Meeting Title: Brainforge Demand Forecasting Discussion Date: 2025-12-11 Meeting participants: Sezim Zhenishbekova, Amber Lin
WEBVTT
1 00:00:54.640 ⇒ 00:00:55.860 Amber Lin: Hello!
2 00:00:56.020 ⇒ 00:00:57.550 Sezim Zhenishbekova: Hi, how are you?
3 00:00:58.380 ⇒ 00:01:01.029 Amber Lin: I’m good. Isn’t it already 5PM for you?
4 00:01:01.370 ⇒ 00:01:04.910 Sezim Zhenishbekova: Yes, it is. It is 5-12 now.
5 00:01:05.480 ⇒ 00:01:08.249 Amber Lin: Don’t make it quick, I don’t wanna keep you after it.
6 00:01:08.250 ⇒ 00:01:08.850 Sezim Zhenishbekova: No, no worries.
7 00:01:08.850 ⇒ 00:01:09.640 Amber Lin: hours.
8 00:01:09.800 ⇒ 00:01:16.660 Sezim Zhenishbekova: No, I… I’m flexible, because I work part-time, that’s why I can, like, shift hours.
9 00:01:16.660 ⇒ 00:01:17.460 Amber Lin: Hmm…
10 00:01:17.460 ⇒ 00:01:21.910 Sezim Zhenishbekova: Yeah, it’s a bit mixed, than usual. What time is… where are you based? I forgot.
11 00:01:21.910 ⇒ 00:01:25.869 Amber Lin: I’m in LA, so it’s 2… 2.12 right now.
12 00:01:25.870 ⇒ 00:01:29.420 Sezim Zhenishbekova: Oh, okay, cool. How long have you been living in LA?
13 00:01:29.820 ⇒ 00:01:35.389 Amber Lin: I moved here, say, 2, 3… 3 years ago.
14 00:01:35.550 ⇒ 00:01:35.960 Sezim Zhenishbekova: Too high.
15 00:01:35.960 ⇒ 00:01:37.139 Amber Lin: Two years ago?
16 00:01:38.140 ⇒ 00:01:40.070 Sezim Zhenishbekova: Were you before?
17 00:01:40.070 ⇒ 00:01:46.020 Amber Lin: I’m originally based from China, and then I went to Canada…
18 00:01:46.210 ⇒ 00:01:47.850 Amber Lin: When I was in high school.
19 00:01:48.530 ⇒ 00:01:54.419 Amber Lin: at my university, I went to Hong Kong for a bit, went to Europe for a bit, and then,
20 00:01:54.840 ⇒ 00:01:56.299 Amber Lin: Back in the States.
21 00:01:56.520 ⇒ 00:01:58.100 Sezim Zhenishbekova: Nice, that’s cool.
22 00:01:58.100 ⇒ 00:01:59.379 Amber Lin: Yeah, what about you?
23 00:01:59.810 ⇒ 00:02:02.900 Sezim Zhenishbekova: Yeah, I’m originally from Kyrgyzstan, I got my.
24 00:02:02.900 ⇒ 00:02:03.270 Amber Lin: Hmm.
25 00:02:03.270 ⇒ 00:02:18.539 Sezim Zhenishbekova: a bachelor’s degree there, and then… and then just traveled to Europe as well, lived in the Netherlands for a bit, and then just decided to come back to the U.S, because I was an exchange student during my high school year in Michigan. So, yeah, and then I just…
26 00:02:18.540 ⇒ 00:02:19.100 Amber Lin: Very cool.
27 00:02:19.100 ⇒ 00:02:22.029 Sezim Zhenishbekova: Here, 3 years ago to get my master’s degree.
28 00:02:22.990 ⇒ 00:02:23.823 Sezim Zhenishbekova: New York.
29 00:02:24.590 ⇒ 00:02:29.629 Amber Lin: Cool. Are you also… are you also on… A student visa right now?
30 00:02:29.630 ⇒ 00:02:32.349 Sezim Zhenishbekova: Yes, are you, are you too? Almost.
31 00:02:32.350 ⇒ 00:02:35.730 Amber Lin: Yeah, I am. Are you on…
32 00:02:35.730 ⇒ 00:02:36.420 Sezim Zhenishbekova: Thank you.
33 00:02:36.420 ⇒ 00:02:38.010 Amber Lin: OPT with them?
34 00:02:38.370 ⇒ 00:02:46.839 Sezim Zhenishbekova: I haven’t done anything yet, because I have another part-time job that’s actually, filing for my STEM extension right now.
35 00:02:46.840 ⇒ 00:02:47.680 Amber Lin: Mmm.
36 00:02:47.680 ⇒ 00:02:52.500 Sezim Zhenishbekova: It’s finishing in January, like, 29th, so…
37 00:02:52.500 ⇒ 00:03:08.579 Amber Lin: I am on the same timeline as you, and today, in the past month, I was so stressed, because I’m trying to get them to register and E-Verify, and, like, the… because Uten doesn’t handle it. Their external finance team handles it, and I don’t think…
38 00:03:08.710 ⇒ 00:03:14.990 Amber Lin: They went through with it, so, like, right now, they don’t… they don’t know their username and password.
39 00:03:15.350 ⇒ 00:03:26.999 Amber Lin: So… so I was like, for a month, I was like, okay, the government shut down, so it’s slower, but it’s really because, people forgot, and so I’m checking.
40 00:03:27.000 ⇒ 00:03:27.390 Sezim Zhenishbekova: Checking in.
41 00:03:27.390 ⇒ 00:03:28.460 Amber Lin: Right now, I was like, can you.
42 00:03:28.460 ⇒ 00:03:28.790 Sezim Zhenishbekova: I’m reading.
43 00:03:28.790 ⇒ 00:03:32.780 Amber Lin: Please finish this, because my timeline is also in February.
44 00:03:32.780 ⇒ 00:03:33.710 Sezim Zhenishbekova: Oh, yeah, exactly.
45 00:03:33.710 ⇒ 00:03:47.000 Amber Lin: It’s really late already. I know, I didn’t… you know, when I asked my other boss to do that, like, I’m working there for 10 hours, but the STEM extension turns that it has to be minimum 20 hours.
46 00:03:47.000 ⇒ 00:04:04.669 Sezim Zhenishbekova: to work more for him. I convinced him, he said, okay, like, I can hire you for 20 hours now, but then… and he’s now, like, I’ve negotiated the terms, I filled everything out for him, and now I’m waiting for e-verification. He’s like, oh, it says, like, I will do this, I’ll do this, like, week past, he didn’t do it yet.
47 00:04:04.670 ⇒ 00:04:05.280 Amber Lin: Oh, man.
48 00:04:05.280 ⇒ 00:04:07.569 Sezim Zhenishbekova: I think I should keep pushing him, then.
49 00:04:07.570 ⇒ 00:04:22.550 Amber Lin: It’s… it’s so… like, if you end up with Brainforge, like, by the time you sign, I think we should have our E-Verify, because we don’t have it yet. We haven’t completed the training, or did the… the tiny exams, so…
50 00:04:22.980 ⇒ 00:04:32.580 Amber Lin: I’ll let you… I’ll let you know, because I think that’s important for you, too, because if they don’t have… if Brainfor doesn’t have E-Verify, like, there’s no point you can work full-time here.
51 00:04:32.580 ⇒ 00:04:42.049 Sezim Zhenishbekova: Yeah, exactly, that’s why, in another job, I’m already signing a contract as well, and I’m like, oh my god, like, it’s gonna be overwhelming. And because I’m working with them too.
52 00:04:42.280 ⇒ 00:04:47.290 Sezim Zhenishbekova: like, part-time. I also need to perform well, because I’m not sure about brain forage, too.
53 00:04:47.290 ⇒ 00:04:48.010 Amber Lin: Yeah.
54 00:04:48.700 ⇒ 00:04:49.440 Amber Lin: Yeah
55 00:04:49.710 ⇒ 00:05:02.860 Amber Lin: Okay. Let me, let me let Robert know, because I think, I, I thought they were deciding on making a decision last week, but I think because, like, insomnia stuff, like, it wasn’t done yet.
56 00:05:03.280 ⇒ 00:05:05.330 Amber Lin: Thinking about this week, we’re…
57 00:05:05.460 ⇒ 00:05:17.150 Amber Lin: like, mid-December, so let me see if they can… that just gives them a better reason to get E-Verify done, so if they get that done, then… then you have more choices, because then…
58 00:05:17.570 ⇒ 00:05:18.510 Amber Lin: Like, I don’t…
59 00:05:18.510 ⇒ 00:05:33.990 Sezim Zhenishbekova: He told me that I have a review session next week, sometime on Tuesday or Wednesday, so I think it will be good. But yeah, other way around, if I will get in or not, I’m like, whatever happens, happens for a reason.
60 00:05:33.990 ⇒ 00:05:36.780 Amber Lin: You know, as long as your other company has a verified, like…
61 00:05:36.780 ⇒ 00:05:41.030 Sezim Zhenishbekova: I’m like, I’m already tired of this, like, I will just go.
62 00:05:41.030 ⇒ 00:05:41.430 Amber Lin: over.
63 00:05:41.430 ⇒ 00:05:43.759 Sezim Zhenishbekova: Well, it happens, I do it. If not, it’s also key.
64 00:05:43.760 ⇒ 00:05:55.960 Amber Lin: No, I also… I also… because their e-verify was taking so long, I searched up, schools that have day one CPT, so if it really ends up that it’s too late, you can…
65 00:05:56.060 ⇒ 00:05:59.800 Amber Lin: Because you did your master’s, so you can do, like, a doctoral…
66 00:06:00.050 ⇒ 00:06:00.560 Sezim Zhenishbekova: Yeah.
67 00:06:00.560 ⇒ 00:06:03.119 Amber Lin: day one CBTF, like, nothing goes well.
68 00:06:03.900 ⇒ 00:06:05.530 Sezim Zhenishbekova: That’s… that’s… yeah.
69 00:06:06.320 ⇒ 00:06:20.479 Sezim Zhenishbekova: Wow, okay, I didn’t know you… you’re… I saw, like, there’s no one from, with international… with status at all here, but yeah, I think it’s me on… and me and you that you’re only, right? Or other…
70 00:06:20.480 ⇒ 00:06:25.470 Amber Lin: So, yeah, there’s no… student… I don’t think.
71 00:06:25.720 ⇒ 00:06:26.350 Sezim Zhenishbekova: Mmm.
72 00:06:27.260 ⇒ 00:06:31.540 Amber Lin: Yeah, okay, I’ll keep that in mind. I’m talking to Utam later, so I’ll…
73 00:06:31.620 ⇒ 00:06:33.350 Sezim Zhenishbekova: Master’s student, by the way?
74 00:06:33.350 ⇒ 00:06:44.659 Amber Lin: No, I went to… I finished my undergrad here, so, like, if everything fails, I still have my master’s and a doctorate.
75 00:06:45.180 ⇒ 00:06:47.390 Amber Lin: Dude, go to school!
76 00:06:47.390 ⇒ 00:06:52.369 Sezim Zhenishbekova: Yeah, it’s good, at least if I can do Masters, I guess it’s fine.
77 00:06:52.890 ⇒ 00:07:09.319 Sezim Zhenishbekova: Yeah, but I think they will do it. I was getting their official… I need it for my OPT filing, even if it’s a contract job, I have to file for, even the job that I… like, as a second job, I have to file it on board, so I was waiting for confirmed
78 00:07:09.800 ⇒ 00:07:14.710 Sezim Zhenishbekova: address, registered legal address for Brainforge, like, for a week, I think.
79 00:07:14.710 ⇒ 00:07:19.660 Amber Lin: Mmm. I, I have that, do you need that? I already got it, I just… Okay, okay.
80 00:07:19.660 ⇒ 00:07:23.229 Sezim Zhenishbekova: my copy system, but they, yeah, that’s what they did.
81 00:07:23.230 ⇒ 00:07:23.720 Amber Lin: tea.
82 00:07:23.720 ⇒ 00:07:27.850 Sezim Zhenishbekova: So, even those small address took them a while to file.
83 00:07:27.850 ⇒ 00:07:32.890 Amber Lin: Yeah, nobody does the operations side, so it takes a long time.
84 00:07:33.430 ⇒ 00:07:38.040 Sezim Zhenishbekova: Good. Okay, so, I didn’t, like, could you please call.
85 00:07:38.040 ⇒ 00:07:51.200 Amber Lin: Yeah, yeah. So, let me show you this, this document. So, Honey Stinger is the brand… oh, let me share screen, and I’ll walk you through.
86 00:07:51.830 ⇒ 00:07:55.140 Amber Lin: And… So…
87 00:07:59.840 ⇒ 00:08:09.790 Amber Lin: So, this is Honey Singer. So, they sell… Energy, waffles, And then they sell… Stuff.
88 00:08:10.210 ⇒ 00:08:19.280 Amber Lin: And bars, and shoes, and that. So they sell, like, consumer food items, sweet items. And then, they are on…
89 00:08:19.530 ⇒ 00:08:25.419 Amber Lin: Amazon. So, here’s their crackers there on Amazon.
90 00:08:25.690 ⇒ 00:08:34.510 Amber Lin: They’re on Shopify, which is their DTC, and then they’re on some other platforms. Walmart, Target, Sam’s Club. Pretty small, but…
91 00:08:35.110 ⇒ 00:08:35.990 Amber Lin: exists.
92 00:08:36.380 ⇒ 00:08:41.459 Amber Lin: And the task right now is,
93 00:08:41.669 ⇒ 00:08:44.670 Amber Lin: The stakeholder wants to forecast demand.
94 00:08:44.790 ⇒ 00:08:56.079 Amber Lin: And this is their high-level forecast, of demand, but I think they most likely just did it very basically on this
95 00:08:57.160 ⇒ 00:09:01.940 Amber Lin: I don’t exactly know how they did it, but it’s probably very, very high level.
96 00:09:01.940 ⇒ 00:09:05.280 Sezim Zhenishbekova: Of, oh, growth rate changes, or yearly changes.
97 00:09:05.920 ⇒ 00:09:08.239 Amber Lin: And right now, they have it…
98 00:09:08.240 ⇒ 00:09:10.170 Sezim Zhenishbekova: How cormless, right?
99 00:09:10.170 ⇒ 00:09:19.379 Amber Lin: I… I… my Excel expired, so probably it is in Excel. I can send the file to you. Let me do that real quick.
100 00:09:20.300 ⇒ 00:09:23.790 Amber Lin: And… Nope.
101 00:09:29.470 ⇒ 00:09:31.750 Amber Lin: Let me send that to you.
102 00:09:38.530 ⇒ 00:09:39.530 Amber Lin: So…
103 00:09:41.190 ⇒ 00:09:49.500 Amber Lin: And… so I guess the first step’s to see how they did it, but I thought about it, and I think, so far, I think…
104 00:09:49.690 ⇒ 00:09:59.030 Amber Lin: First, I have to do 3 separate forecasts for each of the big channels, and also probably have to do it by…
105 00:09:59.500 ⇒ 00:10:04.969 Amber Lin: ASIN, roll it up into months, and then…
106 00:10:05.070 ⇒ 00:10:09.849 Amber Lin: Roll it up into my channel, and then by year.
107 00:10:10.740 ⇒ 00:10:18.570 Amber Lin: So Sof… bar… Like, I’m thinking of…
108 00:10:20.490 ⇒ 00:10:24.320 Amber Lin: I have never done demand forecasting before, so this is…
109 00:10:24.510 ⇒ 00:10:33.909 Amber Lin: Like, I feel like I’m missing a lot of steps, and I… like, it feels so complicated, I don’t know where to start. I’m thinking for Amazon, because they’re a…
110 00:10:34.120 ⇒ 00:10:45.190 Amber Lin: They’re first-party sellers, so they don’t sell to the consumer, they sell to Amazon. So it’s different, a bit different than sales, so there’s still, like, a seasonal…
111 00:10:45.660 ⇒ 00:10:49.150 Amber Lin: like, procurements, Amazon orders from them.
112 00:10:49.150 ⇒ 00:10:50.889 Sezim Zhenishbekova: And then I want to…
113 00:10:51.480 ⇒ 00:11:07.259 Amber Lin: times that by the different trends of seasonalities of, like, Amazon deal days, maybe category performances, and then check if there’s any inventory constraints. Like, that’s what I have so far.
114 00:11:07.390 ⇒ 00:11:14.320 Amber Lin: And then I’ll do Shopify differently, and Shopify will probably just be, like, traffic, conversion.
115 00:11:14.530 ⇒ 00:11:25.440 Amber Lin: average order value, like, that type of stuff. I think the main thing is… Like, I… I… Amazon’s confusing.
116 00:11:25.770 ⇒ 00:11:36.939 Sezim Zhenishbekova: Yeah, I can imagine. So basically, you have this plan, but you’re struggling to systematize where to start and how, right?
117 00:11:37.350 ⇒ 00:11:46.290 Amber Lin: Yeah, and because I’ve never done it before, it’s hard for me to see where things might go wrong, or if I’m leaving things out.
118 00:11:46.390 ⇒ 00:11:50.270 Amber Lin: Have you done demand… Forecasting before?
119 00:11:50.460 ⇒ 00:12:06.810 Sezim Zhenishbekova: So, I haven’t done demand forecasting, but it’s, like, it’s a lot of dependencies that you build. You have the data and just analyze the past performances and give some chain percentage changes that could affect the future demand.
120 00:12:06.920 ⇒ 00:12:23.689 Sezim Zhenishbekova: is, gonna, like, help you out to forecast for the future, and you’ve written down several, like, formulas, like, baseline seasonality, depending on the trend, but, like, what is the trend, right? Like, there are some questions that may arise, so…
121 00:12:24.080 ⇒ 00:12:30.499 Sezim Zhenishbekova: for the baseline seasonality and trend, I would recommend to, like.
122 00:12:31.650 ⇒ 00:12:43.960 Sezim Zhenishbekova: take columns and try to understand, like, what’s built now, and then after you do that, what are the dependencies, what are the quarters, and how they’re doing 2026 forecasts, like, based on the Excel, they have the forecast.
123 00:12:44.480 ⇒ 00:12:54.410 Sezim Zhenishbekova: So, first, just focus on historical purchases, I would say. Calculate average, like, per units per month.
124 00:12:54.610 ⇒ 00:12:56.330 Sezim Zhenishbekova: Per SKU.
125 00:12:56.380 ⇒ 00:13:16.309 Sezim Zhenishbekova: And then… and then split it into different quarters that the month, so that will help you to see what’s the seasonality. Like, in quarter one, for example, sales were much higher, and then in the quarter two, it went down or not. So it gives you… it helps you to see what, like, how it reacts with the season changes.
126 00:13:16.620 ⇒ 00:13:31.370 Amber Lin: I see. Another constraint I have on Amazon is that they really just started growing this May, so anytime before, Amazon probably didn’t directly order from them, so their procurement orders
127 00:13:31.560 ⇒ 00:13:38.660 Amber Lin: which is how they earn the money from Amazon, only started this May, and that’s very limited data to forecast.
128 00:13:39.470 ⇒ 00:13:44.299 Amber Lin: say, January through May, so how would you approach that?
129 00:13:44.500 ⇒ 00:13:52.209 Sezim Zhenishbekova: So, usually when I see that, I start to look for competitors.
130 00:13:52.380 ⇒ 00:13:55.990 Amber Lin: And when I see the competitors, I see, like.
131 00:13:55.990 ⇒ 00:13:59.560 Sezim Zhenishbekova: For example, how much… They had themselves.
132 00:13:59.910 ⇒ 00:14:10.900 Sezim Zhenishbekova: or reviews. Like, because sometimes, like, something similar, like, find something similar and find the data for the whole season, like, average year was sold.
133 00:14:10.960 ⇒ 00:14:20.379 Sezim Zhenishbekova: and then use that. There is also another method that I’ve been using is, doing something like, through Google Trends.
134 00:14:21.050 ⇒ 00:14:28.620 Sezim Zhenishbekova: Like, how many times it was searched, and then just assign how much out of that, like, it might be sold, what’s the percentage of it.
135 00:14:29.150 ⇒ 00:14:39.239 Sezim Zhenishbekova: So that can be the second one. So it’s mostly, like, when you don’t have data in hand that it backs up, I go to market to analyze.
136 00:14:39.240 ⇒ 00:14:39.810 Amber Lin: C.
137 00:14:39.810 ⇒ 00:14:53.509 Sezim Zhenishbekova: some similarities with different products, and if I find numbers, then it’s great, and approximate. And of course, there’s gonna be room for error, but it gives you close, like, close estimations on how it’s gonna run.
138 00:14:53.510 ⇒ 00:15:05.800 Amber Lin: I see. Cool. So, let me try and find that. So, the PO orders, like, no, that’s a bad graph. Like, they only really started
139 00:15:05.910 ⇒ 00:15:07.150 Amber Lin: in…
140 00:15:07.840 ⇒ 00:15:19.090 Amber Lin: May this year, like, before they didn’t have anything. But I also have sales, actual sales status, because this is Amazon’s procurement orders to them.
141 00:15:19.590 ⇒ 00:15:20.839 Amber Lin: This is…
142 00:15:21.310 ⇒ 00:15:32.409 Amber Lin: I assume this is Amazon’s actual sales, so how, like, the directly consumer purchases on Amazon. I have that data, and it is quite seasonal.
143 00:15:32.820 ⇒ 00:15:36.610 Amber Lin: Amazon as well. So, maybe…
144 00:15:37.380 ⇒ 00:15:46.330 Amber Lin: I don’t even have to go to competitors, I can take this and infer relationships between sales and…
145 00:15:47.070 ⇒ 00:15:52.019 Amber Lin: Procurement orders, and then use that to, like… Check in the…
146 00:15:52.020 ⇒ 00:15:58.140 Sezim Zhenishbekova: Yeah, I think it’s also a cool idea to check it out and see how it depends on it, yes.
147 00:15:59.770 ⇒ 00:16:00.440 Amber Lin: Okay.
148 00:16:00.440 ⇒ 00:16:00.810 Sezim Zhenishbekova: pork.
149 00:16:01.140 ⇒ 00:16:02.669 Sezim Zhenishbekova: Which is actual, yeah.
150 00:16:03.230 ⇒ 00:16:07.400 Sezim Zhenishbekova: Cool. And then for Shopify, I guess it’s more…
151 00:16:07.430 ⇒ 00:16:19.479 Amber Lin: with a more regular forecast, there’s some traffic, conversions, average order value. Does that look like a reasonable formula to you? What else would you also consider?
152 00:16:19.920 ⇒ 00:16:26.930 Sezim Zhenishbekova: I think… I like that you added the repeated purchases.
153 00:16:27.840 ⇒ 00:16:34.459 Sezim Zhenishbekova: And, like, purchases, so that you see how many… how often people come back to re-buy it.
154 00:16:36.490 ⇒ 00:16:45.099 Sezim Zhenishbekova: so having that also included in the sales calculation and the projection towards it, that will be… I think it’s a good idea to do.
155 00:16:46.210 ⇒ 00:16:53.209 Sezim Zhenishbekova: So, basically, yeah, other than that, traffic conversion… AOV, what is AOV?
156 00:16:53.210 ⇒ 00:16:54.999 Amber Lin: Average order value, so…
157 00:16:55.000 ⇒ 00:16:56.969 Sezim Zhenishbekova: It says…
158 00:16:57.680 ⇒ 00:17:06.560 Sezim Zhenishbekova: So, conversion rate, if you go dive deep into the traffic for each month, right? Like, what was… you just sorted based on the type of…
159 00:17:07.069 ⇒ 00:17:11.959 Sezim Zhenishbekova: paid or organic, or email. And then traffic is mostly about revenue, right?
160 00:17:12.249 ⇒ 00:17:13.239 Sezim Zhenishbekova: How much revenue?
161 00:17:13.240 ⇒ 00:17:19.619 Amber Lin: Yeah, I can show you their… graph. Like, this is their…
162 00:17:19.800 ⇒ 00:17:26.130 Amber Lin: So, this is Shopify. Their conversion rates from visitors to
163 00:17:26.260 ⇒ 00:17:40.190 Amber Lin: like, active customers who had subscriptions and includes subscriptions and first-time purchasers, which is single purchasers. Their conversion rate is pretty stable, it’s about 50%, but
164 00:17:40.880 ⇒ 00:17:44.780 Amber Lin: Like, they’re… Traffic is declining.
165 00:17:46.000 ⇒ 00:17:53.540 Amber Lin: So, I guess… I would break it down into… Like…
166 00:17:54.040 ⇒ 00:17:59.430 Amber Lin: I don’t know, like, single customers versus… Repeat customer?
167 00:17:59.430 ⇒ 00:18:06.260 Sezim Zhenishbekova: Yeah, like, separating them will be good, just to see, like, maybe some, like, one first-time purchase customers.
168 00:18:06.370 ⇒ 00:18:09.870 Amber Lin: And for the repeat customers, and then see how.
169 00:18:09.870 ⇒ 00:18:12.260 Sezim Zhenishbekova: Often, repeat customers come.
170 00:18:13.100 ⇒ 00:18:20.249 Sezim Zhenishbekova: To see that rate, and then you can estimate, what’s the percentage of their return, like, what’s the.
171 00:18:20.620 ⇒ 00:18:22.760 Sezim Zhenishbekova: How often they repeat.
172 00:18:22.880 ⇒ 00:18:25.879 Sezim Zhenishbekova: They’re coming to you, basically.
173 00:18:25.880 ⇒ 00:18:26.230 Amber Lin: Okay.
174 00:18:26.230 ⇒ 00:18:33.240 Sezim Zhenishbekova: Buy this thing, and then that’s gonna give you the base repeat rate that you actually can incorporate into the formula.
175 00:18:33.240 ⇒ 00:18:35.160 Amber Lin: I see.
176 00:18:35.430 ⇒ 00:18:44.820 Amber Lin: I guess the next question is, how do you account for promotions? Like, I know I should include promotional lift, I don’t know how to…
177 00:18:45.410 ⇒ 00:18:46.890 Amber Lin: Put that into the forecast.
178 00:18:46.890 ⇒ 00:18:50.199 Sezim Zhenishbekova: So for Shopify, or Amazon, or…
179 00:18:50.430 ⇒ 00:18:52.610 Amber Lin: Just for Shopify to make it simple.
180 00:18:52.610 ⇒ 00:18:58.199 Sezim Zhenishbekova: So basically, when I see promotion, I look at the sales, like, when the promotion was released.
181 00:18:58.220 ⇒ 00:19:02.289 Amber Lin: I can actually show you the graph, what it looks like.
182 00:19:03.500 ⇒ 00:19:07.650 Amber Lin: That’s… B… Like, this is…
183 00:19:08.040 ⇒ 00:19:12.260 Amber Lin: This is the orders using discount codes.
184 00:19:12.760 ⇒ 00:19:13.180 Sezim Zhenishbekova: borders.
185 00:19:13.180 ⇒ 00:19:18.920 Amber Lin: not… not using discount codes. I would say, like, these are the promotional periods.
186 00:19:20.620 ⇒ 00:19:29.010 Sezim Zhenishbekova: So basically, and when was the… is there a forecast when there was no promotion at all?
187 00:19:31.340 ⇒ 00:19:38.589 Amber Lin: I can ask them. I think that would be… it’s like what you did for Insomnia. It’s probably,
188 00:19:38.730 ⇒ 00:19:42.119 Amber Lin: like, a period where there’s no promotions, I can check that.
189 00:19:42.120 ⇒ 00:19:44.460 Sezim Zhenishbekova: Yeah, just, like, base to see what’s the…
190 00:19:44.460 ⇒ 00:19:45.109 Amber Lin: Huh.
191 00:19:45.110 ⇒ 00:19:53.159 Sezim Zhenishbekova: Correlation, like, between them, how much sales went, and then it will help you to calculate what kind of effect does promotion have.
192 00:19:54.610 ⇒ 00:20:02.150 Sezim Zhenishbekova: And then calcoid… like, revenues, right? Like, cost of acquisition of the customer, COX.
193 00:20:03.390 ⇒ 00:20:05.060 Amber Lin: Oh, gotcha. Okay.
194 00:20:05.660 ⇒ 00:20:21.350 Sezim Zhenishbekova: And then usually, like, when I forecast also, like, when I have, like, the base repeat rate, I would just keep multiplying it, like, what’s the base repeat rate after the promo? And then first, just understand, go to the historical data, the way I think.
195 00:20:21.350 ⇒ 00:20:34.929 Sezim Zhenishbekova: Then try to understand what’s the historical trend line like for the past, and then break it down to seasonability, and how certain promotions, certain websites can affect the result.
196 00:20:34.930 ⇒ 00:20:54.090 Sezim Zhenishbekova: like, define it and number it in percentages, because it’s hard to have specific numbers, so once I have that percentages, then I just take those percentages, compile them, and then do the forecast. And then… and then once you have that percentage number of growth month to month.
197 00:20:56.090 ⇒ 00:21:02.870 Sezim Zhenishbekova: You can add 10% and deduct 10% to see the best and worst forecast scenarios.
198 00:21:02.870 ⇒ 00:21:06.480 Amber Lin: Oh… so… .
199 00:21:06.480 ⇒ 00:21:12.559 Sezim Zhenishbekova: So, like, so you see, like, what can be the best case, and what can be the worst, and what we are saying is the average.
200 00:21:12.560 ⇒ 00:21:21.220 Amber Lin: Let’s see. Just deducting 10% and adding 10% really does the thing. Yeah, so your recommendation is to find the growth rate.
201 00:21:21.220 ⇒ 00:21:27.640 Amber Lin: Do I find individual… see, say, each of these is a different rate.
202 00:21:27.640 ⇒ 00:21:43.299 Amber Lin: when I apply it to, I say, Excel, how do I apply it? Do I just put a rate on it, or do I have multiple pages, just like what you did for Insomnia, and then merge them in one final page together?
203 00:21:43.440 ⇒ 00:22:01.339 Sezim Zhenishbekova: So, the Insomnia one was a bit interesting, right? Like, because of its year-over-year, like, percentages, but it doesn’t really forecast. That’s why I didn’t have much of a gross rate, but I just showed how it was growing over the years, and how each promotion has certain percentages, but…
204 00:22:01.340 ⇒ 00:22:11.609 Sezim Zhenishbekova: Finding the gross rate… yeah, like, it depends all on what you see based on the revenue. Like, I think the easiest one will be the revenue, like, how much product you sold.
205 00:22:11.610 ⇒ 00:22:30.340 Sezim Zhenishbekova: and how revenue grew over time, and just see the trend, like, I think on average, the revenue growth was, for example, 10%, and that you can take as a gross rate. But then also consider, if you want to add to that revenue gross rate something else that can also work, like.
206 00:22:30.530 ⇒ 00:22:48.270 Sezim Zhenishbekova: What kind of effect does promotions have, how much does it lifts? And then… and then just, have it very dynamic. For example, some… probably clients also want to play around with it. For example, what will happen with the forecast if we have…
207 00:22:48.270 ⇒ 00:22:51.859 Sezim Zhenishbekova: some promotion that’s gonna be announced. Like, we’re planning to…
208 00:22:52.240 ⇒ 00:23:10.759 Sezim Zhenishbekova: three to four more promotions, and we want to see how it might affect, and how much we need to have products, in the warehouse. So, like, so that they can tweak it, and they kind of have approximate idea of how much promos can bring the sales growth. And I think…
209 00:23:10.760 ⇒ 00:23:16.309 Amber Lin: How do I build that in? Do I give them a box to fill that in, or…
210 00:23:16.490 ⇒ 00:23:28.349 Sezim Zhenishbekova: Yeah, yeah, you can ref… yeah, yeah, I think that can also… yeah, I think it’s a bit of an art type of thing, you know, when you start thinking of it, but you can…
211 00:23:28.840 ⇒ 00:23:37.549 Sezim Zhenishbekova: Assign, like, certain column with the inputs, like, gross rate, approximate gross rate this much, approximate promotion
212 00:23:37.740 ⇒ 00:23:40.730 Sezim Zhenishbekova: Gross rate per promotion, for example.
213 00:23:41.370 ⇒ 00:23:48.499 Sezim Zhenishbekova: And then… and then every time when you calculate things, you just need to refer to that cell, so it’s more dynamic, yeah.
214 00:23:48.500 ⇒ 00:23:49.980 Amber Lin: Gotcha, okay.
215 00:23:50.290 ⇒ 00:23:53.390 Amber Lin: But your plans seem solid, I think it’s just a matter of…
216 00:23:53.390 ⇒ 00:23:56.320 Sezim Zhenishbekova: Underst… yeah, I think you’re in the right track.
217 00:23:57.620 ⇒ 00:24:04.370 Amber Lin: Okay, that’s… that’s great to hear. I think the actual implementation is going to be a lot harder.
218 00:24:04.370 ⇒ 00:24:07.390 Sezim Zhenishbekova: like, tell me about, because I’m with Eden, I’m like…
219 00:24:07.390 ⇒ 00:24:09.970 Amber Lin: What are you doing on Eden?
220 00:24:09.970 ⇒ 00:24:13.339 Sezim Zhenishbekova: So I have to build also a forecast model, just like yours.
221 00:24:13.340 ⇒ 00:24:19.289 Amber Lin: But I’m doing it for, like, retention, based on retention rate, and I’ve never…
222 00:24:20.460 ⇒ 00:24:28.739 Sezim Zhenishbekova: And I just realized I don’t have access to BigQuery, like, properly. I need to have DeskPocket Tableau, so I’m in, like, very…
223 00:24:28.960 ⇒ 00:24:32.390 Sezim Zhenishbekova: You can’t even get started.
224 00:24:32.580 ⇒ 00:24:37.479 Sezim Zhenishbekova: Yeah, yeah, so I understand it, so you’re doing great, I think. You already have the action plan.
225 00:24:38.250 ⇒ 00:24:43.130 Amber Lin: Appreciate it. Is there any way I can help you with the access,
226 00:24:43.130 ⇒ 00:24:50.140 Sezim Zhenishbekova: I had already a call with, with a couple people. I had a call with Zavaish, and…
227 00:24:50.290 ⇒ 00:24:56.349 Sezim Zhenishbekova: And I’m having a call tomorrow with Henry, so it’s all good for now. Okay. I think it’s just a matter of…
228 00:24:56.630 ⇒ 00:25:04.189 Sezim Zhenishbekova: sitting and playing around with the data, and just trying to understand how one is opposite another, I think.
229 00:25:04.400 ⇒ 00:25:14.080 Sezim Zhenishbekova: Other than that, yeah, I’m gonna record Insomnia Cookies, like, video walkthrough of what I’ve done and what I changed, and then I think it’s…
230 00:25:14.080 ⇒ 00:25:14.650 Amber Lin: Awesome.
231 00:25:14.650 ⇒ 00:25:15.150 Sezim Zhenishbekova: We’re gonna let.
232 00:25:15.150 ⇒ 00:25:20.109 Amber Lin: It looks really, it looks really nice, like, the colors makes it so much easier to read.
233 00:25:20.320 ⇒ 00:25:24.089 Amber Lin: I was, like, thinking on how to make it more…
234 00:25:24.090 ⇒ 00:25:28.390 Sezim Zhenishbekova: affordable. But yeah, thank you for helping me to structure it, too.
235 00:25:28.390 ⇒ 00:25:31.420 Amber Lin: Oh, thank you for helping here, I’m glad, like, we can help each other.
236 00:25:31.420 ⇒ 00:25:33.080 Sezim Zhenishbekova: Yeah,
237 00:25:34.060 ⇒ 00:25:34.770 Amber Lin: Alright.
238 00:25:35.000 ⇒ 00:25:37.850 Sezim Zhenishbekova: Alrighty. Yes, thank you so much, Amber, and…
239 00:25:37.850 ⇒ 00:25:42.419 Amber Lin: Of course. I’ll ask them about the E-Verify stuff. I’ll let you know, too.
240 00:25:42.420 ⇒ 00:25:46.870 Sezim Zhenishbekova: Sounds good. But for now, I don’t really need it, I think, but maybe in the future, if I get.
241 00:25:46.870 ⇒ 00:25:49.270 Amber Lin: If you have to switch employers, then…
242 00:25:49.270 ⇒ 00:25:49.660 Sezim Zhenishbekova: Yeah, well…
243 00:25:49.660 ⇒ 00:25:50.750 Amber Lin: It’ll be an issue.
244 00:25:50.750 ⇒ 00:25:56.840 Sezim Zhenishbekova: Yeah, then I will be able to… that will be amazing. You would do the work for me.
245 00:25:56.840 ⇒ 00:25:58.999 Amber Lin: Sounds good. Have to do it anyways. Okay.
246 00:25:59.000 ⇒ 00:26:02.489 Sezim Zhenishbekova: Right, bye! Good night, see you. Good day.