Meeting Title: Brainforge Final Interview Date: 2026-03-30 Meeting participants: Anthony Chukwuemeka Orji, Greg Stoutenburg, Robert Tseng, Amber Lin, Anthony
WEBVTT
1 00:04:38.630 ⇒ 00:04:40.280 Anthony Chukwuemeka Orji: Hi, Greg, good morning!
2 00:04:40.520 ⇒ 00:04:42.159 Greg Stoutenburg: Hey, good morning, how are you?
3 00:04:42.290 ⇒ 00:04:43.650 Anthony Chukwuemeka Orji: I’m doing well.
4 00:04:44.360 ⇒ 00:04:45.579 Anthony Chukwuemeka Orji: How about you?
5 00:04:46.100 ⇒ 00:04:47.290 Greg Stoutenburg: Doing alright!
6 00:04:47.620 ⇒ 00:04:49.140 Anthony Chukwuemeka Orji: I hope you enjoyed your weekend.
7 00:04:49.280 ⇒ 00:04:52.520 Greg Stoutenburg: Yeah, you too. I’m yawning. It was too fun.
8 00:04:55.620 ⇒ 00:05:03.949 Anthony Chukwuemeka Orji: Yeah, I woke up with a hiccup. I’m surprised he already stopped. I was battling for you to stop before I started the interview.
9 00:05:03.950 ⇒ 00:05:05.330 Greg Stoutenburg: Now’s not the time!
10 00:05:05.330 ⇒ 00:05:08.710 Anthony Chukwuemeka Orji: I don’t know the time. Yeah.
11 00:05:08.710 ⇒ 00:05:09.810 Greg Stoutenburg: That’s funny.
12 00:05:10.670 ⇒ 00:05:13.619 Greg Stoutenburg: Alright, so we’re just waiting for the rest of the team to hop on, then.
13 00:05:13.620 ⇒ 00:05:14.400 Anthony Chukwuemeka Orji: Okay.
14 00:05:14.640 ⇒ 00:05:15.470 Greg Stoutenburg: Hey, Robert.
15 00:05:16.150 ⇒ 00:05:17.449 Robert Tseng: Hey, FD. Hey, Greg.
16 00:05:18.930 ⇒ 00:05:19.740 Anthony Chukwuemeka Orji: Hello?
17 00:05:21.640 ⇒ 00:05:22.770 Anthony Chukwuemeka Orji: Can you see me?
18 00:05:23.200 ⇒ 00:05:24.060 Robert Tseng: Yeah, I can see you.
19 00:05:24.310 ⇒ 00:05:24.970 Anthony Chukwuemeka Orji: Okay.
20 00:05:27.140 ⇒ 00:05:28.150 Greg Stoutenburg: Alright.
21 00:05:29.510 ⇒ 00:05:32.240 Greg Stoutenburg: Cool, so just waiting for Amber, I think.
22 00:05:32.970 ⇒ 00:05:34.990 Greg Stoutenburg: Yeah, let me ping her.
23 00:05:57.770 ⇒ 00:05:59.419 Robert Tseng: Well, I think we can get started.
24 00:05:59.420 ⇒ 00:06:00.399 Greg Stoutenburg: Yeah, let’s just get started.
25 00:06:00.910 ⇒ 00:06:01.900 Greg Stoutenburg: Let’s get started.
26 00:06:02.380 ⇒ 00:06:06.680 Robert Tseng: Yeah, so I guess, Anthony, you have a… you have a presentation for us today?
27 00:06:07.000 ⇒ 00:06:11.230 Anthony Chukwuemeka Orji: Correct. I do have a presentation today.
28 00:06:11.710 ⇒ 00:06:17.069 Anthony Chukwuemeka Orji: First of all, I really appreciate the opportunity to go through this exercise.
29 00:06:17.190 ⇒ 00:06:21.150 Anthony Chukwuemeka Orji: It was really eye-opening working on that presentation.
30 00:06:21.340 ⇒ 00:06:27.540 Anthony Chukwuemeka Orji: I know… I’m assuming one of the things that was tested was a test for time.
31 00:06:27.740 ⇒ 00:06:40.390 Anthony Chukwuemeka Orji: And I put in intermission before the 23 hours mark. However, I did some modification, to extend further the analysis, and I hope to share them today, if time permits.
32 00:06:41.700 ⇒ 00:06:56.840 Robert Tseng: Sure, yeah, I mean, just to kind of give some structure to this call, since we only have 45 minutes together, yeah, I think we’ll give you… we’ll probably give you up to 10 minutes to kind of present your findings. You don’t have to walk through every slide, just kind of talk through it.
33 00:06:56.940 ⇒ 00:07:03.429 Robert Tseng: as if you were presenting to, kind of, like, your… a client. And so…
34 00:07:03.650 ⇒ 00:07:17.710 Robert Tseng: Yeah, and then we’ll do some Q&A off of your deck. Hopefully we’ll wrap that up within 15 minutes, 20 minutes, and then we will… we’re just… I mean, I think we really want to get a chance to, ask you more behavioral questions as well.
35 00:07:17.740 ⇒ 00:07:26.310 Robert Tseng: I think, like, the team so far has been really impressed with, kind of, your relevant experience, and also, you know, just your ability to
36 00:07:26.450 ⇒ 00:07:28.670 Robert Tseng: kind of… share…
37 00:07:28.810 ⇒ 00:07:45.550 Robert Tseng: your, I guess, communication clarity, I guess, and so I think there’s some other things that we’d like to evaluate on this call, and I think the exercise would be a good way to kind of get into it, but then we also would like to, you know, give you some time to ask us some questions at the end, too.
38 00:07:45.820 ⇒ 00:07:58.429 Anthony Chukwuemeka Orji: Okay, alright, I would just, you know, see if I could get in, like, 5 minutes to chip in the remaining extra validation steps I did, other than the deck. But I’m gonna start away if you, you know, permit.
39 00:07:58.800 ⇒ 00:07:59.700 Robert Tseng: Yeah, go ahead.
40 00:08:00.160 ⇒ 00:08:02.589 Anthony Chukwuemeka Orji: Okay, my name is Anthony.
41 00:08:02.810 ⇒ 00:08:09.169 Anthony Chukwuemeka Orji: And this is my presentation for the new vertical analytics.
42 00:08:09.730 ⇒ 00:08:16.789 Anthony Chukwuemeka Orji: What I did was I analyzed about… 13,000, deliveries across,
43 00:08:16.930 ⇒ 00:08:23.850 Anthony Chukwuemeka Orji: First of all, I’m going to share my screen, so that I can show the presentation.
44 00:08:24.170 ⇒ 00:08:32.370 Anthony Chukwuemeka Orji: Well, how is the share button? The thing that confuses me here?
45 00:08:34.350 ⇒ 00:08:36.040 Greg Stoutenburg: It should be the one with the up arrow.
46 00:08:37.490 ⇒ 00:08:39.880 Greg Stoutenburg: It’s like a box with an arrow pointing up in it.
47 00:08:41.409 ⇒ 00:08:42.949 Anthony Chukwuemeka Orji: Because I, I, I…
48 00:08:43.059 ⇒ 00:08:47.199 Anthony Chukwuemeka Orji: I don’t see it here, that’s why. I don’t know if I have permissions to share.
49 00:08:48.049 ⇒ 00:08:49.089 Anthony Chukwuemeka Orji: Okay, I do.
50 00:08:50.529 ⇒ 00:08:52.359 Anthony Chukwuemeka Orji: Okay…
51 00:08:56.349 ⇒ 00:08:58.069 Anthony Chukwuemeka Orji: Okay, can you see my screen?
52 00:09:03.279 ⇒ 00:09:04.389 Anthony Chukwuemeka Orji: Can you see my screen?
53 00:09:05.520 ⇒ 00:09:07.789 Robert Tseng: Not yet. Oh yeah, okay, yeah, it just came up.
54 00:09:08.150 ⇒ 00:09:19.250 Anthony Chukwuemeka Orji: Okay, alright. Okay, I’m gonna start again. My name is Anthony, and this is my presentation for the Brainforge, interview exercise, New Vertical Analytics.
55 00:09:19.350 ⇒ 00:09:38.189 Anthony Chukwuemeka Orji: So what I did was, analyzed about 13,000 deliveries across Dashmat and Goswee, so as to identify where operational breakdowns are happening. And most importantly, so that I could find out which levers were actually, moving, the performance.
56 00:09:38.300 ⇒ 00:09:42.340 Anthony Chukwuemeka Orji: I walked through the key findings, specifically 3 of them.
57 00:09:42.460 ⇒ 00:09:47.819 Anthony Chukwuemeka Orji: The business impacts and the actions that are prioritized based on my findings.
58 00:09:53.630 ⇒ 00:09:56.759 Anthony Chukwuemeka Orji: Okay, at a high level.
59 00:09:57.000 ⇒ 00:09:59.140 Anthony Chukwuemeka Orji: Like I said, there are 3 core issues.
60 00:09:59.360 ⇒ 00:10:10.580 Anthony Chukwuemeka Orji: Late delivery, missing item, And, late delivery, missing item, and, Wednesday being a very bad weekday.
61 00:10:10.840 ⇒ 00:10:16.859 Anthony Chukwuemeka Orji: The three core issues, late delivery is primarily You know, by national acceptance.
62 00:10:17.080 ⇒ 00:10:19.770 Anthony Chukwuemeka Orji: The lease, and not by store readiness.
63 00:10:20.180 ⇒ 00:10:27.680 Anthony Chukwuemeka Orji: Whereas the missing items are concentrated in grocery, and increases sharply with order size.
64 00:10:28.040 ⇒ 00:10:33.100 Anthony Chukwuemeka Orji: Performance degrades midweek, particularly on Wednesday.
65 00:10:33.240 ⇒ 00:10:36.399 Anthony Chukwuemeka Orji: Across both delivery and missing item.
66 00:10:36.720 ⇒ 00:10:40.980 Anthony Chukwuemeka Orji: Based on all of that, what I’ll do is to prioritize three actions.
67 00:10:41.100 ⇒ 00:10:48.980 Anthony Chukwuemeka Orji: I will introduce, size, or other size thresholds, so as to reduce missing items.
68 00:10:49.110 ⇒ 00:10:53.340 Anthony Chukwuemeka Orji: I will type in Dasha Acceptance Service Level Agreement.
69 00:10:53.450 ⇒ 00:10:57.379 Anthony Chukwuemeka Orji: And then I’ll focus operational fixes on Ghostway 2.
70 00:10:57.810 ⇒ 00:11:03.560 Anthony Chukwuemeka Orji: Where delays and cancellations concentrated and are more frequent?
71 00:11:09.210 ⇒ 00:11:15.270 Anthony Chukwuemeka Orji: I’ll start with assumptions, and then data context, and then walk through
72 00:11:15.440 ⇒ 00:11:20.969 Anthony Chukwuemeka Orji: the three key findings, I’ll give recommendations, and I’ll give next steps.
73 00:11:25.240 ⇒ 00:11:42.150 Anthony Chukwuemeka Orji: Now, this is an overview and a methodology. This is a one-month data set, so I focused on directional insights alone, rather than, you know, long-term trends. I wanted to see specifically what was driving operational issues.
74 00:11:42.320 ⇒ 00:11:57.709 Anthony Chukwuemeka Orji: Also, I treated, the item price as a proxy for basket value, so I didn’t, call it item price, because I believe, it is a basket value rather than the item price, because it is not… it does not reflect the full revenue.
75 00:11:57.830 ⇒ 00:12:05.230 Anthony Chukwuemeka Orji: Since fees and discounts aren’t, included. Another thing to point out was that the data say that,
76 00:12:05.440 ⇒ 00:12:21.970 Anthony Chukwuemeka Orji: The prices were in cents, but upon confirmation doing transformation and wrangling, I did some averages and realized that it couldn’t be in cents, so I moved forward with the assumption that it was in dollars, based on how the data structure, is.
77 00:12:25.940 ⇒ 00:12:43.579 Anthony Chukwuemeka Orji: Now, what this slide shows us is that the averages actually mask the story, and if we depend on just the averages, we may not get the real insight. Because when I started the manual transformation and wrangling, trying to do, I saw the averages being covered, and there was a confusion between CLAT and D2R.
78 00:12:43.990 ⇒ 00:12:48.880 Anthony Chukwuemeka Orji: So the averages, it marks the story, and the breakdowns are what matters.
79 00:12:48.940 ⇒ 00:12:59.259 Anthony Chukwuemeka Orji: Now, this slide summarizes the overall performance. About 13,000 deliveries, 6.2% of them are late deliveries.
80 00:12:59.270 ⇒ 00:13:16.559 Anthony Chukwuemeka Orji: And 4.8… 6.2% of them are item missing, and 4.8% of them are late deliveries. Now, what’s important is not the averages, but how these averages break down operationally, and that’s what I’ll be working through in the next slides.
81 00:13:21.440 ⇒ 00:13:24.279 Anthony Chukwuemeka Orji: Now, the first finding is what this slide shows.
82 00:13:25.120 ⇒ 00:13:29.879 Anthony Chukwuemeka Orji: The first finding is that order size is the number one predictor.
83 00:13:30.250 ⇒ 00:13:46.699 Anthony Chukwuemeka Orji: And 8 items is the inflection point. When I was doing the analysis, I realized that missing items are not inuly distributed. You know, they are actually concentrated in the gross fee category. Now, the key inflection is the order size.
84 00:13:46.860 ⇒ 00:13:52.420 Anthony Chukwuemeka Orji: And once orders exceed around 8 items, The missing weight increases sharply.
85 00:13:52.460 ⇒ 00:14:06.779 Anthony Chukwuemeka Orji: Now, this suggests, peaking complexity and operational overload, not random errors. I associated this with myself as a customer using DoorDash, and I’ve, you know, encountered a whole lot of this issue.
86 00:14:06.800 ⇒ 00:14:16.519 Anthony Chukwuemeka Orji: Most of the time. When I have orders going above a certain amount, it is definitely… I’ll definitely be having missing items, or I’ll definitely need to substitute.
87 00:14:16.590 ⇒ 00:14:20.260 Anthony Chukwuemeka Orji: So that was one of the assumptions that drew me closer to that.
88 00:14:20.690 ⇒ 00:14:24.749 Anthony Chukwuemeka Orji: When translated into dollars, we see that grocery store has
89 00:14:24.980 ⇒ 00:14:28.670 Anthony Chukwuemeka Orji: Roughly 14% to 16% of basket value.
90 00:14:28.890 ⇒ 00:14:32.119 Anthony Chukwuemeka Orji: At risk, while Dashmat is near, zero.
91 00:14:32.980 ⇒ 00:14:50.780 Anthony Chukwuemeka Orji: You know, what this means is that the right fix is a targeted control at large goes to others. That is why I spoke about creating the threshold, and from the data, I realized the threshold is from 8 items, as one of the recommendations. So I’m going to move to the next slide.
92 00:14:52.920 ⇒ 00:14:55.679 Anthony Chukwuemeka Orji: Now, the next slides talk about the second finding.
93 00:14:56.050 ⇒ 00:15:00.559 Anthony Chukwuemeka Orji: Which is, Dispatch delay, not distance.
94 00:15:00.890 ⇒ 00:15:04.390 Anthony Chukwuemeka Orji: And clots is the liver.
95 00:15:04.870 ⇒ 00:15:12.989 Anthony Chukwuemeka Orji: Now, class being the time between when the order is placed, and when the driver or the dasher accepts the order.
96 00:15:13.160 ⇒ 00:15:16.940 Anthony Chukwuemeka Orji: It’s the highest level, because if the class is already late.
97 00:15:17.250 ⇒ 00:15:32.970 Anthony Chukwuemeka Orji: The D2R, the distance when the driver gets to the store, it compounds, and at the end of the day, it compounds the late delivery. Now, the strongest driver of late delivery is the dash acceptance, not the delay, which is the class.
98 00:15:33.320 ⇒ 00:15:44.240 Anthony Chukwuemeka Orji: Once class exceeds around 7 minutes, we realize that late delivery more than doubles, and at 15 plus minutes, it exceeds, you know.
99 00:15:44.700 ⇒ 00:15:46.920 Anthony Chukwuemeka Orji: Exceeds about 28%.
100 00:15:47.440 ⇒ 00:15:55.690 Anthony Chukwuemeka Orji: D2L contributes, but it is secondary, and it is fundamentally a dispatch timing problem, not a distance problem.
101 00:15:56.280 ⇒ 00:15:58.769 Anthony Chukwuemeka Orji: What this does is that it shifts ownership.
102 00:15:58.930 ⇒ 00:16:08.610 Anthony Chukwuemeka Orji: From store operations towards dispatch and data behavior, allowing stakeholders to concentrate towards that solution.
103 00:16:09.400 ⇒ 00:16:15.719 Anthony Chukwuemeka Orji: Now, the highest level fix here is improving acceptance speed, not the routing.
104 00:16:16.290 ⇒ 00:16:18.550 Anthony Chukwuemeka Orji: Now, I’m going to move on to the third finding.
105 00:16:19.470 ⇒ 00:16:22.779 Anthony Chukwuemeka Orji: which is Wednesday, as the worst operational day.
106 00:16:23.870 ⇒ 00:16:32.580 Anthony Chukwuemeka Orji: And it breaks both marks in every metric. It influences the late delivery, also influences the missing rates.
107 00:16:32.800 ⇒ 00:16:34.560 Anthony Chukwuemeka Orji: The missing weights.
108 00:16:34.810 ⇒ 00:16:41.159 Anthony Chukwuemeka Orji: When it is the worst performing day, it has the highest lead delivery, the highest mission rate, simultaneously.
109 00:16:41.400 ⇒ 00:16:44.840 Anthony Chukwuemeka Orji: What it indicates is… A coordination gap.
110 00:16:45.370 ⇒ 00:16:51.159 Anthony Chukwuemeka Orji: across store readiness, Dasha availability and demand on that specific day.
111 00:16:51.510 ⇒ 00:16:54.159 Anthony Chukwuemeka Orji: And this is not an isolated problem, because
112 00:16:54.660 ⇒ 00:17:01.380 Anthony Chukwuemeka Orji: For it to have such effect on those two findings, it means it is not isolated, and it is a system
113 00:17:01.600 ⇒ 00:17:04.450 Anthony Chukwuemeka Orji: Or process synchronization issue.
114 00:17:06.690 ⇒ 00:17:09.539 Anthony Chukwuemeka Orji: Now, I’m going to move on to the recommendations.
115 00:17:11.589 ⇒ 00:17:21.380 Anthony Chukwuemeka Orji: From the findings, The recommendations would be to implement other size threshold alerts for large grocery orders.
116 00:17:21.790 ⇒ 00:17:24.889 Anthony Chukwuemeka Orji: I will introduce thresholds or batching.
117 00:17:25.099 ⇒ 00:17:29.510 Anthony Chukwuemeka Orji: For large grocery orders, to reduce the picking errors.
118 00:17:29.740 ⇒ 00:17:36.300 Anthony Chukwuemeka Orji: 75% of orders with 13 plus items have at least one missing item.
119 00:17:36.530 ⇒ 00:17:45.209 Anthony Chukwuemeka Orji: And grocery is about 14.2%, Y-Dashmat is 0.2%, which is the lowest. That’s a 13.9 addressable gap.
120 00:17:46.040 ⇒ 00:17:49.460 Anthony Chukwuemeka Orji: The open question is whether the app
121 00:17:49.770 ⇒ 00:17:57.640 Anthony Chukwuemeka Orji: The Dasha app supports in-house scoring and scanning, and whether we can separate Dasha misses
122 00:17:57.890 ⇒ 00:18:01.660 Anthony Chukwuemeka Orji: From 2 stock outs with store inventory runs.
123 00:18:04.310 ⇒ 00:18:06.910 Anthony Chukwuemeka Orji: Now, I’m gonna move to the next, recommendation.
124 00:18:10.210 ⇒ 00:18:19.549 Anthony Chukwuemeka Orji: The next recommendation would be to tighten Dasha Acceptance SLE, and this is based on this 7-minute threshold that was realized.
125 00:18:19.930 ⇒ 00:18:27.269 Anthony Chukwuemeka Orji: I would tighten the actual acceptance expectations, especially in the early hours, between 3 a.m. and 6 AM,
126 00:18:27.380 ⇒ 00:18:30.300 Anthony Chukwuemeka Orji: Which records the most late deliveries.
127 00:18:30.780 ⇒ 00:18:37.149 Anthony Chukwuemeka Orji: Going from, at the 7-minute threshold.
128 00:18:37.440 ⇒ 00:18:41.680 Anthony Chukwuemeka Orji: Late weight stays under 2.5%. Therefore.
129 00:18:41.810 ⇒ 00:18:45.580 Anthony Chukwuemeka Orji: I’m gonna, implement a 7,
130 00:18:46.250 ⇒ 00:18:48.429 Anthony Chukwuemeka Orji: A 7 minutes class threshold.
131 00:18:48.650 ⇒ 00:18:52.179 Anthony Chukwuemeka Orji: And added to the service level agreement, for the dashes.
132 00:18:52.730 ⇒ 00:18:57.809 Anthony Chukwuemeka Orji: I’m gonna move on to the next one, which is the final recommendation.
133 00:18:58.590 ⇒ 00:19:05.440 Anthony Chukwuemeka Orji: Which would be to… Target grocery, too. In a very proportional manner… manner.
134 00:19:05.760 ⇒ 00:19:13.089 Anthony Chukwuemeka Orji: Now, glossary 2… House routing and operational issues, clearly, from the analysis.
135 00:19:13.330 ⇒ 00:19:17.159 Anthony Chukwuemeka Orji: So, what it needs is routing and operational improvement.
136 00:19:17.350 ⇒ 00:19:19.540 Anthony Chukwuemeka Orji: I would do operational fixes.
137 00:19:19.650 ⇒ 00:19:23.370 Anthony Chukwuemeka Orji: And these operational fixes need to be targeted uniformly.
138 00:19:23.690 ⇒ 00:19:29.290 Anthony Chukwuemeka Orji: Goss’s D2R of 8.07 minutes is 3 times that of Dashmat’s.
139 00:19:29.580 ⇒ 00:19:36.469 Anthony Chukwuemeka Orji: And that 5 minutes gap, You know, it’s estimated to cut late rates from 1 to 2 points.
140 00:19:36.950 ⇒ 00:19:42.810 Anthony Chukwuemeka Orji: Cancel rate is also dropping, will also drop from 3.1% to 1.1%.
141 00:19:43.060 ⇒ 00:19:48.450 Anthony Chukwuemeka Orji: The next step… From this analysis.
142 00:19:49.550 ⇒ 00:19:54.220 Anthony Chukwuemeka Orji: would be to do an A-B testing experimentation.
143 00:19:54.420 ⇒ 00:20:07.829 Anthony Chukwuemeka Orji: to confirm whether the attributes… attribution that I’ve given to CLAT is really correct and not just a mere correlation. I’m going to stop there for now to see if there are any questions before I move further.
144 00:20:10.010 ⇒ 00:20:12.490 Anthony Chukwuemeka Orji: Okay, I’m gonna move further.
145 00:20:12.750 ⇒ 00:20:26.000 Robert Tseng: Yeah, actually, Anthony, I think… I think we’re good on the… on the presentation, so we can just… we can just pause there. We’ll ask a couple questions, and then we… I think we can… we can, sorry, it’s been… it’s been 10 minutes on… on the presentation, so we don’t… we don’t need to hear you go through the rest of it.
146 00:20:26.390 ⇒ 00:20:26.720 Anthony Chukwuemeka Orji: Okay.
147 00:20:26.720 ⇒ 00:20:42.500 Robert Tseng: Yeah, so I guess, my question would be, yeah, when you’re kind of putting together, some recommendations here, I’m seeing that you have, like, high feasibility for which one, but if I were the client, and I were to… and, I mean, I’m the client, so I’m gonna ask, like.
148 00:20:42.550 ⇒ 00:20:47.859 Robert Tseng: Which one of these… which one of these recommendations is, the most high impact, and why?
149 00:20:48.410 ⇒ 00:20:54.140 Anthony Chukwuemeka Orji: Now, I’m gonna show exactly why I wanted to show that. I’m gonna… is it okay if I open this up?
150 00:20:54.310 ⇒ 00:20:58.089 Anthony Chukwuemeka Orji: Okay, so this is what I wanted to show.
151 00:21:00.470 ⇒ 00:21:07.310 Anthony Chukwuemeka Orji: So, the recommendation that is high impact, and why?
152 00:21:07.480 ⇒ 00:21:12.790 Anthony Chukwuemeka Orji: would be the easiest win from all three recommendations. Would be…
153 00:21:13.580 ⇒ 00:21:26.420 Anthony Chukwuemeka Orji: the other size, the threshold for the other size. I ran… I ran a power curve analysis on all prospective A-B tests using a simulation. I don’t know if you can see my screen.
154 00:21:27.200 ⇒ 00:21:28.059 Greg Stoutenburg: We see that.
155 00:21:28.060 ⇒ 00:21:29.719 Anthony Chukwuemeka Orji: You can see the power curve.
156 00:21:30.080 ⇒ 00:21:30.780 Greg Stoutenburg: Yep.
157 00:21:30.780 ⇒ 00:21:37.330 Anthony Chukwuemeka Orji: Okay, now, from the power curve analysis, we can see that the other size allowed threshold.
158 00:21:37.900 ⇒ 00:21:45.750 Anthony Chukwuemeka Orji: would easily supersede the 80% power threshold, even from the start, at n is equal to 51.
159 00:21:46.150 ⇒ 00:21:57.520 Anthony Chukwuemeka Orji: Meaning that with small size data, we can easily see the order size allowed threshold bringing expected results.
160 00:21:57.960 ⇒ 00:22:05.829 Anthony Chukwuemeka Orji: Within just, 4 days, within just 4 days, I also have this here, which shows that
161 00:22:08.600 ⇒ 00:22:11.530 Anthony Chukwuemeka Orji: But, within just 4 days.
162 00:22:11.680 ⇒ 00:22:19.100 Anthony Chukwuemeka Orji: And so, I believe that answers the question. The other size alert would be the first recommendation that would give the highest effects.
163 00:22:21.150 ⇒ 00:22:36.009 Greg Stoutenburg: Anthony, just to clarify here, I mean, we’re sort of in the discussion period, so just to clarify here, what exactly would the order alert do? Like, when you think about what the mechanism is for how the order alert is going to impact Dasher performance, can you just say a little bit more about that?
164 00:22:36.480 ⇒ 00:22:43.340 Anthony Chukwuemeka Orji: Now, the order alerts, when we realize that when item supersedes
165 00:22:43.470 ⇒ 00:22:47.430 Anthony Chukwuemeka Orji: 8 items. We have dashes…
166 00:22:47.620 ⇒ 00:23:01.070 Anthony Chukwuemeka Orji: having incorrect or missing reports, or missing items. When we create that threshold and add it to the service level agreement, we create some type of an alert to the Dasher
167 00:23:01.270 ⇒ 00:23:14.939 Anthony Chukwuemeka Orji: You know, to keep them aware that they need to get all the items correct. That is one. With the threshold, with the other size a lot, we can also use that to assign
168 00:23:15.270 ⇒ 00:23:16.560 Anthony Chukwuemeka Orji: To the top.
169 00:23:29.510 ⇒ 00:23:33.599 Robert Tseng: Anthony, you’re, your audio cut out, we can’t hear you anymore.
170 00:25:27.000 ⇒ 00:25:28.490 Anthony: Hi, can you hear me?
171 00:25:28.770 ⇒ 00:25:29.280 Greg Stoutenburg: Hey!
172 00:25:29.280 ⇒ 00:25:30.390 Robert Tseng: We can hear you.
173 00:25:30.390 ⇒ 00:25:36.940 Anthony: I don’t know what happened, I’m sorry about that, my computer just went off. That was… that is very strange.
174 00:25:37.250 ⇒ 00:25:41.859 Greg Stoutenburg: Yeah, it looks like you’re signed in on another window, but it’s just, like, it’s just unresponsive, there’s nothing there.
175 00:25:42.370 ⇒ 00:25:56.410 Anthony: it just… it just went off. That’s very strange. Apologies. If I could continue, I was speaking about the order size threshold for, the dashas. So the first thing would be to… when I put the order size threshold.
176 00:25:56.410 ⇒ 00:26:04.600 Anthony: What that will help to do, it will help us assign those orders that go past the 8-item threshold to top dashers.
177 00:26:05.690 ⇒ 00:26:13.380 Anthony: Or, if it is assigned to anybody who’s not a top dasher, there’s probably gonna be, like, a notification flag to the dasher.
178 00:26:13.540 ⇒ 00:26:15.569 Anthony: Reminding the data to…
179 00:26:15.740 ⇒ 00:26:31.669 Anthony: optimally and effectively make sure there are no missing items. But I think the first suggestion would be using that 8 size, threshold to assign to top dashers who can effectively, make sure there are no missing items.
180 00:26:32.990 ⇒ 00:26:38.480 Robert Tseng: Okay, yeah, let’s just, let’s just pause there. So, I mean, I guess we can give you some feedback, like,
181 00:26:38.610 ⇒ 00:26:51.130 Robert Tseng: On the… yeah, I think your analysis to kind of set the threshold for A-plus items is interesting. I mean, I’m sure there’s some… I mean, I’m sure your analysis kind of shows that. I think that it’s…
182 00:26:51.130 ⇒ 00:26:51.660 Anthony: It’s…
183 00:26:51.660 ⇒ 00:26:55.530 Robert Tseng: I mean, I don’t feel like it moves the needle that much. I feel like…
184 00:26:56.000 ⇒ 00:27:15.650 Robert Tseng: I mean, there’s other… there’s other reasons why, dashers were not… are not, like, fulfilling, orders on time. Like, I… or, like, not… not filling… fulfilling orders accurately. So, things like, what’s the probability of missing… missing an item? Also, like.
185 00:27:15.650 ⇒ 00:27:28.269 Robert Tseng: you know, there… there could be just a miscommunication between systems, between the point of… point of sale system and the app and the supply side. Like, I feel like we zeroed in on just, like, this sending alerts to dashers, and it’s…
186 00:27:28.400 ⇒ 00:27:44.679 Robert Tseng: I mean, you can justify it, sure, I just… I just feel like it was too much of a jump to make off of… as, like, to bank as your number one recommendation, just to send dash or alerts. So, also, like, I think there’s a more succinct way to communicate that, and
187 00:27:44.680 ⇒ 00:27:57.250 Robert Tseng: you know, as a client, you’d probably want to… the client would probably debate with you on whether that’s really going to move the needle in their business. So, I guess, like, I mean, we could… that’s… that’s probably… that’s my response to the,
188 00:27:57.860 ⇒ 00:28:04.809 Robert Tseng: Do the recommendation in the context of this, deck, but just kind of, like, zooming out from this whole exercise.
189 00:28:04.810 ⇒ 00:28:29.670 Robert Tseng: Yeah, sounds like you’re really thorough in kind of going deep, testing different models and approaches to validate, kind of, a very specific kind of technical threshold. So, I think your technical bar is very high, and I think there’s, yeah, there’s a lot you can do there. I would say for this role specifically, as, like, a senior strategist, you’re often presenting to non-technical stakeholders how you kind of cover all of your bases in a client
190 00:28:29.670 ⇒ 00:28:44.529 Robert Tseng: presentation, I think is really important. So, I think for that reason, like, I’m not… I mean, I personally wouldn’t think that you’re the best fit for this role. So, I mean, I’ll leave some time for Greg to kind of give any feedback.
191 00:28:44.530 ⇒ 00:28:51.290 Robert Tseng: But I think that’s… that’s kind of my… that’s my… that’s my… that’s my decision at… at that… at this point.
192 00:28:51.290 ⇒ 00:28:58.339 Anthony: But I didn’t hear you well. You, you, however, I didn’t hear you well. You said,
193 00:28:58.500 ⇒ 00:29:07.679 Anthony: For the way I translated the business problem to solutions, you don’t think I’m the best person for this role? Is that… I didn’t hear that clearly.
194 00:29:08.180 ⇒ 00:29:13.349 Robert Tseng: Yeah, that’s… that’s not what I said, but, I just… the way that you…
195 00:29:13.950 ⇒ 00:29:33.099 Robert Tseng: communicated your insight, you know, we kind of tried to focus you on what’s the number one recommendation you would make. You set this threshold of 8 plus orders. I think there are other considerations, like, I think you went too deep too fast, and didn’t really, like, kind of give… like, there’s other considerations that I feel like you just kind of.
196 00:29:33.100 ⇒ 00:29:34.310 Anthony: Right.
197 00:29:34.310 ⇒ 00:29:35.380 Robert Tseng: you didn’t address.
198 00:29:35.630 ⇒ 00:29:46.409 Anthony: Right. There are other considerations for the missing items. For instance, grocery 2 is another major factor. The grocery 2 store is another major factor that drives
199 00:29:46.520 ⇒ 00:30:02.960 Anthony: are missing items. If you look at it, a whole lot of the missing items are actually concentrated on that. And that is why, the Ghost V2 is also a test in the power curve that needs to be simulated. And the approach would be to
200 00:30:03.210 ⇒ 00:30:08.880 Anthony: Fix the operational issues in God’s V2, so…
201 00:30:09.010 ⇒ 00:30:25.780 Anthony: So that was also, I wasn’t… I wasn’t saying the recommendation is just eat, you know, others. I was just giving that as one of the recommendations, and it was not only the driving factor. But I respect your, your, decision.
202 00:30:27.780 ⇒ 00:30:39.819 Robert Tseng: Yeah, I know that there’s a lot that you don’t get to cover in these kind of presentations, and it’s kind of like a knowing your audience, as well, so, you know, maybe we should make it clear in the…
203 00:30:39.820 ⇒ 00:30:53.179 Robert Tseng: instructions as well, but, you know, if you’re presenting to a VP-level, C-level client, they’re not gonna care so much about all of the… the how… the how. Like, I don’t need to see all the models you use, I think I would only really care about the
204 00:30:53.240 ⇒ 00:31:16.479 Robert Tseng: recommendation, and we would be more of a discussion of me, like, asking you, did you consider this? Did you consider that? And you would have to kind of defend that in a short period of time. So, I think what you presented felt more like a technical presentation. You’re convening it to a… kind of like a, you know, a very senior technical person, which is great in… in… in a particular role. I just don’t think it’s… it’s this role.
205 00:31:19.330 ⇒ 00:31:19.920 Anthony: Wait.
206 00:31:20.060 ⇒ 00:31:20.890 Anthony: Okay.
207 00:31:21.480 ⇒ 00:31:22.130 Robert Tseng: Yeah.
208 00:31:23.490 ⇒ 00:31:25.139 Robert Tseng: Any other thoughts, Greg? Or…
209 00:31:25.140 ⇒ 00:31:45.140 Greg Stoutenburg: Yeah, I mean, I don’t really have anything to add. My thought similarly was, you know, I see the serious technical expertise, and… but would share the, share the alignment step concern about working with, you know, executive clients about what they would want to see, and what they think would help defend an experimentation roadmap.
210 00:31:45.570 ⇒ 00:31:46.380 Greg Stoutenburg: Yep.
211 00:31:47.280 ⇒ 00:31:47.700 Anthony: Okay.
212 00:31:47.700 ⇒ 00:31:48.210 Robert Tseng: Yeah.
213 00:31:49.070 ⇒ 00:32:08.200 Robert Tseng: Okay, well, yeah, thank you for your time, Anthony. I appreciate your… the effort you put here. Yeah, if there’s any other role that’s more kind of in line, I think we hire for just data engineers, analytics engineers. Like, I feel like those might have been, like, maybe there’s a role that’ll fit on the more technical side that’s not so…
214 00:32:08.380 ⇒ 00:32:21.620 Robert Tseng: I mean, this is… this is like, I need to feel confident that I can put you in front of a VP or C-level person, and just kind of, you know, give a very clear presentation to influence their business needs, and I don’t really feel like we got
215 00:32:22.160 ⇒ 00:32:30.669 Robert Tseng: I don’t think we got to that level of confidence in this call, but perhaps there’s another role that we’ll hire for that may be better fit for your skill set.
216 00:32:31.090 ⇒ 00:32:37.670 Anthony: Fine, yeah, I do understand, but I do have some questions to ask, maybe just a couple of questions to ask.
217 00:32:37.670 ⇒ 00:32:39.649 Robert Tseng: Yeah, sure, I can take a couple questions.
218 00:32:39.790 ⇒ 00:32:46.800 Anthony: Yeah, the first question is, I’m… I’m thinking maybe, the stakeholders would have worked on this analysis,
219 00:32:47.100 ⇒ 00:32:53.850 Anthony: Before, you know, giving it to us to work on. So, from your perspective, what would you say would be the…
220 00:32:53.960 ⇒ 00:32:57.040 Anthony: The greatest recommendation, from these data sets.
221 00:32:59.180 ⇒ 00:33:10.210 Robert Tseng: I think there’s a few different approaches, like, I think, what… I mean, I don’t think there’s, like, a silver bullet for the recommendation, but, like, even your particular example of, like.
222 00:33:10.210 ⇒ 00:33:10.670 Anthony: Okay, excellent.
223 00:33:10.670 ⇒ 00:33:15.170 Robert Tseng: setting an 8-order threshold. I think I would have liked to see more of, like.
224 00:33:15.770 ⇒ 00:33:18.350 Robert Tseng: Actually, when you look into the…
225 00:33:19.080 ⇒ 00:33:24.809 Robert Tseng: there… there’s a certain percentage of orders that have missing… missing items. I think you nailed that.
226 00:33:25.080 ⇒ 00:33:25.870 Anthony: 6%.
227 00:33:25.870 ⇒ 00:33:36.159 Robert Tseng: kind of, like, what… like, breaking that down further, what are the different causes? It could be, like, Dasher-related, it could be that, you know, Dasher Diligence, or whatever, they need the alerts.
228 00:33:36.160 ⇒ 00:33:47.229 Robert Tseng: you know, I would slide that in there. I would say, like, it could be a systems error. It could be a difference between the point-of-sale data and also, kind of, the inventory, wherever that’s coming from within the mart.
229 00:33:47.400 ⇒ 00:34:02.469 Robert Tseng: And so there’s… there is, like, a need to work with these, to hold these other grocery stores accountable, to make sure that when an order is being placed on the app, it’s actually showing up appropriately in the inventory. Like, I think that seems very likely, it’s a very high-leverage thing.
230 00:34:02.470 ⇒ 00:34:10.250 Robert Tseng: Very common, data problem when you’re working with, a platform that integrates multiple third-party systems.
231 00:34:10.400 ⇒ 00:34:19.750 Robert Tseng: And I think there could also just be, like, like a… I mean, aside from the integrations problem, there could be a, like, a…
232 00:34:20.340 ⇒ 00:34:44.989 Robert Tseng: just the or… like, the order is just not accurately reflected in the app as well, so it could just be, like, a pure app error. And so, I think you would take that same problem that you noticed, but then you would have to tackle it from multiple angles. I think it’s up to you to build the story of, like, which one you feel like is the most important to go after, but that shows me more that you kind of covered your bases on, like, you looked at the demand side, you looked at the supply side, you looked at the platform
233 00:34:44.989 ⇒ 00:34:45.819 Robert Tseng: form side.
234 00:34:46.020 ⇒ 00:35:00.660 Robert Tseng: you… there is some consideration to which one has the highest leverage, and then you’re… then you’re picking one and committing to it. So, I think, like, that’s kind of the kind of storytelling that we would probably be looking for more when you’re making a recommendation like this.
235 00:35:01.360 ⇒ 00:35:09.629 Anthony: Great. Thank you very much for your feedback. I really appreciate it. And thank you so much for the time and, you know, the experience so far. It’s been great.
236 00:35:10.500 ⇒ 00:35:15.499 Robert Tseng: Great, alright, yeah, thanks, Anthony, and, yeah, best of luck in your…
237 00:35:15.820 ⇒ 00:35:16.160 Anthony: night.
238 00:35:16.160 ⇒ 00:35:17.080 Robert Tseng: In your search.
239 00:35:17.410 ⇒ 00:35:17.810 Anthony: Alright.
240 00:35:17.810 ⇒ 00:35:19.090 Greg Stoutenburg: Thanks, Anthony. Take care.