Meeting Title: Honey Stinger Demand Plan Date: 2025-12-10 Meeting participants: Henry Zhao, Amber Lin
WEBVTT
1 00:01:17.070 ⇒ 00:01:18.310 Amber Lin: Hi there!
2 00:01:22.460 ⇒ 00:01:23.380 Henry Zhao: Alone.
3 00:01:24.520 ⇒ 00:01:26.710 Amber Lin: How are you? Seem tired.
4 00:01:26.710 ⇒ 00:01:29.059 Henry Zhao: Oh, yeah, rough days, it’s like…
5 00:01:29.480 ⇒ 00:01:32.550 Amber Lin: No, the morning was rough. I was in that meeting, I was like, oof.
6 00:01:33.100 ⇒ 00:01:38.230 Amber Lin: That is… that is a tough meeting to handle. I… I could not have it as good as you did.
7 00:01:38.930 ⇒ 00:01:47.610 Henry Zhao: Yeah, thank you. So I’m out the next two days, but I might have to do some work because, I don’t feel good about giving this all to you, right? I think it’s not fair to you.
8 00:01:48.380 ⇒ 00:01:52.399 Henry Zhao: But basically, right, we just need to look at these recommendations from Robert.
9 00:01:52.590 ⇒ 00:01:57.469 Henry Zhao: And kind of build some sort of demand plan, right? Like, just basically forecast 2026 POs.
10 00:01:58.330 ⇒ 00:01:59.990 Henry Zhao: So that they can prepare for that?
11 00:02:02.410 ⇒ 00:02:11.779 Amber Lin: I have not done any forecasting or demand plans before. Could you give me a framework to think about it, or how to do it?
12 00:02:12.260 ⇒ 00:02:20.500 Henry Zhao: Yeah, so what I’m thinking of is, like, something like a subscribe and save. I kind of want to just plot the, like, the growth over time, like, on a chart.
13 00:02:20.670 ⇒ 00:02:21.090 Amber Lin: and see…
14 00:02:21.090 ⇒ 00:02:26.599 Henry Zhao: like, what chart, what, that fits under, right? So this seems to me like a linear one.
15 00:02:26.840 ⇒ 00:02:38.999 Henry Zhao: Where I would have just ignored this minus 20%, because it seems like it’s an edge case, right? And also, I’d rather have extra inventory than not enough, so I would probably just ignore this 20%.
16 00:02:39.030 ⇒ 00:02:46.930 Henry Zhao: and do a line here, and just assume that every MOM is gonna be that much moving forward, and then maybe add a 5% buffer.
17 00:02:49.170 ⇒ 00:02:51.949 Henry Zhao: Because that’s, I think, what Acosta did for this,
18 00:02:52.120 ⇒ 00:02:57.139 Henry Zhao: So for 2026 forecasted POs, it looks like they’re just doing, like, a 5% buffer.
19 00:02:57.590 ⇒ 00:03:01.889 Henry Zhao: And I’ve asked him about that, I was like, why did you do this 5% buffer? I’m just gonna kinda understand that.
20 00:03:02.220 ⇒ 00:03:04.430 Henry Zhao: And see if that’s what we want to do also.
21 00:03:04.630 ⇒ 00:03:09.870 Henry Zhao: And then I was gonna break it down by ASIN to also kind of see… if…
22 00:03:10.500 ⇒ 00:03:16.650 Henry Zhao: forecast by ASIN and then grouped up all together kind of aligns with the…
23 00:03:17.260 ⇒ 00:03:24.070 Henry Zhao: overall forecast that I put out, so I would do, like… so I’m gonna actually just put this somewhere else. Let’s just do one example for strawberry waffle.
24 00:03:25.560 ⇒ 00:03:26.280 Henry Zhao: I’ll bet.
25 00:03:26.790 ⇒ 00:03:27.809 Henry Zhao: We want to loom.
26 00:03:28.290 ⇒ 00:03:29.650 Henry Zhao: No, go away.
27 00:03:32.430 ⇒ 00:03:33.579 Henry Zhao: I want to make a loom.
28 00:03:35.870 ⇒ 00:03:38.379 Henry Zhao: So annoying. Sorry, one second.
29 00:03:40.540 ⇒ 00:03:54.310 Henry Zhao: So, like, we can talk now, and then I will, like, let you take a look at it, just see, like, what you think you can do, and I can help out, like, whatever you think you can’t do. Or we can just kind of divide up the opportunities. We can just divide the workload right now.
30 00:03:55.040 ⇒ 00:04:02.059 Henry Zhao: Okay, so orders… So I would look at this, and just kind of see what pattern it follows.
31 00:04:05.060 ⇒ 00:04:11.810 Henry Zhao: Yeah, but this is tough, like, it doesn’t really follow a pattern, and you’ve already seen that in your analysis. So I was kind of just asking ChatGPT, I was like.
32 00:04:12.820 ⇒ 00:04:15.609 Henry Zhao: Because I’ve never really done forecasting when it’s this erratic.
33 00:04:15.740 ⇒ 00:04:21.720 Henry Zhao: So I’ve kind of just been reading this. Like, predicting future inventory from erratic is common forecasting challenge. Let me copy this to you.
34 00:04:27.220 ⇒ 00:04:33.520 Amber Lin: How… what was the stuff that Byron sent over? I know… I think they have a high-level…
35 00:04:33.820 ⇒ 00:04:36.810 Amber Lin: Top-down forecasting?
36 00:04:40.150 ⇒ 00:04:44.280 Amber Lin: But… is that what’s in the inventory tracker?
37 00:04:46.280 ⇒ 00:04:50.380 Henry Zhao: Yeah, it’s just, like, weekly and monthly PO amounts.
38 00:04:52.720 ⇒ 00:04:57.900 Amber Lin: This is the actual amounts. I thought they had a high-level forecast.
39 00:04:58.040 ⇒ 00:05:00.449 Henry Zhao: They do, that’s this, that’s this right here.
40 00:05:01.530 ⇒ 00:05:04.529 Amber Lin: Oh, how… do you know how they did that?
41 00:05:04.530 ⇒ 00:05:07.030 Henry Zhao: No, I asked, so I’m waiting for a response on that.
42 00:05:07.570 ⇒ 00:05:14.440 Henry Zhao: But some of this makes sense to me. So, like, I’ve combined basically the last few years here, so I have 2025 here, and then.
43 00:05:14.440 ⇒ 00:05:14.850 Amber Lin: 2021.
44 00:05:14.850 ⇒ 00:05:17.260 Henry Zhao: 2025, 2024, 2023.
45 00:05:17.580 ⇒ 00:05:20.289 Henry Zhao: Like, one thing that’s easy about…
46 00:05:21.260 ⇒ 00:05:25.810 Henry Zhao: some of these, it’s like, it’s very consistent, like, this minus 29% is common in February.
47 00:05:26.240 ⇒ 00:05:33.980 Henry Zhao: But some things are very erratic. Like, this one is 135% March in 2025, 84% in 2024, 12% in 2023.
48 00:05:34.210 ⇒ 00:05:39.060 Henry Zhao: So I really need some help thinking about, like, how we want to forecast based on historical amounts.
49 00:05:39.570 ⇒ 00:05:47.339 Amber Lin: I see. This month-over-month change, that is, that’s the period-over-period, right?
50 00:05:47.340 ⇒ 00:05:48.270 Henry Zhao: Yeah,
51 00:05:48.470 ⇒ 00:05:50.339 Amber Lin: Oh.
52 00:05:50.340 ⇒ 00:05:52.289 Henry Zhao: We might want to look at year over year also.
53 00:05:54.160 ⇒ 00:06:00.910 Amber Lin: I see, the period change, that’s based on… not… not based on accepted, just based on requested, right?
54 00:06:02.400 ⇒ 00:06:05.070 Henry Zhao: So we can also look at year-over-year change.
55 00:06:05.210 ⇒ 00:06:07.780 Henry Zhao: So how do they calculate that?
56 00:06:09.800 ⇒ 00:06:12.849 Henry Zhao: I figure out which… what they’re doing for year-over-year change.
57 00:06:13.550 ⇒ 00:06:16.290 Henry Zhao: So, let me see if it’s this one.
58 00:06:21.360 ⇒ 00:06:23.689 Henry Zhao: 63, yup, that seems right, okay.
59 00:06:23.950 ⇒ 00:06:28.989 Henry Zhao: So it looks like his requested PO amounts… So imagine this…
60 00:06:34.520 ⇒ 00:06:36.719 Henry Zhao: Yeah, still very erratic, like…
61 00:06:40.640 ⇒ 00:06:41.799 Henry Zhao: You know what I mean? Like…
62 00:06:42.340 ⇒ 00:06:45.720 Henry Zhao: It’ll be hard to forecast, because there’s not really a set rules.
63 00:06:48.440 ⇒ 00:06:49.920 Amber Lin: I see…
64 00:06:56.160 ⇒ 00:07:06.160 Amber Lin: Let’s see, the high POs are… Let’s see… They’re usually high… Spiking…
65 00:07:07.100 ⇒ 00:07:12.419 Amber Lin: No, it doesn’t always spike in May. Like, in 2025…
66 00:07:12.850 ⇒ 00:07:20.310 Amber Lin: Actually, scroll up a little bit… 2025 overall… Let’s see…
67 00:07:25.970 ⇒ 00:07:27.200 Amber Lin: Huh.
68 00:07:30.310 ⇒ 00:07:31.310 Amber Lin: Hmm.
69 00:07:31.310 ⇒ 00:07:34.360 Henry Zhao: I think maybe if we do it by ASIN, it’ll be a little bit easier,
70 00:07:37.760 ⇒ 00:07:41.099 Henry Zhao: But I don’t know why, they just have this… Amount of data.
71 00:07:42.820 ⇒ 00:07:49.530 Amber Lin: What’s the… On the ASIN side, what if we… Hmm.
72 00:07:50.280 ⇒ 00:07:51.770 Henry Zhao: How much did they have here?
73 00:07:51.770 ⇒ 00:07:53.860 Amber Lin: Right now, are we thinking of…
74 00:07:54.260 ⇒ 00:08:00.499 Amber Lin: I guess right now, what we’re thinking is, what factors goes into that demand forecast?
75 00:08:03.120 ⇒ 00:08:11.350 Amber Lin: What was your experience before of building forecasts? Like, usually, what factors do you… do you consider?
76 00:08:13.620 ⇒ 00:08:16.340 Henry Zhao: I guess it depends on the industry and the product, right?
77 00:08:16.510 ⇒ 00:08:20.289 Henry Zhao: So… Yeah, I haven’t done really forecasting in a very long time.
78 00:08:20.890 ⇒ 00:08:22.269 Henry Zhao: That’s another issue.
79 00:08:24.110 ⇒ 00:08:26.930 Henry Zhao: And usually it’s, like, pretty steady growth.
80 00:08:27.400 ⇒ 00:08:35.330 Henry Zhao: It’s not like that erratic. And also, if you want data more than… more older than April 28th, all the data is in here in Mother Duck. I don’t know if you’re…
81 00:08:35.919 ⇒ 00:08:36.900 Henry Zhao: Can you…
82 00:08:36.909 ⇒ 00:08:37.759 Amber Lin: I didn’t…
83 00:08:37.760 ⇒ 00:08:38.129 Henry Zhao: I mean…
84 00:08:38.130 ⇒ 00:08:41.999 Amber Lin: I didn’t see it. Can you show me where… what…
85 00:08:42.000 ⇒ 00:08:45.620 Henry Zhao: You have to log in with the honeyster at brainforge.ai account.
86 00:08:45.620 ⇒ 00:08:47.759 Amber Lin: Oh, the… I see, that makes so much sense.
87 00:08:47.760 ⇒ 00:08:51.960 Henry Zhao: And then I’ve attached it, yeah, I’ve attached it to Acosta Data.
88 00:08:51.960 ⇒ 00:08:53.099 Amber Lin: Cool, okay.
89 00:08:53.770 ⇒ 00:09:02.569 Amber Lin: That’s very helpful. I’ll go from there, because I don’t know, like, my Excel is not that great.
90 00:09:02.870 ⇒ 00:09:07.059 Amber Lin: Let me log in to the other account.
91 00:09:19.380 ⇒ 00:09:24.040 Amber Lin: How much time do you have left today or tomorrow to work on this?
92 00:09:24.040 ⇒ 00:09:26.480 Henry Zhao: I can open up as much time as you need me to open up.
93 00:09:27.620 ⇒ 00:09:29.689 Henry Zhao: But I’m about to board a flight in 15 minutes.
94 00:09:30.740 ⇒ 00:09:39.709 Amber Lin: I see. I can’t… I probably can’t look at it today or until later today, because I need to finish up the slides, so…
95 00:09:39.860 ⇒ 00:09:41.700 Amber Lin: Are you on vacation?
96 00:09:42.280 ⇒ 00:09:44.219 Henry Zhao: Yeah, I’m supposed to be out the next 2 days.
97 00:09:44.220 ⇒ 00:09:47.789 Amber Lin: And you shouldn’t be working when you’re out on vacation.
98 00:09:47.920 ⇒ 00:09:51.680 Henry Zhao: Yeah, but I’m not in a good spot with Robert right now, so I kinda wanna… you know what I mean?
99 00:09:51.900 ⇒ 00:09:58.489 Amber Lin: Well, I… I think trust your ability is just, like, the analysis is a new type of work, so…
100 00:09:58.490 ⇒ 00:09:59.110 Henry Zhao: Yeah.
101 00:09:59.110 ⇒ 00:10:11.120 Amber Lin: It’s just on Eden, like, this part, I can try to see what I can do, especially if we both haven’t done forecasting before, like, we can try and see.
102 00:10:11.880 ⇒ 00:10:12.970 Amber Lin: Will we come up.
103 00:10:12.970 ⇒ 00:10:23.190 Henry Zhao: I’m gonna create a loom right now, and just kind of talk about what we’ve discussed right now, and the challenges, and we’ll see if Robert says anything. And then we can think about maybe again later tonight, or tomorrow morning, does that work?
104 00:10:23.330 ⇒ 00:10:30.200 Amber Lin: Sure. I guess one last thing. What granularity are we forecasting on? I’m gonna just do my month.
105 00:10:30.790 ⇒ 00:10:33.059 Amber Lin: By month and by ASIN?
106 00:10:33.750 ⇒ 00:10:48.249 Henry Zhao: Yeah, by month and by ASIN, and ASIN is by week, so ASIN will allow me to, like, if I forecast this for the next, like, 26 weeks, I can then combine it into months and see if it roughly adds up to our monthly forecast. So, in other words, it’s like another signal, you know what I mean?
107 00:10:48.460 ⇒ 00:10:53.489 Amber Lin: Okay, so we’re doing… We’re doing weekly ASINs, rolling it up.
108 00:10:53.490 ⇒ 00:10:53.910 Henry Zhao: decades.
109 00:10:53.910 ⇒ 00:10:58.430 Amber Lin: That up into months, overall.
110 00:10:58.570 ⇒ 00:11:00.120 Henry Zhao: Yeah. Cool.
111 00:11:00.120 ⇒ 00:11:04.700 Amber Lin: Okay, I might do a roll-up into categories, because I think that… that might…
112 00:11:04.700 ⇒ 00:11:05.200 Henry Zhao: investing.
113 00:11:05.200 ⇒ 00:11:21.640 Amber Lin: That might work the best. But I’ll try to write down what factors can go into there from simple to complex, and I don’t think we can do too complex models right now, and I will let you know. I don’t think we’re doing any machine learnings right now, right?
114 00:11:21.640 ⇒ 00:11:22.469 Henry Zhao: No, I know.
115 00:11:22.470 ⇒ 00:11:26.299 Amber Lin: Excel, no models, okay.
116 00:11:26.520 ⇒ 00:11:27.250 Amber Lin: That makes sense.
117 00:11:27.250 ⇒ 00:11:29.440 Henry Zhao: Yeah. Cool.
118 00:11:29.440 ⇒ 00:11:34.150 Amber Lin: Yeah, that’s good. Board your flight, have fun, don’t think about it today, not much we can do.
119 00:11:34.150 ⇒ 00:11:37.920 Henry Zhao: Yeah, okay, I’ll make the loom and just kind of share where we’re at. Cool.
120 00:11:38.680 ⇒ 00:11:39.340 Henry Zhao: Okay.
121 00:11:40.460 ⇒ 00:11:41.220 Henry Zhao: Thank you.
122 00:11:41.850 ⇒ 00:11:44.170 Amber Lin: Thanks. Have fun on vacation. Bye!
123 00:11:44.170 ⇒ 00:11:44.860 Henry Zhao: Bye-bye.