Agboola: what I have been doing a lot recently.
28 00:05:44.990 ⇒ 00:05:52.470 Demilade Agboola: It’s tomorrow and Friday, it’s a two-day… like, a lot of Nigerian weddings tend to be two-day events.
29 00:05:52.760 ⇒ 00:05:54.110 Demilade Agboola: So it’s the city of that.
30 00:05:54.230 ⇒ 00:05:55.100 Demilade Agboola: Yeah.
31 00:05:55.610 ⇒ 00:06:15.099 Jake Nathan: That’s awesome, that’s… that’s exciting. Yeah, I was, last weekend, I was… I’d never been a best man before, but yeah, I was a best man in one of my friends’ weddings, and yeah, it’s… it’s a lot of fun, but it also is, yeah, like you’re saying, a lot of running around, you’re kind of like the… like, everyone’s kind of… like, if they have anything they need, they’re just like, hey, can you do this for me? So it’s a… it’s a…
32 00:06:15.100 ⇒ 00:06:15.580 Demilade Agboola: Yay.
33 00:06:15.580 ⇒ 00:06:31.479 Jake Nathan: a little bit, I guess, stressful, but, like, but then it’s very satisfying, like, once you kind of watch the whole event unfold, and obviously you’re… it’s fun to be with the friends, so, that’s cool, man. That’s… that’s really cool. I hope… I hope it’s a fun…
34 00:06:31.740 ⇒ 00:06:33.770 Jake Nathan: fun event.
35 00:06:34.180 ⇒ 00:06:35.690 Demilade Agboola: Should be funky.
36 00:06:36.060 ⇒ 00:06:41.920 Jake Nathan: Yeah, this, and the speech, too. I… I know what you’re saying, I feel like I had it in my head, but…
37 00:06:42.300 ⇒ 00:06:47.550 Jake Nathan: it wasn’t until I really, like, wrote it down, and then that made it a lot easier, because…
38 00:06:47.790 ⇒ 00:06:58.779 Jake Nathan: I don’t know, sometimes… sometimes I get nervous before speaking, sometimes I don’t, but what about you? Are you, like… do you get nervous at all before you speak, or can you just, you know, go up there and it’s all good?
39 00:06:59.410 ⇒ 00:07:04.440 Demilade Agboola: I mean, I’m not… I’m not a… I won’t necessarily be natural.
40 00:07:04.590 ⇒ 00:07:08.079 Demilade Agboola: But, like, back in, like, high school.
41 00:07:08.200 ⇒ 00:07:14.540 Demilade Agboola: Like, I was, like, we have, like, Perfect and stuff, so I was, like, the head boy of the school.
42 00:07:14.950 ⇒ 00:07:19.610 Demilade Agboola: So, I had to talk quite a bit to the school.
43 00:07:20.120 ⇒ 00:07:28.749 Demilade Agboola: So that kind of prepared me for, like, just some level of public speaking. So I’m not the worst person in the world, but, I mean, I’ve actually thought of taking
44 00:07:29.070 ⇒ 00:07:36.840 Demilade Agboola: like, public speaking, speech classes, just generally speaking. I feel like there’s… there’s a level of expertise
45 00:07:37.010 ⇒ 00:07:42.040 Demilade Agboola: That, you know, just changes a lot of the dynamics that you can experience in everyday life.
46 00:07:42.400 ⇒ 00:07:47.200 Demilade Agboola: But yeah, I mean, I have no issues with that, I don’t really get nervous.
47 00:07:47.870 ⇒ 00:07:52.559 Demilade Agboola: I would just kind of get… go on there and just get things done.
48 00:07:53.060 ⇒ 00:07:59.149 Demilade Agboola: And, you know, this… this is… this is… it’s a… for a Nigerian wedding, this is small, but we’re still expecting, like.
49 00:07:59.290 ⇒ 00:08:00.860 Demilade Agboola: 200-ish people?
50 00:08:00.990 ⇒ 00:08:02.160 Demilade Agboola: A little bit over.
51 00:08:02.980 ⇒ 00:08:04.519 Demilade Agboola: But, I mean.
52 00:08:04.520 ⇒ 00:08:05.650 Jake Nathan: That’s a solid runner.
53 00:08:06.170 ⇒ 00:08:07.070 Demilade Agboola: Yeah.
54 00:08:07.390 ⇒ 00:08:13.950 Demilade Agboola: So it’s not the large… like, I know, like, Nigerians have really large weddings. My sister’s wedding was, like, a thousand plus, like, a thousand.
55 00:08:13.950 ⇒ 00:08:16.660 Jake Nathan: Oh my gosh, that’s crazy.
56 00:08:16.920 ⇒ 00:08:22.610 Demilade Agboola: Yeah, so if I had to… if I was, like, the best man at that sort of wedding, I think I might be a little bit more nervous.
57 00:08:23.660 ⇒ 00:08:38.590 Jake Nathan: I mean, yeah, a thousand people, that’s like a… yeah, that’s… I mean, that sounds… could be really fun, but also, yeah, that’s… I mean, that’s a lot of people. A lot, like, a lot of food, a lot of, like, everything is… like, a thousand people makes it, I’m sure, like, a lot at once.
58 00:08:38.940 ⇒ 00:08:44.219 Demilade Agboola: Yeah, and insurance like big weddings, too. Some people have, like, 3,000, 4,000 people at their weddings.
59 00:08:44.220 ⇒ 00:08:46.199 Jake Nathan: Oh my goodness.
60 00:08:46.480 ⇒ 00:08:51.219 Jake Nathan: That is… I mean, that honestly sounds pretty… pretty fun.
61 00:08:51.480 ⇒ 00:09:07.909 Jake Nathan: Well, they… well, I don’t want to take up too much of your time, because if you’ve got best man duties, I want to let you go do that. But, yeah, I know we were talking a little bit with Utom and Slack, and I think, you know, what we were going to go with is basically a… a…
62 00:09:08.470 ⇒ 00:09:24.580 Jake Nathan: lesson or playbook around, you know, these high-intensity shopping days. So, you know, Black Friday seems to be the next one that’s gonna come up, and then the cool thing is, like, if you prepare for Black Friday, you know, like you were telling me, Mother’s Day.
63 00:09:24.610 ⇒ 00:09:36.360 Jake Nathan: Father’s Day, Valentine’s Day, like, there’s just so many… it seems like that is, like, a big part of e-com, is, like, preparing for those big days. So, yeah, if you,
64 00:09:37.050 ⇒ 00:09:43.719 Jake Nathan: I think that’s… that’s the one we can go with. And so, I guess, like, the first question I have for you is.
65 00:09:43.960 ⇒ 00:09:48.290 Jake Nathan: Like, why is a high-intensity day like this
66 00:09:48.910 ⇒ 00:09:55.310 Jake Nathan: something to even prepare for? Like, why is it unique or challenging from a data perspective, if that makes sense?
67 00:09:56.670 ⇒ 00:10:06.699 Demilade Agboola: So it’s important to prepare for such days because the volume of the data that is coming in increases, in some cases, exponentially.
68 00:10:07.540 ⇒ 00:10:15.290 Demilade Agboola: And that increase… If not well catered to, can cause…
69 00:10:15.690 ⇒ 00:10:19.149 Demilade Agboola: Can cause, slower data refreshes.
70 00:10:19.280 ⇒ 00:10:24.680 Demilade Agboola: And just make it harder, like, causing more latency, basically, in seeing your data.
71 00:10:24.960 ⇒ 00:10:31.040 Demilade Agboola: So that’s point number one. Point number two is, on such days, due to the
72 00:10:31.820 ⇒ 00:10:38.580 Demilade Agboola: fast pace of what’s happening. It’s very important to be as on top of your data as possible.
73 00:10:38.660 ⇒ 00:10:52.090 Demilade Agboola: So, on such days, if you only used to look at morning reports, like, oh, what happened the previous day, well, if you come the day after and just look at the report, then you’re late. You know, there’s nothing you could have… there’s nothing you could do.
74 00:10:52.190 ⇒ 00:11:01.619 Demilade Agboola: or, you know, made any… you can’t make any changes in the past anymore, right? So trying to be on top of things on such days, especially, is very important. You want to be…
75 00:11:02.200 ⇒ 00:11:05.019 Demilade Agboola: Right there, in that day.
76 00:11:05.280 ⇒ 00:11:24.079 Demilade Agboola: being able to say, hey, things have… things have fallen off over the past hour. Is our website down, or is something… like, is there something… is there an issue paying with a particular page, or something? Right? You want to be as on things as possible, so that you don’t lose the traction that you have, or the volume of transactions that you have on that day.
77 00:11:25.300 ⇒ 00:11:27.089 Jake Nathan: Gotcha. That makes sense.
78 00:11:27.700 ⇒ 00:11:36.809 Jake Nathan: That totally makes sense. And then, we’ll dive into each point, but did you have, another point, or the… you’ve had two so far that are great.
79 00:11:37.630 ⇒ 00:11:42.070 Demilade Agboola: Those were the two main points I had about why you want to be on top of that.
80 00:11:42.760 ⇒ 00:11:49.210 Jake Nathan: Gotcha. Okay, so yeah, so back to number one, like, point number running around, like.
81 00:11:49.370 ⇒ 00:11:52.320 Jake Nathan: Because there’s a lot more traffic.
82 00:11:52.880 ⇒ 00:11:56.749 Jake Nathan: this could increase the latency of your data. Like, why…
83 00:11:56.990 ⇒ 00:12:06.349 Jake Nathan: why is that a problem? Like, what… what’s kind of the consequence of, you know, if there’s more latency, what, why could that be an issue?
84 00:12:07.650 ⇒ 00:12:13.219 Demilade Agboola: So if you have low latency, or slower, or a larger latency with your data.
85 00:12:13.370 ⇒ 00:12:18.499 Demilade Agboola: Effectively, what happens is you’re only going to be able to make decisions when that data is ready.
86 00:12:18.670 ⇒ 00:12:30.160 Demilade Agboola: And if you are only making decisions at a slower cadence to what you ideally would want to, well, it means you’re constantly chasing that day.
87 00:12:31.840 ⇒ 00:12:41.260 Demilade Agboola: I’m just… so we… like, just so we’re clear, even though it’s technically a high-intensity day, in reality, it tends to be a high intensity week.
88 00:12:41.700 ⇒ 00:12:51.999 Demilade Agboola: Right? Because people don’t just place, especially for, like, Valentine’s Day or something, people don’t just place on that day. People are placing in, like, the 7 to 10 days leading up to Valentine’s Day.
89 00:12:52.320 ⇒ 00:13:04.839 Demilade Agboola: let’s just say sounds 14 days leading up Valentine’s Day. The orders start to rack up and rack up and rack up. And so now, you really want to be on top of things as quickly as possible. You want to have your, you know.
90 00:13:05.030 ⇒ 00:13:17.629 Demilade Agboola: read your May Day dashboards ready, you want to be able to know when orders are dropping, your failure rates on transactions, for instance, you want to have it as close to real-time visibility basis, something like that.
91 00:13:17.890 ⇒ 00:13:33.419 Demilade Agboola: you want to see, what are the main orders that… what are people ordering? You want to see if you’re out of stock. As much as possible, you want to see these things, and you want to see it as close to real time as possible, so you can make those decisions, you know.
92 00:13:33.640 ⇒ 00:13:35.040 Demilade Agboola: As quickly as possible.
93 00:13:36.240 ⇒ 00:13:40.089 Jake Nathan: Yeah, that makes sense, and to that point, like you were saying.
94 00:13:40.200 ⇒ 00:13:55.729 Jake Nathan: how… so, now that we kind of know it’s a problem, or someone reading this would know, hey, this is a problem, I need to start preparing, and I like what you said about, you know, Valentine’s Day, you can’t prepare on February 13th for February 14th, you really need to…
95 00:13:55.930 ⇒ 00:13:59.620 Jake Nathan: I mean, I guess one question is, when do you think…
96 00:13:59.730 ⇒ 00:14:09.980 Jake Nathan: Like, is it a month in advance? Is it 3 months in advance? Like, when do you start preparing for a high-intensity shopping day? Like, how much time should you give yourself?
97 00:14:11.350 ⇒ 00:14:15.949 Demilade Agboola: So this really depends on the scale that we’re talking about.
98 00:14:16.380 ⇒ 00:14:21.570 Demilade Agboola: And usually, if you consult with us, we will help you figure that out.
99 00:14:21.710 ⇒ 00:14:27.959 Demilade Agboola: But usually, you want to be able to get things up and running, like your system’s up and running.
100 00:14:28.130 ⇒ 00:14:30.550 Demilade Agboola: At least 2 weeks before.
101 00:14:30.880 ⇒ 00:14:36.370 Jake Nathan: So you can start, like, as the volume is ramping up, if there are any breakages, you can kind of see it.
102 00:14:36.590 ⇒ 00:14:38.619 Demilade Agboola: in a non…
103 00:14:39.040 ⇒ 00:14:46.769 Demilade Agboola: critical states. You can kind of see, you know, if it’s the end of January, and you see orders are beginning to ramp up, for instance.
104 00:14:46.910 ⇒ 00:14:53.389 Demilade Agboola: If things are breaking, you can kind of adjust them, look at your dashboard, see if things are, like, where you want them to be.
105 00:14:53.880 ⇒ 00:15:00.419 Demilade Agboola: And you can make those tweaks. Obviously, a breakage on, you know, the 2nd of February is…
106 00:15:01.330 ⇒ 00:15:20.129 Demilade Agboola: less harmful or less, like, problematic than a breakage on the 13th of February. So, as much as possible, you want to be ahead of the curve. So yes, ideally, you’d… worst case scenario, want to be done 2 weeks prior to the D-Day, and you would ideally want to start working on this process at least 2 months before.
107 00:15:20.760 ⇒ 00:15:22.020 Demilade Agboola: You wanna stop? Okay.
108 00:15:22.260 ⇒ 00:15:24.589 Demilade Agboola: You want to start getting things in place.
109 00:15:26.540 ⇒ 00:15:39.149 Jake Nathan: Yeah, I like that. So then, now it’s 2 months before, you mentioned a little bit, but can you go into more detail about, like, let’s say, hey, I know I want to be prepared for this.
110 00:15:39.220 ⇒ 00:15:45.310 Jake Nathan: I want to be prepared. I have 2 months. I’m ready to start preparing. Like, what…
111 00:15:45.350 ⇒ 00:15:59.699 Jake Nathan: what specific actions should I start taking? Like, you mentioned a dashboard, you mentioned a few things, but, like, if you were in this situation, like, what are the, like, next steps to preparing?
112 00:16:01.010 ⇒ 00:16:03.989 Demilade Agboola: Number one step would always be defining
113 00:16:04.760 ⇒ 00:16:09.969 Demilade Agboola: The important things that we’re looking for, like, the important metrics we’re looking for on that day.
114 00:16:10.330 ⇒ 00:16:18.370 Demilade Agboola: So, for instance, an important metric can be stock. What are the values, like, how much stock do we have of each?
115 00:16:18.590 ⇒ 00:16:20.979 Demilade Agboola: Commodity that we have, right?
116 00:16:20.980 ⇒ 00:16:21.990 Jake Nathan: Number two.
117 00:16:21.990 ⇒ 00:16:29.169 Demilade Agboola: could be transaction failure rates. So obviously, if transactions are failing, you’re losing money. You want to be ahead of it as quickly as possible.
118 00:16:29.460 ⇒ 00:16:38.910 Demilade Agboola: And given the fact that it’s a high-intensity day, or, like, a high-intensity period, people are not literally waiting. Like, if I’m trying to buy flowers on Valentine’s Day.
119 00:16:39.030 ⇒ 00:16:48.789 Demilade Agboola: and it’s failing, I’m gonna go to another, like, vendor and try and get flowers, basically. Or if I’m trying to get something, like, if I’m trying to get, like, headphones on Black Friday.
120 00:16:49.280 ⇒ 00:17:00.700 Demilade Agboola: my transactions are failing, I’m just going to go to another, you know, vendor. I’m just trying to find another spot where I can do that. So, obviously, on such days, you want your, like, you might want visibility into certain metrics.
121 00:17:00.830 ⇒ 00:17:05.139 Demilade Agboola: these… Are sort of similar, but also vary across different industries.
122 00:17:05.280 ⇒ 00:17:20.670 Demilade Agboola: So the important thing is, we need to get a baseline of those metrics. So your everyday dashboard’s great, but, like, they don’t answer these questions. So once you identify the important metrics, the next step is what models do we have, or what processes do we have.
123 00:17:20.760 ⇒ 00:17:29.819 Demilade Agboola: that build out these metrics, and you need to isolate them. So once you isolate them, that allows you to be able to say, hey, these are our critical models.
124 00:17:30.230 ⇒ 00:17:36.460 Demilade Agboola: You need to have… A critical model line, and that lets you know, hey.
125 00:17:37.520 ⇒ 00:17:45.279 Demilade Agboola: everything might fail, right? But these models, we need to be on top of it. So, at that point, once you have that defined.
126 00:17:45.570 ⇒ 00:17:56.169 Demilade Agboola: The next thing you need to do is, number one, You need to… Optimize those models.
127 00:17:56.810 ⇒ 00:18:01.390 Demilade Agboola: So, right now, when I’m thinking optimization, I’m thinking…
128 00:18:01.660 ⇒ 00:18:05.769 Demilade Agboola: You need to ensure, like, the models are as incrementally…
129 00:18:06.170 ⇒ 00:18:11.780 Demilade Agboola: smart as possible. So incremental models just basically means instead of the models building from scratch.
130 00:18:12.180 ⇒ 00:18:16.100 Demilade Agboola: You would have them viewed from when they last… where they last stopped.
131 00:18:16.520 ⇒ 00:18:17.450 Demilade Agboola: Right?
132 00:18:17.890 ⇒ 00:18:20.960 Demilade Agboola: You’d also do… think about the…
133 00:18:21.390 ⇒ 00:18:24.529 Demilade Agboola: Like, what joins are you doing?
134 00:18:24.790 ⇒ 00:18:30.070 Demilade Agboola: Is there a smarter way to do it? Do I need to sort this thing faster? So when I mean, like, sort.
135 00:18:32.220 ⇒ 00:18:39.910 Demilade Agboola: when dbt runs and builds tables, it builds them in a… in a sorting order, so think of a sort like a dictionary.
136 00:18:39.910 ⇒ 00:18:53.889 Demilade Agboola: Like, if I scrambled a dictionary and told you to find the word giant, it will take you forever to find giant. However, if you open the dictionary, you would know, go to G, then go to I, then go to A, and then you would probably see giant around there.
137 00:18:55.190 ⇒ 00:19:13.709 Demilade Agboola: Sorting is something similar to that. It’s creating an easy-to-find index of data within your system, and that allows, like, your dashboards to find things faster, it makes it very much easier to be able to retrieve data and, you know, visualize that data, and that’s helpful.
138 00:19:14.080 ⇒ 00:19:20.060 Demilade Agboola: And so, yeah, things like that. You need to optimize your data, because that’s very important.
139 00:19:20.350 ⇒ 00:19:22.380 Demilade Agboola: So that you don’t have
140 00:19:23.230 ⇒ 00:19:25.890 Demilade Agboola: You reduce latency as much as possible.
141 00:19:26.880 ⇒ 00:19:34.030 Demilade Agboola: And then, once we have that, you need to also do… An infrastructure,
142 00:19:35.400 ⇒ 00:19:45.519 Demilade Agboola: an infrastructure breakdown. I think this should actually come before all the part of, like, analyzing your models, but immediately after you’re done, like, defining your metrics, you need to analyze your infrastructure.
143 00:19:45.940 ⇒ 00:19:52.490 Demilade Agboola: You want to be able to see your data quickly, cool. But do you have the tools that allow you to do that?
144 00:19:52.600 ⇒ 00:20:07.240 Demilade Agboola: Because sometimes it’s easy to say, yeah, I want my data quickly, but if you’re not using, you know, warehouses that are built for faster processing times, or like, you know, lower latency times, if you’re not using
145 00:20:07.630 ⇒ 00:20:10.949 Demilade Agboola: Dashboards, that means that you…
146 00:20:12.060 ⇒ 00:20:14.450 Demilade Agboola: The refreshes can keep up with such.
147 00:20:14.630 ⇒ 00:20:25.190 Demilade Agboola: you’re just gonna… you’re just basically, you know, wasting time. So you need to do an infrastructure audit, and then you can start thinking about, like, building your models to…
148 00:20:25.690 ⇒ 00:20:28.980 Demilade Agboola: Incrementally, like, to work optimally.
149 00:20:31.300 ⇒ 00:20:34.570 Jake Nathan: Gotcha. Yeah, so, if… or go ahead.
150 00:20:35.200 ⇒ 00:20:41.609 Demilade Agboola: Alright, so, yeah, once you then have your models working optimally, at this point.
151 00:20:41.830 ⇒ 00:20:51.819 Demilade Agboola: You need to set up schedule runs and schedule jobs where you isolate these models and this, your newly defined,
152 00:20:53.600 ⇒ 00:21:12.660 Demilade Agboola: optimized models, basically. The idea is, you need to create a path that is different from everything, like a run that is different from everything. That way, you can have your monitoring and your tests set up on just that run, so if there’s any issues, you can flag them really quickly, because it’s scheduled away from
153 00:21:12.930 ⇒ 00:21:14.369 Demilade Agboola: All the other stuff.
154 00:21:14.680 ⇒ 00:21:28.709 Demilade Agboola: So the other, you know, business KPIs that matter on a normal day, like, that’s great, they should still run, but you want to be able to have visibility on just one… like, you want to take them apart, you don’t want to have everything you need.
155 00:21:28.890 ⇒ 00:21:34.070 Demilade Agboola: in one space. You want to separately build that out, such that if that breaks.
156 00:21:34.230 ⇒ 00:21:47.690 Demilade Agboola: all… all… all hell breaks loose. You’re able to, like, go straight there and not start separating things. You know, these are the critical models we’ve built, these are what’s power in the dashboard, and we’re alert to anything that’s happening to them immediately.
157 00:21:49.950 ⇒ 00:21:54.809 Jake Nathan: That makes sense. So yeah, you want to have your, like, carve out a space just to…
158 00:21:54.950 ⇒ 00:22:01.500 Jake Nathan: To focus on the critical models, so that way… and that’s probably… Like, like you said, April.
159 00:22:02.020 ⇒ 00:22:02.880 Jake Nathan: Go ahead.
160 00:22:03.410 ⇒ 00:22:05.570 Demilade Agboola: Yeah, you want to carve out,
161 00:22:05.900 ⇒ 00:22:10.870 Demilade Agboola: Jobs specifically for them, you know, jobs or environments specifically for them.
162 00:22:11.090 ⇒ 00:22:16.720 Demilade Agboola: That way, it’s easy to identify any problems that are going to affect that model.
163 00:22:19.200 ⇒ 00:22:34.670 Demilade Agboola: alone, like, affect that, like, dashboard alone, or whatever dashboard you’re using alone. So it could be your Black Friday dash… your Black Friday jobs, it could be your Valentine’s Day jobs, it could be your Mother’s Day jobs, your Father’s Day jobs, but those jobs need to be explicit.
164 00:22:35.370 ⇒ 00:22:43.960 Demilade Agboola: very, like, no bloatware, you’re not having anything that is unnecessary to the end goal. Everything needs to be very clearly selected to…
165 00:22:44.170 ⇒ 00:22:53.040 Demilade Agboola: reflect on the end goal of business… like, high volume intense. High volume… critical models.
166 00:22:53.520 ⇒ 00:22:58.810 Demilade Agboola: Once you have those separated, yeah, then you can have them in your own infrastructure where
167 00:22:59.140 ⇒ 00:23:02.800 Demilade Agboola: In their own job, where you can just kind of monitor and ensure everything is going well.
168 00:23:04.080 ⇒ 00:23:11.450 Jake Nathan: Gotcha. Okay, that… that totally makes sense. And, yeah, is that,
169 00:23:11.630 ⇒ 00:23:29.499 Jake Nathan: before I ask you any more follow-up questions, so… so far you’ve said, like, first define the metrics, then analyze the infrastructure, then you figure out what models, then you optimize those models, and then you separately… or you… yeah, you separate them, schedules, like, separate runs.
170 00:23:29.630 ⇒ 00:23:35.630 Jake Nathan: In order to… So yeah, are there any additional steps?
171 00:23:36.090 ⇒ 00:23:42.900 Demilade Agboola: The final step would probably be some stuff around, like, cadence. How often do you want those, like, critical day models to run?
172 00:23:44.790 ⇒ 00:23:47.080 Jake Nathan: That needs to be defined across the team.
173 00:23:47.420 ⇒ 00:23:57.590 Demilade Agboola: And it needs to be figured out based off Business needs, and… Like… Yeah, basically business needs.
174 00:23:57.800 ⇒ 00:24:06.600 Demilade Agboola: So, it’s possible, for instance, you might be seeing the data every, like, 15 minutes as it happens, but are you really doing stuff every 15 minutes, or is every hour fine?
175 00:24:06.890 ⇒ 00:24:13.379 Demilade Agboola: Or you might be every 30 minutes, but, like, the idea is you need to be able to find out what KDAS works for these… this process.
176 00:24:13.830 ⇒ 00:24:22.369 Demilade Agboola: Bear in mind, compute is always the more expensive thing about data.
177 00:24:23.110 ⇒ 00:24:31.929 Demilade Agboola: storage is cheaper, compute is where, like, you know, cost can start to rise. So you want to be smart about how often you run these things.
178 00:24:32.310 ⇒ 00:24:40.549 Demilade Agboola: So you need that perfect balance between business output and value, as well as the actual cost.
179 00:24:42.960 ⇒ 00:24:49.210 Jake Nathan: That… that totally makes sense. If there’s, like, it varies so much between each company, but…
180 00:24:49.450 ⇒ 00:25:02.750 Jake Nathan: Is there, like, a standard, hey, if you… if you were just going into any company, you didn’t even know what they were selling, what would you… would it be 15 minutes, 30 minutes, an hour? Like, what’s just kind of, like, a standard cadence?
181 00:25:04.120 ⇒ 00:25:09.929 Demilade Agboola: It varies too wildly to just categorically say.
182 00:25:12.120 ⇒ 00:25:20.560 Demilade Agboola: However, I will say for a decent number of companies, unless you’re talking, like, maybe huge volume.
183 00:25:20.780 ⇒ 00:25:22.940 Demilade Agboola: Like, then Amazon and stuff.
184 00:25:23.230 ⇒ 00:25:34.990 Demilade Agboola: A number of companies between, like, 45 minutes intervals will probably give you enough data to be able to know what’s going on, whilst still being able to make
185 00:25:35.210 ⇒ 00:25:40.010 Demilade Agboola: whatever changes you need. You know, obviously, again, case by case.
186 00:25:40.150 ⇒ 00:25:43.170 Demilade Agboola: So some people definitely need way more.
187 00:25:43.500 ⇒ 00:25:46.189 Demilade Agboola: time. Like, way less time, sorry.
188 00:25:46.350 ⇒ 00:25:53.089 Demilade Agboola: And some people might need, you know, a bit more time. They’re fine with a bit more time. Like, if it’s an hour and still 45 minutes, they’re fine.
189 00:25:53.740 ⇒ 00:25:59.389 Demilade Agboola: But for some people, you’re thinking, like, literally, how can we get down to, like, 10 minutes, or how can we get down to, like, 5 minutes?
190 00:25:59.510 ⇒ 00:26:07.079 Demilade Agboola: Because you know that, like, anything that goes bad needs to be seen by the business stakeholders immediately.
191 00:26:08.840 ⇒ 00:26:10.619 Jake Nathan: That makes sense.
192 00:26:10.830 ⇒ 00:26:17.679 Jake Nathan: Gotcha. So… Once you’re able to figure out the cadence, that’s kind of the final step in the puzzle.
193 00:26:18.300 ⇒ 00:26:20.140 Jake Nathan: And you’re ready to go after that.
194 00:26:20.660 ⇒ 00:26:23.279 Demilade Agboola: Yeah, yeah, ready to go into implementation, yes.
195 00:26:24.610 ⇒ 00:26:40.659 Jake Nathan: Gotcha. And, okay, that’s… this is a super, super helpful playbook. So, back, I guess, to the… the first one, metrics. So, you mentioned stock is definitely really important if you can’t… if you don’t have something in stock, then that’s…
196 00:26:40.810 ⇒ 00:26:52.349 Jake Nathan: a big issue. Failed transactions. Are there any other metrics that come to mind, like, critical metrics in this case, that you… you’ve kind of seen a pattern in the companies that you’ve worked with?
197 00:26:53.680 ⇒ 00:26:59.220 Demilade Agboola: Yeah, so, like, stock can just be referred to as inventory, so, like, inventory is…
198 00:27:03.320 ⇒ 00:27:07.390 Demilade Agboola: So, inventory… transactions…
199 00:27:14.920 ⇒ 00:27:17.920 Demilade Agboola: Those tend to be the major ones,
200 00:27:19.040 ⇒ 00:27:23.709 Demilade Agboola: Those are, like, the critical, oh, oh, shoot, something’s going on here.
201 00:27:23.970 ⇒ 00:27:26.759 Demilade Agboola: We need to get on top of it.
202 00:27:27.090 ⇒ 00:27:27.720 Jake Nathan: Yeah.
203 00:27:28.330 ⇒ 00:27:36.640 Jake Nathan: That makes sense. So… so now that we kind of have the… the playbook laid out.
204 00:27:36.790 ⇒ 00:27:41.130 Jake Nathan: Are there any other, I guess…
205 00:27:41.360 ⇒ 00:27:47.230 Jake Nathan: you don’t have to, like, obviously I wouldn’t share any, like, client information in particular, but are there any…
206 00:27:47.400 ⇒ 00:28:02.810 Jake Nathan: Stories that you have from your client work, from either something that went wrong, or, just really any personal narrative that you have from the clients that you’ve worked with of why, like, the steps that you laid out are important.
207 00:28:03.720 ⇒ 00:28:10.799 Demilade Agboola: Another metric to add is also, like, just conversion rates, especially if, like, you have your marketing team, like, looking at that.
208 00:28:11.650 ⇒ 00:28:16.660 Demilade Agboola: I want to just know if, like, your conversion is dropping.
209 00:28:17.230 ⇒ 00:28:23.450 Demilade Agboola: Could it be, like, you know, they might need… like, your marketing team might need to be able to make decisions right there, right then.
210 00:28:23.630 ⇒ 00:28:29.769 Demilade Agboola: about… where to push more ads, for instance. Maybe Facebook is giving you higher quality
211 00:28:30.040 ⇒ 00:28:35.100 Demilade Agboola: conversion than saying, you know, like, Instagram or Twitter or something.
212 00:28:36.430 ⇒ 00:28:38.050 Jake Nathan: That makes sense.
213 00:28:38.380 ⇒ 00:28:40.820 Jake Nathan: I think those are good, like, yeah, 3.
214 00:28:41.070 ⇒ 00:28:42.940 Jake Nathan: Three metrics that are important.
215 00:28:47.150 ⇒ 00:28:48.110 Jake Nathan: Awesome.
216 00:28:50.810 ⇒ 00:29:01.549 Jake Nathan: And yeah, kind of back to my other question just about, whether it’s with… I think you said you worked with Urban Stems, you worked with,
217 00:29:03.010 ⇒ 00:29:05.039 Jake Nathan: Did you work with Eden as well?
218 00:29:05.800 ⇒ 00:29:08.769 Demilade Agboola: Yeah, I work with Eden, but Eden doesn’t necessarily have a, like.
219 00:29:09.020 ⇒ 00:29:12.520 Demilade Agboola: high intensity… I’m not experiencing high intensity. Yeah.
220 00:29:13.020 ⇒ 00:29:16.219 Jake Nathan: That’s true. Well, I guess with urban stems, like, is there…
221 00:29:16.490 ⇒ 00:29:22.290 Jake Nathan: like, did everything go pretty smoothly? And I won’t mention, like, actual, you know,
222 00:29:23.420 ⇒ 00:29:33.870 Jake Nathan: urban stems, whatever, like, UTAM will make sure, like, it’s all, you know, what we’re saying is appropriate to publish. But, yeah, just, like, any stories or any, like.
223 00:29:34.080 ⇒ 00:29:35.450 Jake Nathan: Anything from there.
224 00:29:36.560 ⇒ 00:29:44.530 Demilade Agboola: Yeah, I think… Prior to us coming in, urban Stems had, like, polluted.
225 00:29:45.110 ⇒ 00:29:49.409 Demilade Agboola: I’m obviously gonna… I trust you’re gonna, like, filter out, like, I’m just gonna talk, but, like…
226 00:29:49.680 ⇒ 00:29:50.140 Jake Nathan: Totally.
227 00:29:50.570 ⇒ 00:29:54.039 Demilade Agboola: Urban Sims had, like, floated, infrastructure.
228 00:29:54.350 ⇒ 00:29:58.129 Demilade Agboola: It took a while, the jobs were not running, and it kind of…
229 00:29:58.700 ⇒ 00:30:08.290 Demilade Agboola: took… for them to get their numbers, they had a number of failures in their jobs. So, but for other understands, we’re able to nail down
230 00:30:08.500 ⇒ 00:30:14.769 Demilade Agboola: Critical numbers that they needed for inventory analysis on that day.
231 00:30:15.790 ⇒ 00:30:18.330 Demilade Agboola: And then build out and exclude the…
232 00:30:19.040 ⇒ 00:30:21.460 Demilade Agboola: important models that are in there to have.
233 00:30:21.970 ⇒ 00:30:29.840 Demilade Agboola: We’re able to rebuild, and in some cases, add new information that they needed that hadn’t been available.
234 00:30:30.130 ⇒ 00:30:34.479 Demilade Agboola: And then we’re able to kind of stress test it prior to launch.
235 00:30:34.680 ⇒ 00:30:40.319 Demilade Agboola: And then, when we saw the numbers were fine, audited, we then launched it, and…
236 00:30:40.490 ⇒ 00:30:58.180 Demilade Agboola: yeah, they… they… they had a pretty good Mother’s Day, compared to what they were expecting. Because they had also had… so it was Mother’s Day, and they had had Valentine’s Day previously. Valentine’s Day was a hassle. Like, it was really… they had so many complaints.
237 00:30:58.380 ⇒ 00:31:00.300 Demilade Agboola: We weren’t working with them then.
238 00:31:00.660 ⇒ 00:31:04.700 Demilade Agboola: But Mother’s Day was… they felt pretty good.
239 00:31:05.230 ⇒ 00:31:08.529 Demilade Agboola: After Mother’s Day. They had a whole, like.
240 00:31:09.320 ⇒ 00:31:15.200 Demilade Agboola: retro, and again, they were pretty satisfied. I mean, it wasn’t 100% perfect, but, like.
241 00:31:15.310 ⇒ 00:31:18.670 Demilade Agboola: The, you know, the improvement was… was massive.
242 00:31:20.130 ⇒ 00:31:25.430 Jake Nathan: Yeah, do you remember off the top of your head, and I could ask you, Tom, this too,
243 00:31:25.600 ⇒ 00:31:31.139 Jake Nathan: But do you… do you remember, like, did they… were there any numbers that they said, like, did this…
244 00:31:31.870 ⇒ 00:31:36.320 Jake Nathan: You know, was it 10% better, like, any sort of metrics?
245 00:31:37.480 ⇒ 00:31:39.080 Demilade Agboola: You might have to ask Utan for that.
246 00:31:39.670 ⇒ 00:31:40.050 Jake Nathan: Okay.
247 00:31:40.700 ⇒ 00:31:41.370 Demilade Agboola: It’s mine.
248 00:31:41.970 ⇒ 00:31:47.680 Jake Nathan: Gotcha. And you mentioned, like, in general, they were happy. The…
249 00:31:48.250 ⇒ 00:32:02.930 Jake Nathan: I’m guessing, like, a lot of it was on their data team, like… like, when you’re preparing, like, obviously, I mean, the whole company in some way has to prepare, but, like, when you’re kind of talking about this, you kind of mentioned a few different teams, like, in your eyes, like, what are the big…
250 00:32:03.050 ⇒ 00:32:11.050 Jake Nathan: stakeholders that need to be aligned here? Like, what teams need to be aligned, for… for kind of that playbook that you just laid out?
251 00:32:13.300 ⇒ 00:32:21.610 Demilade Agboola: So every… first off, everyone needs to be, but the important people that need to be aligned are the executives, because they need to understand
252 00:32:21.970 ⇒ 00:32:25.320 Demilade Agboola: What they need and what’s, like, critical for that day.
253 00:32:25.720 ⇒ 00:32:28.859 Demilade Agboola: Operations, always critical.
254 00:32:29.180 ⇒ 00:32:32.109 Demilade Agboola: the data team, Always critical.
255 00:32:32.940 ⇒ 00:32:39.459 Demilade Agboola: And then, depending on, like, what they have planned, again, this is… this can be subjective, could be marketing.
256 00:32:40.260 ⇒ 00:32:43.130 Demilade Agboola: Marketing might be important. Some places it might not be.
257 00:32:43.750 ⇒ 00:32:46.250 Demilade Agboola: It might just be operations in some places.
258 00:32:46.500 ⇒ 00:32:52.470 Demilade Agboola: But yeah, it’s just important to be… for everyone to be able to know, every team to be aligned on what… what…
259 00:32:53.540 ⇒ 00:32:55.469 Demilade Agboola: What their role is on that day.
260 00:32:55.640 ⇒ 00:32:58.409 Demilade Agboola: And what they need to see for it to be a successful day.
261 00:33:00.120 ⇒ 00:33:05.779 Jake Nathan: Yeah, that… that makes sense. And that almost, to me, feels like a… almost bonus, like.
262 00:33:05.990 ⇒ 00:33:13.059 Jake Nathan: Final step after that of, like, you have everything, and then now it’s, like, the communication of, like you’re saying,
263 00:33:13.350 ⇒ 00:33:29.970 Jake Nathan: having the executives, the operations, the… obviously the data team, and the marketing team, like, once… once you have everything, like, all those models built out, everything ready to go, I’m sure part of it is also communicating that to other teams to make sure they understand what’s going on as well.
264 00:33:30.900 ⇒ 00:33:31.870 Demilade Agboola: Yes.
265 00:33:34.740 ⇒ 00:33:41.639 Jake Nathan: Cool. And so, yeah, this was… this was super… I think this is gonna be awesome. I’m excited to put this together, and…
266 00:33:41.640 ⇒ 00:33:56.039 Jake Nathan: I feel confident. I mean, also, I have the recording, so I can listen back, but remind me one more time, just so I am super clear on the… like, the… basically, the two problems of why this is unique is, one, you’re saying the latency, and then two.
267 00:33:56.040 ⇒ 00:34:02.900 Jake Nathan: you’re saying it’s, like, I guess the problem is a lot of teams are…
268 00:34:03.120 ⇒ 00:34:12.790 Jake Nathan: not proactive, and so they’re looking at data, you know, they’re doing… like, they’re… instead of being… or yeah, go ahead, in your eyes, what is it?
269 00:34:13.580 ⇒ 00:34:17.719 Demilade Agboola: If the data comes in, if you have, like, If you have…
270 00:34:20.909 ⇒ 00:34:23.089 Demilade Agboola: If you have data that isn’t…
271 00:34:23.449 ⇒ 00:34:25.889 Demilade Agboola: If you haven’t prayed for that day.
272 00:34:26.150 ⇒ 00:34:37.430 Demilade Agboola: On that day, you’ll constantly be behind that day, because most people don’t necessarily have, like, real-time data infrastructure, or close to real-time data infrastructure, because it’s generally considered unnecessary.
273 00:34:37.560 ⇒ 00:34:50.489 Demilade Agboola: So usually what you end up having is daily reports, so you see what happened the previous day, or you can see what happened over the past 6 hours, or whatever cadence. Usually 6, 12, or 24 hours cadence.
274 00:34:51.040 ⇒ 00:34:53.199 Demilade Agboola: That’s usually fairly common.
275 00:34:54.280 ⇒ 00:34:59.440 Demilade Agboola: The problem with that sort of cadence is, on such a high-intensity, high-volume day.
276 00:34:59.900 ⇒ 00:35:10.120 Demilade Agboola: you can lose… like, think about it, if the cadence is 6am, 12pm, 6pm, and then 12 a.m.
277 00:35:10.410 ⇒ 00:35:14.370 Demilade Agboola: Effectively, if you look at it at 12 PM,
278 00:35:15.410 ⇒ 00:35:18.470 Demilade Agboola: you’ve already lost so much of your day. If you make an.
279 00:35:18.470 ⇒ 00:35:18.830 Jake Nathan: 18.
280 00:35:18.830 ⇒ 00:35:20.700 Demilade Agboola: points in time.
281 00:35:21.480 ⇒ 00:35:30.259 Demilade Agboola: you might be lucky to still save that day. But if you had a 24-hour cadence, it means if you realize something went wrong.
282 00:35:30.410 ⇒ 00:35:41.690 Demilade Agboola: maybe you look at your daily report and see that sales one usually load that day. Well, unfortunately, you can’t do anything to save that day. So, you want to be, as much as possible.
283 00:35:42.060 ⇒ 00:35:45.669 Demilade Agboola: Be as close to any of the problems that could occur.
284 00:35:47.190 ⇒ 00:35:57.850 Demilade Agboola: And in that case, the cadence allows you to be able to handle that. So a one-hour cadence means that if things are going askew, you can see within the hour.
285 00:35:58.530 ⇒ 00:36:08.670 Demilade Agboola: and potentially get it fixed out within the next hour, or within 2 hours, or something, and maybe save your day. If you don’t have that, and it’s a 12-hour sync.
286 00:36:08.830 ⇒ 00:36:14.800 Demilade Agboola: It could be the end of the day, or the end of the business day, and it’s kind of late to be able to just try and save that day.
287 00:36:16.430 ⇒ 00:36:19.299 Jake Nathan: Totally, that makes sense. So, kind of the issue is that…
288 00:36:19.800 ⇒ 00:36:23.140 Jake Nathan: Teams tend to fall behind because they…
289 00:36:23.370 ⇒ 00:36:31.809 Jake Nathan: They don’t change their cadence for this specific High-intensity day, they… Or, yeah, period there.
290 00:36:32.480 ⇒ 00:36:39.529 Demilade Agboola: So, teams can fall behind, because if they use what they… if they use the infrastructure that is built for the regular days.
291 00:36:39.980 ⇒ 00:36:44.829 Demilade Agboola: You will fall behind, because regular days are not, you know, high intensity days.
292 00:36:45.090 ⇒ 00:36:46.330 Demilade Agboola: a high volume.
293 00:36:47.380 ⇒ 00:36:58.520 Jake Nathan: Gotcha. And then I’d say almost a third issue is, they’re all kind of similar, but the third thing is, like, you’re… you’re saying, really, you should be starting to…
294 00:36:58.740 ⇒ 00:37:06.710 Jake Nathan: Talk about this even up to 2 months in advance. And so, teams need to, kind of.
295 00:37:07.220 ⇒ 00:37:17.940 Jake Nathan: I’m sure there’s a lot of teams that are scrambling, and so, like, that’s another issue. So, being able to, like, giving yourself enough time to prepare seems like,
296 00:37:18.260 ⇒ 00:37:36.890 Jake Nathan: a mistake that they make as well. So, yeah, and then I think, like, we just talked about, like, you get the metrics, you analyze the infrastructure, you figure out your models, you optimize them, you carve them out, and then you talk to the team as a whole and make sure everyone’s aligned. So, I think,
297 00:37:37.210 ⇒ 00:37:42.689 Jake Nathan: Do you have any final… final words, or any… anything else that we didn’t cover?
298 00:37:42.850 ⇒ 00:37:45.279 Jake Nathan: That you… you want to make sure we do cover?
299 00:37:46.270 ⇒ 00:37:51.039 Demilade Agboola: Not right now, I think… I think that pretty much… that pretty much is it.
300 00:37:53.530 ⇒ 00:38:03.530 Jake Nathan: Awesome. Well, yeah, this was… this is super helpful. Thanks for coming ready with… with a playbook, and kind of next steps for me are I’m gonna…
301 00:38:03.840 ⇒ 00:38:14.909 Jake Nathan: Like, listen back to our interview and start… start writing this down and… and, coming up with… with an article, and then,
302 00:38:15.130 ⇒ 00:38:19.109 Jake Nathan: you, me, and U, Tom, can… like, I’ll send it in our…
303 00:38:19.540 ⇒ 00:38:29.149 Jake Nathan: Slack, and we can just make sure if there’s any edits or anything you want to change, and then we can, we’ll loop in Hannah and the design team to…
304 00:38:29.440 ⇒ 00:38:35.600 Jake Nathan: To get it designed, and then, you know, we’ll publish it on the website and have
305 00:38:35.600 ⇒ 00:38:51.149 Jake Nathan: like, a PDF that we’ll send out as kind of, like, a sales enablement tool, so, like, when we’re having more sales conversations. So, anyway, this has been… this has been super helpful, and hey, also, good luck on the… on the best man speech tomorrow. Maybe I gave you a little,
306 00:38:51.430 ⇒ 00:38:56.489 Jake Nathan: A little preparation, just, like, getting some reps speaking, so you’re gonna do a great job.
307 00:38:57.040 ⇒ 00:39:04.109 Demilade Agboola: Okay, sounds good. Thank you very much. I look forward to it, and I’m looking forward to, like, you know, what you come up with. I bet it’s gonna be…
308 00:39:04.240 ⇒ 00:39:05.730 Demilade Agboola: It’s gonna knock my socks off.
309 00:39:06.100 ⇒ 00:39:10.339 Jake Nathan: Let’s go. Awesome. Alright, well, I’ll talk to you later.
310 00:39:10.660 ⇒ 00:39:12.050 Demilade Agboola: Alright, talk to you later. Take care.
311 00:39:12.050 ⇒ 00:39:12.870 Jake Nathan: Okay, bye.