Amazon Data Dudes Data Discrepancies
[00:00:00] Will: All right. Welcome everybody to this episode of Amazon Data Dudes. We are very excited today. I'm one of your hosts, Will Christensen, and we have Brandon Chekits here. Say hello. Hey, how's it going? And Jordan Chekits as well. Hey, everybody. Awesome. And today we have a special guest. We have Alexi Bell.
[00:00:20] Will: Alexi actually works at an Amazon advertising agency. Happens to be the very own Cellulabs advertising agency. I'm going to let her introduce a little bit more about herself, but she's got some really cool experience. She's going to talk a little bit about that today on our show. Episode of Amazon Data Dudes.
[00:00:37] Alexi: Awesome. Thanks for having me, guys. I'm Alexi Bell. I'm a account director at Seller Labs. I've been working at Amazon the last couple of years and have spent thousands and thousands of hours in Amazon. I wish I could count that high. And just seeing how much Amazon changes over the last, especially year and a half has been wild.
[00:00:57] Alexi: and just when we pull data one time, then we pulled again and it continues to change and alter. And so allowing us to have as consistent of data as possible for our clients and for ourselves to make the best decisions when it comes to advertising is. Critical. And it's just not, it's not consistent with Amazon.
[00:01:15] Alexi: So I would love to see consistent numbers and be able to provide the best data for our clients to make those
[00:01:23] Will: decisions. So correct me if I'm wrong, as you, you said thousands and thousands of hours, you've been like in seller central. Oh, yeah. Holy cow. I can honestly say that. thousands. I don't know about you, but that's not a place that I live, breathe, eat and sleep.
[00:01:40] Will: I spent a little more time in the code. Alexi, that is amazing. So today, the thing we want to talk about is a little bit about the data that's coming from the different systems and the consistency. I'm actually going to pass the baton over here. I have a data expert on Amazon. This is someone who has been dealing with Amazon data since, before many Amazon sellers were born.
[00:02:03] Will: there are Amazon sellers today who are born, who now are Amazon sellers, he was doing it before those Amazon sellers were even born. So Brandon, tell us a little bit more about Amazon data and what you've seen over the years. You're currently muted, you make
[00:02:21] Brandon: me sound really old when you say it like that, but so I don't know if I've been doing since before people were born.
[00:02:27] Brandon: I've been looking at Amazon data for seller labs for over 10 years, though, and ingesting a lot of that data, putting it in our own databases and trying to make sense out of it. But 1 of the things I think is useful is to think about the Amazon. The data we see in seller central is like, where is it coming from?
[00:02:42] Brandon: And what's their purpose and showing it to you? Because there is actually opinions in a lot of those different screens, like some of them are made to demonstrate financial accuracy and some of them were made to show you like, just general trends and stuff like that. So it helps to understand on Amazon side where the data is coming from and maybe how it helped, how it got there.
[00:03:02] Brandon: Got it.
[00:03:03] Will: And so over time, as that data is formed and fashioned and moved, you've seen different reports released via the API. you've seen different reports, via the UX UI. So that's the programmatic way, the API and the front end platform that they're actually showing you in Seller Central.
[00:03:22] Will: Are you seeing Amazon? are they constantly trying to give you the best data? What's been your opinion on how they
[00:03:29] Brandon: handle it? Amazon is different than most other, systems that we've talked with. And most other systems, if you're talking about a Shopify or somebody like there is one source of truth and it's the accuracy, it's accurate.
[00:03:40] Brandon: And no matter how you look at it, the numbers still add up and it doesn't matter. there's one database or one system that tells you the truth. And Amazon is very much not that way. If you think about how Amazon is structured from a corporate standpoint and from a team standpoint, like they have this concept of two person teams and one team doesn't know what the other person is doing.
[00:04:00] Brandon: and so like their data is fragmented and skewed and all, and segmented and all these different ways. And it it's a, by product of how their organization is run as well. So there is, from a seller's perspective, there is no source of this is all, this is the accurate. Source of all, orders, for example, that have ever been placed on the platform.
[00:04:22] Will: so what you're saying is, the left hand doesn't necessarily know what the right hand is doing. And and that is part of Amazon's competitive advantage, right? They're constantly trying to innovate and disruptive innovate and grow and change. And some of that has really served Jeff Bezos.
[00:04:37] Will: and we, we've seen some cool things with that, but here's it, here's a by product, right? That the left hand doesn't know what the right hand is doing. And so, this team over here doesn't realize they've messed up the data from over here is what you're saying. And a lot
[00:04:49] Brandon: of times it's good enough, right?
[00:04:51] Brandon: So it's good enough to demonstrate what they need. So there's no need to go back and figure it out. We'll point out an example. I think today that shows. Pretty clearly what's going on with that, but or stuff changes, legal things happen, orders are deleted, canceled returns, whatever. 1 of these screens, 1 of the data from 1 of these might have been taken at 1 point and then later change.
[00:05:10] Brandon: Then. So what is true? That kind of varies depending on who you're talking to. And it is on terms, what screen you're looking
[00:05:17] Will: at. So I want to take, Oh yeah, I was going to, I'm actually going right to you, Jordan. So Jordan, I want your take on this as an Amazon seller. So one, I want to know how many years have you been selling?
[00:05:28] Will: If you were to just throw out a rough number of units that you've sold in the entirety of everything that's there. I wonder if you even have that number in the back of your mind, like what are we talking here? How experienced a seller are you? And what have you seen? When it comes to this Amazon data.
[00:05:41] Will: Yeah. So,
[00:05:42] Jordan: I've been selling on Amazon since 2009. So this is before FBA. This is before Amazon advertising, man. It used to be so easy to create an Amazon account. You throw it up there and you, it was easy to calculate your money. It's oh, here's the Amazon fee. Here's my cost of goods sold. I'm going to sell it at this price.
[00:05:58] Jordan: I know I make money. Margins were much healthier back then. Obviously competition's good for the economy, but sometimes it's hard on us as sellers. in that, so 2009, I've probably sold, I would guess close to 10 million units, in that time. So a lot of, inventory, So I've, been around the block, so to speak.
[00:06:20] Jordan: the thing is like the history of, every year it's just, it's progressing so quickly, right? Like crazy quickly. And it's like margins are slimmer than ever. Amazon advertising came out and so many people started diving into it because they were told, Hey, you've got to advertise. They didn't know what they were doing.
[00:06:36] Jordan: So they were just throwing stupid amounts of money at Amazon advertising. And it made it really hard for people who did know what they were doing to compete. And so before you know it, you went from. 30, 40 percent profit margins to some people were losing money on inventory and they didn't even know it.
[00:06:52] Jordan: And so, our, standard profit margin right now is probably about anywhere from 7 to 10%, depending on the skew that's after Amazon advertising and everything. And so. Profit margins aren't what they used to be. The numbers, in my opinion, are more important than ever. And back in the day, if Amazon's data, had a discrepancy of two or 3%, it wasn't a big deal.
[00:07:15] Jordan: when your profit margins are 7%, two or 3% is a big deal. Two or three percent's
[00:07:20] Will: 50%
[00:07:20] Jordan: of your profit margins, it's a, huge deal, right? And so the, data, needing the data to be accurate, in my opinion, is more important than ever because we are dealing with such. Such a competitive, industry and margins are competitive you have to know your numbers.
[00:07:38] Jordan: If you don't know your numbers, you're not going to last
[00:07:40] Will: long. So Jordan, have you seen discrepancy? like you've looked at two different places in seller central and you've seen that. So we have an example we're going to share today. Jordan, in the past, this has been something you've seen as well.
[00:07:55] Jordan: Yeah. Yeah. tons of discrepancies. Occasionally Amazon will come out with a new report, right? So you pull the new report and you're like, that doesn't look right. And you compare it to another report and you're like, Hey, look, these two reports from Amazon are 5 percent different. Like these numbers should be the same, but they're not.
[00:08:10] Jordan: And I think it goes, I've never heard it explained as well as what Brandon said, how it's just you got this left and this right hand, and I think they're trying to solve similar things, but they're going about it different ways in situations. And so the numbers are slightly different.
[00:08:25] Jordan: Yeah. and so, yeah, I see discrepancies literally every day, like you pull reports. There's discrepancies. So
[00:08:32] Brandon: most people are grown accustomed to being off, right? As long as we're in the ballpark. But that's how we got this conversation is what 2 percent used to be acceptable. But 2 percent off now is it can be significant.
[00:08:44] Brandon: So we want to make sure we're as accurate as possible.
[00:08:47] Will: Beautiful. So we're going to show an example, but the last thing I want to do is, so is this across the board and I want Alexi and Jordan to weigh in here across the board, like advertising reports, order reports, business by the traffic reports, what, are we talking about, Alexi, what do you see?
[00:09:06] Alexi: I'd say the most consistent thing is it consistently being off. I'd say it's probably the best way to put it. I think it's even more sometimes when you pull an advertise report and the advertise sales are greater than the total sales during that time. that would, I'd say it's like the biggest.
[00:09:23] Alexi: Inconsistency and it sometimes it's not, that much off, but your total sales have to be bigger than your ad sales in that same time period. But then when you add in, the attribution window and stuff and such with ads, then you get a much larger discrepancy than, you would have thought.
[00:09:41] Alexi: So if everything would look at the same window instead of five different seven day, 14 day, 30 day, 60 day windows. I would think we'd have more consistency, but when you add in all those different windows. where is the consistency
[00:09:55] Jordan: there?
[00:09:55] Will: Yeah. Yeah. So Jordan, would you say it's outside of advertising as well?
[00:10:00] Will: Yeah, definitely. So
[00:10:02] Jordan: I'm currently, I've got six Amazon accounts and a lot of the inconsistencies that we see, are across the board, regardless of which Amazon account I'm in, definitely you go to your business reports within seller central and, those reports within business reports are usually pretty consistent 1 with another, because there are a lot of reports there.
[00:10:21] Jordan: But, to Alexi's point, I agree when you're comparing business reports to your advertising reports. That's where I see the largest discrepancies.
[00:10:29] Will: You would say that it's probably very important to be strategic about, okay, so I'm seeing a discrepancy. Let's get some logic behind choosing which data point is going to be more accurate because that should help you to make better decisions about your business.
[00:10:44] Will: Absolutely.
[00:10:45] Jordan: Yeah. and I love that. I love that Alexi talked about the timeframes, right? Like advertising report has, a certain number of days you can pull reports. And it's I get so frustrated because it's I want to look at my last year's numbers, but I can't in advertising.
[00:10:59] Jordan: So I can't, condense all the data that I want to condense and pull the year over year, trends. Like I want to, because I'm limited on the advertising. side, but not on the business report side. That for me is consistently one of the most frustrating things, to help. Cause there's seasonality to a lot of our products and I'm sure pretty much every seller has some kind of seasonality and it's month over month doesn't cut it.
[00:11:24] Jordan: A lot of times you have to look at those year over year trends. So
[00:11:27] Will: Brandon, you've had people come to you with this problem over the years as you founded Cellular Labs and they were asking like, okay, if we're going to take product data and store it for sellers? And we do that by the way, anybody who's watching this, if you're on Cellulabs Pro, We store your data for you.
[00:11:44] Will: So anything where Amazon drops off and loses that, we've got that for you. We've got your back. When you store data like that, you want it to be as accurate as possible for these sellers. Brandon, what decisions as the data architect, as the founder, as the lead programmer here at Cellular Labs, what have you done?
[00:12:01] Will: to what choices have you made to make that more accurate what we typically try to save as granular data as possible. So, and in some cases, and we're talking to some specifics here, like the Amazon business report section is very summarized data. So, it's pulling in data generally from, from across your entire account and showing you like, a chart by day or chart by month.
[00:12:23] Brandon: So that's very granular data. It's been summarized up into very. Very large data. So we typically try to save as granular data as possible so that we can really debug in there and get details. So, for instance, we, we save, from the business report section. We save the child traffic, the page detail traffic by child, ace report, which is the most detailed one we can drill into any particular child SKU or parent SKU and give you traffic for that one and so we try to pull in the most granular data possible. And then we can aggregate it back up. To, give you those reports at the high level.
[00:12:57] Will: Beautiful. So that granularity gives you more access and control and understanding of the data.
[00:13:04] Will: But if you try to go back and comparison it, compare it to Amazon Seller Central, you're gonna see discrepancies and you're gonna go wait. The seller labs know what they're doing. And the answer to that is this. This episode of Amazon data dudes, we do know what we're doing and we're trying to help you drill all the way down to that nitty gritty so that when you look at the high level and the low level from the data, that's, imagine the amount of work you'd have to do.
[00:13:27] Will: So Brandon walk through, what does it look like in terms of us grabbing this data? what would, an Amazon seller have to do manually in order to have the same data that we're, getting a seller who's not on Cellular Labs
[00:13:40] Brandon: sure. so we've actually pulling tons of data.
[00:13:42] Brandon: So we pull in order data. So we pull all your orders, what, date and time they shipped, the amounts of them, the products that were on them, the sale, like the price they sold for the discounts that were on it. We pull in all that data. We also pull in from the financial data, all the individual financial transactions that come in with that.
[00:13:59] Brandon: So we don't know how much went to product sales and to shipping and to Amazon fees and the FBA fees. Like we've got tons of detail there. And then we also pull in data. some overlapping data from some of that is from the, we'll talk about the business report section, that child traffic report that I showed, we pull in that data as well, because it has some data as far as like page views.
[00:14:21] Brandon: And so we're
[00:14:22] Will: talking like, if someone were to manually go do this, we're probably talking like an hour or 2 every day of downloading and putting it into Excel folders or something. And then even then it would become, unruly and difficult to keep up on.
[00:14:36] Brandon: yeah, certainly Excel and folders on your desk, but we put it into a database that's much like structured and very easy to get out.
[00:14:42] Brandon: If you understand it. not
[00:14:44] Jordan: only that, but not only does it, is it easier, but there are, reports and numbers you could look at now that literally. You can't, if you're just relying on Amazon, like you're so limited on some of your historical data in Amazon, but with seller labs, you keep the historical data.
[00:15:03] Jordan: And now I can look at advertising spend over a year. Like that to me is one of the
[00:15:07] Will: big advantages. Because your app it's automatically being captured, right? Like you're, signed up, you're, paying that monthly fee. and one of the things that's paying for is essentially an automated, totally automatic robot VA that's pulling in all this data, putting it into places where it's all sorted.
[00:15:24] Will: And then we're being able to divulge that. And our new data warehouse product actually does some really cool stuff with that and combining and making that available in a Google sheet or in Google data studio. So Brandon, we've got an example here. I want you to dive right into that. Can you show us and it's specifically around that exact report that Brandon was mentioning with the traffic, and the child liaison.
[00:15:44] Will: you show us what we're looking at here in terms of some of our discrepancies that we've seen? Absolutely,
[00:15:50] Brandon: let me put my shared screen here, go up, and I think I've got one on the screen here. This is for, this is one of the, in the, I'll navigate back to here again. So I've gone up to here to the report section, to business reports.
[00:16:03] Brandon: This is one I think most people are familiar with, the sales and traffic report here. It shows you I think the default view is to show you by day. I'm going to group it up by month and look at a longer term trend. So I want to look at what Jordan was talking about. There is a longer term trend going back to, year over year or whatever.
[00:16:20] Brandon: So this is a pretty common report. I think most sellers are familiar with. so what I've done here again, you saw me just do it. I wanted the sales and traffic made at the date range, go back a couple of years and grouped it by month. And so we can see in here, ordered product sales, break that down B to B page, total order items, all these little details about it.
[00:16:39] Brandon: But, I think we can demonstrate what I'm talking about here, but just at the order product sales column. So, yeah, let me go. So 1 of the things I wanted to do here, and I pointed out that seller labs. We download detailed page sales and traffic by child ASIN report. So we're, this is the one that has the most granular data in it.
[00:16:57] Brandon: So it's got things like at every SKU level. It's got I have to scroll bar, use their scroll bar here to go over. It's got the additional columns in it, like sessions page views by box percentage, and then all your conversion rates and stuff like that over here. So it's got more detail on this 1 than the other page has.
[00:17:17] Brandon: So we essentially download that report every day. But to show you what we're talking about here, I'm going to go back to 1 of the months I had in here. I think we're looking at, January of 2021. If I go select that
[00:17:30] Will: month, and if you were to just look at it right there. You would just see the total aggregate, but if you go grab it now, okay.
[00:17:37] Will: So is there a total at the bottom of this report? Can we just scroll down and see a, total for that?
[00:17:45] Brandon: I wish there was, there's no total down here and I'm convinced Amazon doesn't show you a total here. Cause the numbers don't add up, but, most all, most of the, all of the pages have got a total
[00:17:53] Will: at the bottom.
[00:17:53] Will: So if you scroll to the right there, yeah, there's nothing
[00:17:55] Brandon: there. there's no further rows down here. There's no total on it. So in order to sum all this up, I have to actually go up here and download the report. Okay. and so I download that and then I'm going to open it in Excel here.
[00:18:09] Brandon: All right. Let's see. There we go. Open up the Excel screen. So if I show you the same report here, so that's the same timeframe. And if I look at the number I was pointing that was the order product sales column. So you can see the total that we should have in there, according to the child traffic report, is 198, 164.
[00:18:31] Brandon: 55. Okay. But if I hop back over to the Amazon, sales and traffic report, if I look at the same time period for January, it's actually off by a pretty significant amount. It says $185,000, 3 44 0.49. Bring that. That th 13,000 bucks. Yeah, $13,000 off there. So 14,000 outta 200 is about 7% off or something's.
[00:18:55] Will: Okay, so hold on, You, said 7% off, Jordan, what did you say the profit margin was averaging Right now, ,
[00:19:02] Jordan: our average profit margins are seven to 10% hold, depending on the S skew, crap. So, so this is crazy because this is actually all just within the business report section too. So this, most of the discrepancies I've noticed have been between business reports and your advertising reports, but this is still all just within business reports.
[00:19:20] Jordan: This is supposed to be some of the more
[00:19:21] Will: accurate data. Jordan, I want you to imagine really quick that this is your account. and this, you look at this and it is off by that much. what do you do at that point in terms of how do you know what, decisions to make?
[00:19:36] Jordan: you can only make the decisions, as well as the data. That leads you to that decision. I don't know how to phrase that better. That was not worded greatly, but yeah, this is hugely concerning to me.
[00:19:52] Brandon: So let's go by these numbers or something. Let's say you were trying to predict your revenue for January of 2023.
[00:19:59] Brandon: And you're looking at 185, 000 dollars here for this January. And 100, 000 here for this January, what if this number was really 200 and you're showing it's showing 185. So if you're predicting how many units you need to buy or making decisions on that, then you're going to be undercutting yourself.
[00:20:19] Brandon: You're short. Yeah. So you're not gonna buy enough inventory. that's one way in which you might use that, but. Just using incorrect data means you make flawed decisions. Yeah, you use
[00:20:30] Jordan: data for everything, right? Like inventory, forecasting, profitability. I'm sure I'm not the only seller where, you get to the end of the year and you feel like you've had a good year, and then your P& L statement.
[00:20:41] Jordan: It's just Oh, we broke even Oh, you just, you're like, okay, like I, a lot of times you just don't know why, or so it's,
[00:20:52] Will: this is Brandon as the data expert here, which of these two reports would you choose to go off of? Or would you advise Amazon sellers to focus
[00:21:00] Brandon: on? so we've done some comparison on this and we dug into this one quite a lot and we actually pulled in.
[00:21:06] Brandon: So we actually have many sources of this order data. So we, pull in the order data from the orders API, which is very similar to that child traffic, the child report. And the numbers are much, much closer to the child traffic report than they are to the sales. But the high level report.
[00:21:21] Will: So what you're saying is you've done the dirty work of connecting those dots.
[00:21:25] Will: And you would recommend that, that report. That's the, can you go back to that page to show us which one you're talking about?
[00:21:32] Brandon: Yes. so we actually pull in data from many places. This is, this one is the one that is the real data that we get from the orders. API is very close match
[00:21:40] Will: to this data. Okay. So that data, and can you show us in seller central?
[00:21:43] Will: Can you go to that page real quick? That was this
[00:21:46] Brandon: tab I had open here. Okay.
[00:21:47] Will: So, so what you're telling me is the best place to go at it is from here. So if I want to go compare any random month in the past, I have to go hit that button and download it and do man, that sounds like a lot of work. If you're not a cellular apps customer.
[00:22:00] Brandon: Yeah. that's, so that's what we do in the background for you. So we download that report every single day, save it to our database and then keep it for years. So beautiful.
[00:22:09] Will: Do you mind stopping the, sharing there? So, obviously there was a piece of this episode that was about helping you understand who we are at Seller Labs and why we believe we have some really powerful data, helping you understand that we've done some of that dirty work and looked at it.
[00:22:25] Will: and some of you may have been like, wow, what a sales pitch. What we're really trying to help you understand is. We understand this data and we want you to learn how to use it in the best way possible. And we've done the dirty work for you. We've compared the different pieces. We've had hundreds of people point out the different discrepancies and we've chosen the best source possible for every data point you're seeing inside cellular labs.
[00:22:50] Will: that's, what it boils down to.
[00:22:52] Brandon: we've gone back and forth and made changes on all this stuff. And this is the result of years of, fine tuning and adjusting. So pretty good. Beautiful.
[00:23:01] Will: awesome. That is it for this episode of, Amazon data dudes, Alexi.
[00:23:05] Will: Thank you so much for coming today. We really appreciate having you here. And Brandon, thanks for your expertise on the data side. And Jordan, thanks for your insight as a seller. Yeah, good to be here. thanks guys.
[00:00:00] Will: All right. Welcome everybody to this episode of Amazon Data Dudes. We are very excited today. I'm one of your hosts, Will Christensen, and we have Brandon Chekits here. Say hello. Hey, how's it going? And Jordan Chekits as well. Hey, everybody. Awesome. And today we have a special guest. We have Alexi Bell.
[00:00:20] Will: Alexi actually works at an Amazon advertising agency. Happens to be the very own Cellulabs advertising agency. I'm going to let her introduce a little bit more about herself, but she's got some really cool experience. She's going to talk a little bit about that today on our show. Episode of Amazon Data Dudes.
[00:00:37] Alexi: Awesome. Thanks for having me, guys. I'm Alexi Bell. I'm a account director at Seller Labs. I've been working at Amazon the last couple of years and have spent thousands and thousands of hours in Amazon. I wish I could count that high. And just seeing how much Amazon changes over the last, especially year and a half has been wild.
[00:00:57] Alexi: and just when we pull data one time, then we pulled again and it continues to change and alter. And so allowing us to have as consistent of data as possible for our clients and for ourselves to make the best decisions when it comes to advertising is. Critical. And it's just not, it's not consistent with Amazon.
[00:01:15] Alexi: So I would love to see consistent numbers and be able to provide the best data for our clients to make those
[00:01:23] Will: decisions. So correct me if I'm wrong, as you, you said thousands and thousands of hours, you've been like in seller central. Oh, yeah. Holy cow. I can honestly say that. thousands. I don't know about you, but that's not a place that I live, breathe, eat and sleep.
[00:01:40] Will: I spent a little more time in the code. Alexi, that is amazing. So today, the thing we want to talk about is a little bit about the data that's coming from the different systems and the consistency. I'm actually going to pass the baton over here. I have a data expert on Amazon. This is someone who has been dealing with Amazon data since, before many Amazon sellers were born.
[00:02:03] Will: there are Amazon sellers today who are born, who now are Amazon sellers, he was doing it before those Amazon sellers were even born. So Brandon, tell us a little bit more about Amazon data and what you've seen over the years. You're currently muted, you make
[00:02:21] Brandon: me sound really old when you say it like that, but so I don't know if I've been doing since before people were born.
[00:02:27] Brandon: I've been looking at Amazon data for seller labs for over 10 years, though, and ingesting a lot of that data, putting it in our own databases and trying to make sense out of it. But 1 of the things I think is useful is to think about the Amazon. The data we see in seller central is like, where is it coming from?
[00:02:42] Brandon: And what's their purpose and showing it to you? Because there is actually opinions in a lot of those different screens, like some of them are made to demonstrate financial accuracy and some of them were made to show you like, just general trends and stuff like that. So it helps to understand on Amazon side where the data is coming from and maybe how it helped, how it got there.
[00:03:02] Brandon: Got it.
[00:03:03] Will: And so over time, as that data is formed and fashioned and moved, you've seen different reports released via the API. you've seen different reports, via the UX UI. So that's the programmatic way, the API and the front end platform that they're actually showing you in Seller Central.
[00:03:22] Will: Are you seeing Amazon? are they constantly trying to give you the best data? What's been your opinion on how they
[00:03:29] Brandon: handle it? Amazon is different than most other, systems that we've talked with. And most other systems, if you're talking about a Shopify or somebody like there is one source of truth and it's the accuracy, it's accurate.
[00:03:40] Brandon: And no matter how you look at it, the numbers still add up and it doesn't matter. there's one database or one system that tells you the truth. And Amazon is very much not that way. If you think about how Amazon is structured from a corporate standpoint and from a team standpoint, like they have this concept of two person teams and one team doesn't know what the other person is doing.
[00:04:00] Brandon: and so like their data is fragmented and skewed and all, and segmented and all these different ways. And it it's a, by product of how their organization is run as well. So there is, from a seller's perspective, there is no source of this is all, this is the accurate. Source of all, orders, for example, that have ever been placed on the platform.
[00:04:22] Will: so what you're saying is, the left hand doesn't necessarily know what the right hand is doing. And and that is part of Amazon's competitive advantage, right? They're constantly trying to innovate and disruptive innovate and grow and change. And some of that has really served Jeff Bezos.
[00:04:37] Will: and we, we've seen some cool things with that, but here's it, here's a by product, right? That the left hand doesn't know what the right hand is doing. And so, this team over here doesn't realize they've messed up the data from over here is what you're saying. And a lot
[00:04:49] Brandon: of times it's good enough, right?
[00:04:51] Brandon: So it's good enough to demonstrate what they need. So there's no need to go back and figure it out. We'll point out an example. I think today that shows. Pretty clearly what's going on with that, but or stuff changes, legal things happen, orders are deleted, canceled returns, whatever. 1 of these screens, 1 of the data from 1 of these might have been taken at 1 point and then later change.
[00:05:10] Brandon: Then. So what is true? That kind of varies depending on who you're talking to. And it is on terms, what screen you're looking
[00:05:17] Will: at. So I want to take, Oh yeah, I was going to, I'm actually going right to you, Jordan. So Jordan, I want your take on this as an Amazon seller. So one, I want to know how many years have you been selling?
[00:05:28] Will: If you were to just throw out a rough number of units that you've sold in the entirety of everything that's there. I wonder if you even have that number in the back of your mind, like what are we talking here? How experienced a seller are you? And what have you seen? When it comes to this Amazon data.
[00:05:41] Will: Yeah. So,
[00:05:42] Jordan: I've been selling on Amazon since 2009. So this is before FBA. This is before Amazon advertising, man. It used to be so easy to create an Amazon account. You throw it up there and you, it was easy to calculate your money. It's oh, here's the Amazon fee. Here's my cost of goods sold. I'm going to sell it at this price.
[00:05:58] Jordan: I know I make money. Margins were much healthier back then. Obviously competition's good for the economy, but sometimes it's hard on us as sellers. in that, so 2009, I've probably sold, I would guess close to 10 million units, in that time. So a lot of, inventory, So I've, been around the block, so to speak.
[00:06:20] Jordan: the thing is like the history of, every year it's just, it's progressing so quickly, right? Like crazy quickly. And it's like margins are slimmer than ever. Amazon advertising came out and so many people started diving into it because they were told, Hey, you've got to advertise. They didn't know what they were doing.
[00:06:36] Jordan: So they were just throwing stupid amounts of money at Amazon advertising. And it made it really hard for people who did know what they were doing to compete. And so before you know it, you went from. 30, 40 percent profit margins to some people were losing money on inventory and they didn't even know it.
[00:06:52] Jordan: And so, our, standard profit margin right now is probably about anywhere from 7 to 10%, depending on the skew that's after Amazon advertising and everything. And so. Profit margins aren't what they used to be. The numbers, in my opinion, are more important than ever. And back in the day, if Amazon's data, had a discrepancy of two or 3%, it wasn't a big deal.
[00:07:15] Jordan: when your profit margins are 7%, two or 3% is a big deal. Two or three percent's
[00:07:20] Will: 50%
[00:07:20] Jordan: of your profit margins, it's a, huge deal, right? And so the, data, needing the data to be accurate, in my opinion, is more important than ever because we are dealing with such. Such a competitive, industry and margins are competitive you have to know your numbers.
[00:07:38] Jordan: If you don't know your numbers, you're not going to last
[00:07:40] Will: long. So Jordan, have you seen discrepancy? like you've looked at two different places in seller central and you've seen that. So we have an example we're going to share today. Jordan, in the past, this has been something you've seen as well.
[00:07:55] Jordan: Yeah. Yeah. tons of discrepancies. Occasionally Amazon will come out with a new report, right? So you pull the new report and you're like, that doesn't look right. And you compare it to another report and you're like, Hey, look, these two reports from Amazon are 5 percent different. Like these numbers should be the same, but they're not.
[00:08:10] Jordan: And I think it goes, I've never heard it explained as well as what Brandon said, how it's just you got this left and this right hand, and I think they're trying to solve similar things, but they're going about it different ways in situations. And so the numbers are slightly different.
[00:08:25] Jordan: Yeah. and so, yeah, I see discrepancies literally every day, like you pull reports. There's discrepancies. So
[00:08:32] Brandon: most people are grown accustomed to being off, right? As long as we're in the ballpark. But that's how we got this conversation is what 2 percent used to be acceptable. But 2 percent off now is it can be significant.
[00:08:44] Brandon: So we want to make sure we're as accurate as possible.
[00:08:47] Will: Beautiful. So we're going to show an example, but the last thing I want to do is, so is this across the board and I want Alexi and Jordan to weigh in here across the board, like advertising reports, order reports, business by the traffic reports, what, are we talking about, Alexi, what do you see?
[00:09:06] Alexi: I'd say the most consistent thing is it consistently being off. I'd say it's probably the best way to put it. I think it's even more sometimes when you pull an advertise report and the advertise sales are greater than the total sales during that time. that would, I'd say it's like the biggest.
[00:09:23] Alexi: Inconsistency and it sometimes it's not, that much off, but your total sales have to be bigger than your ad sales in that same time period. But then when you add in, the attribution window and stuff and such with ads, then you get a much larger discrepancy than, you would have thought.
[00:09:41] Alexi: So if everything would look at the same window instead of five different seven day, 14 day, 30 day, 60 day windows. I would think we'd have more consistency, but when you add in all those different windows. where is the consistency
[00:09:55] Jordan: there?
[00:09:55] Will: Yeah. Yeah. So Jordan, would you say it's outside of advertising as well?
[00:10:00] Will: Yeah, definitely. So
[00:10:02] Jordan: I'm currently, I've got six Amazon accounts and a lot of the inconsistencies that we see, are across the board, regardless of which Amazon account I'm in, definitely you go to your business reports within seller central and, those reports within business reports are usually pretty consistent 1 with another, because there are a lot of reports there.
[00:10:21] Jordan: But, to Alexi's point, I agree when you're comparing business reports to your advertising reports. That's where I see the largest discrepancies.
[00:10:29] Will: You would say that it's probably very important to be strategic about, okay, so I'm seeing a discrepancy. Let's get some logic behind choosing which data point is going to be more accurate because that should help you to make better decisions about your business.
[00:10:44] Will: Absolutely.
[00:10:45] Jordan: Yeah. and I love that. I love that Alexi talked about the timeframes, right? Like advertising report has, a certain number of days you can pull reports. And it's I get so frustrated because it's I want to look at my last year's numbers, but I can't in advertising.
[00:10:59] Jordan: So I can't, condense all the data that I want to condense and pull the year over year, trends. Like I want to, because I'm limited on the advertising. side, but not on the business report side. That for me is consistently one of the most frustrating things, to help. Cause there's seasonality to a lot of our products and I'm sure pretty much every seller has some kind of seasonality and it's month over month doesn't cut it.
[00:11:24] Jordan: A lot of times you have to look at those year over year trends. So
[00:11:27] Will: Brandon, you've had people come to you with this problem over the years as you founded Cellular Labs and they were asking like, okay, if we're going to take product data and store it for sellers? And we do that by the way, anybody who's watching this, if you're on Cellulabs Pro, We store your data for you.
[00:11:44] Will: So anything where Amazon drops off and loses that, we've got that for you. We've got your back. When you store data like that, you want it to be as accurate as possible for these sellers. Brandon, what decisions as the data architect, as the founder, as the lead programmer here at Cellular Labs, what have you done?
[00:12:01] Will: to what choices have you made to make that more accurate what we typically try to save as granular data as possible. So, and in some cases, and we're talking to some specifics here, like the Amazon business report section is very summarized data. So, it's pulling in data generally from, from across your entire account and showing you like, a chart by day or chart by month.
[00:12:23] Brandon: So that's very granular data. It's been summarized up into very. Very large data. So we typically try to save as granular data as possible so that we can really debug in there and get details. So, for instance, we, we save, from the business report section. We save the child traffic, the page detail traffic by child, ace report, which is the most detailed one we can drill into any particular child SKU or parent SKU and give you traffic for that one and so we try to pull in the most granular data possible. And then we can aggregate it back up. To, give you those reports at the high level.
[00:12:57] Will: Beautiful. So that granularity gives you more access and control and understanding of the data.
[00:13:04] Will: But if you try to go back and comparison it, compare it to Amazon Seller Central, you're gonna see discrepancies and you're gonna go wait. The seller labs know what they're doing. And the answer to that is this. This episode of Amazon data dudes, we do know what we're doing and we're trying to help you drill all the way down to that nitty gritty so that when you look at the high level and the low level from the data, that's, imagine the amount of work you'd have to do.
[00:13:27] Will: So Brandon walk through, what does it look like in terms of us grabbing this data? what would, an Amazon seller have to do manually in order to have the same data that we're, getting a seller who's not on Cellular Labs
[00:13:40] Brandon: sure. so we've actually pulling tons of data.
[00:13:42] Brandon: So we pull in order data. So we pull all your orders, what, date and time they shipped, the amounts of them, the products that were on them, the sale, like the price they sold for the discounts that were on it. We pull in all that data. We also pull in from the financial data, all the individual financial transactions that come in with that.
[00:13:59] Brandon: So we don't know how much went to product sales and to shipping and to Amazon fees and the FBA fees. Like we've got tons of detail there. And then we also pull in data. some overlapping data from some of that is from the, we'll talk about the business report section, that child traffic report that I showed, we pull in that data as well, because it has some data as far as like page views.
[00:14:21] Brandon: And so we're
[00:14:22] Will: talking like, if someone were to manually go do this, we're probably talking like an hour or 2 every day of downloading and putting it into Excel folders or something. And then even then it would become, unruly and difficult to keep up on.
[00:14:36] Brandon: yeah, certainly Excel and folders on your desk, but we put it into a database that's much like structured and very easy to get out.
[00:14:42] Brandon: If you understand it. not
[00:14:44] Jordan: only that, but not only does it, is it easier, but there are, reports and numbers you could look at now that literally. You can't, if you're just relying on Amazon, like you're so limited on some of your historical data in Amazon, but with seller labs, you keep the historical data.
[00:15:03] Jordan: And now I can look at advertising spend over a year. Like that to me is one of the
[00:15:07] Will: big advantages. Because your app it's automatically being captured, right? Like you're, signed up, you're, paying that monthly fee. and one of the things that's paying for is essentially an automated, totally automatic robot VA that's pulling in all this data, putting it into places where it's all sorted.
[00:15:24] Will: And then we're being able to divulge that. And our new data warehouse product actually does some really cool stuff with that and combining and making that available in a Google sheet or in Google data studio. So Brandon, we've got an example here. I want you to dive right into that. Can you show us and it's specifically around that exact report that Brandon was mentioning with the traffic, and the child liaison.
[00:15:44] Will: you show us what we're looking at here in terms of some of our discrepancies that we've seen? Absolutely,
[00:15:50] Brandon: let me put my shared screen here, go up, and I think I've got one on the screen here. This is for, this is one of the, in the, I'll navigate back to here again. So I've gone up to here to the report section, to business reports.
[00:16:03] Brandon: This is one I think most people are familiar with, the sales and traffic report here. It shows you I think the default view is to show you by day. I'm going to group it up by month and look at a longer term trend. So I want to look at what Jordan was talking about. There is a longer term trend going back to, year over year or whatever.
[00:16:20] Brandon: So this is a pretty common report. I think most sellers are familiar with. so what I've done here again, you saw me just do it. I wanted the sales and traffic made at the date range, go back a couple of years and grouped it by month. And so we can see in here, ordered product sales, break that down B to B page, total order items, all these little details about it.
[00:16:39] Brandon: But, I think we can demonstrate what I'm talking about here, but just at the order product sales column. So, yeah, let me go. So 1 of the things I wanted to do here, and I pointed out that seller labs. We download detailed page sales and traffic by child ASIN report. So we're, this is the one that has the most granular data in it.
[00:16:57] Brandon: So it's got things like at every SKU level. It's got I have to scroll bar, use their scroll bar here to go over. It's got the additional columns in it, like sessions page views by box percentage, and then all your conversion rates and stuff like that over here. So it's got more detail on this 1 than the other page has.
[00:17:17] Brandon: So we essentially download that report every day. But to show you what we're talking about here, I'm going to go back to 1 of the months I had in here. I think we're looking at, January of 2021. If I go select that
[00:17:30] Will: month, and if you were to just look at it right there. You would just see the total aggregate, but if you go grab it now, okay.
[00:17:37] Will: So is there a total at the bottom of this report? Can we just scroll down and see a, total for that?
[00:17:45] Brandon: I wish there was, there's no total down here and I'm convinced Amazon doesn't show you a total here. Cause the numbers don't add up, but, most all, most of the, all of the pages have got a total
[00:17:53] Will: at the bottom.
[00:17:53] Will: So if you scroll to the right there, yeah, there's nothing
[00:17:55] Brandon: there. there's no further rows down here. There's no total on it. So in order to sum all this up, I have to actually go up here and download the report. Okay. and so I download that and then I'm going to open it in Excel here.
[00:18:09] Brandon: All right. Let's see. There we go. Open up the Excel screen. So if I show you the same report here, so that's the same timeframe. And if I look at the number I was pointing that was the order product sales column. So you can see the total that we should have in there, according to the child traffic report, is 198, 164.
[00:18:31] Brandon: 55. Okay. But if I hop back over to the Amazon, sales and traffic report, if I look at the same time period for January, it's actually off by a pretty significant amount. It says $185,000, 3 44 0.49. Bring that. That th 13,000 bucks. Yeah, $13,000 off there. So 14,000 outta 200 is about 7% off or something's.
[00:18:55] Will: Okay, so hold on, You, said 7% off, Jordan, what did you say the profit margin was averaging Right now, ,
[00:19:02] Jordan: our average profit margins are seven to 10% hold, depending on the S skew, crap. So, so this is crazy because this is actually all just within the business report section too. So this, most of the discrepancies I've noticed have been between business reports and your advertising reports, but this is still all just within business reports.
[00:19:20] Jordan: This is supposed to be some of the more
[00:19:21] Will: accurate data. Jordan, I want you to imagine really quick that this is your account. and this, you look at this and it is off by that much. what do you do at that point in terms of how do you know what, decisions to make?
[00:19:36] Jordan: you can only make the decisions, as well as the data. That leads you to that decision. I don't know how to phrase that better. That was not worded greatly, but yeah, this is hugely concerning to me.
[00:19:52] Brandon: So let's go by these numbers or something. Let's say you were trying to predict your revenue for January of 2023.
[00:19:59] Brandon: And you're looking at 185, 000 dollars here for this January. And 100, 000 here for this January, what if this number was really 200 and you're showing it's showing 185. So if you're predicting how many units you need to buy or making decisions on that, then you're going to be undercutting yourself.
[00:20:19] Brandon: You're short. Yeah. So you're not gonna buy enough inventory. that's one way in which you might use that, but. Just using incorrect data means you make flawed decisions. Yeah, you use
[00:20:30] Jordan: data for everything, right? Like inventory, forecasting, profitability. I'm sure I'm not the only seller where, you get to the end of the year and you feel like you've had a good year, and then your P& L statement.
[00:20:41] Jordan: It's just Oh, we broke even Oh, you just, you're like, okay, like I, a lot of times you just don't know why, or so it's,
[00:20:52] Will: this is Brandon as the data expert here, which of these two reports would you choose to go off of? Or would you advise Amazon sellers to focus
[00:21:00] Brandon: on? so we've done some comparison on this and we dug into this one quite a lot and we actually pulled in.
[00:21:06] Brandon: So we actually have many sources of this order data. So we, pull in the order data from the orders API, which is very similar to that child traffic, the child report. And the numbers are much, much closer to the child traffic report than they are to the sales. But the high level report.
[00:21:21] Will: So what you're saying is you've done the dirty work of connecting those dots.
[00:21:25] Will: And you would recommend that, that report. That's the, can you go back to that page to show us which one you're talking about?
[00:21:32] Brandon: Yes. so we actually pull in data from many places. This is, this one is the one that is the real data that we get from the orders. API is very close match
[00:21:40] Will: to this data. Okay. So that data, and can you show us in seller central?
[00:21:43] Will: Can you go to that page real quick? That was this
[00:21:46] Brandon: tab I had open here. Okay.
[00:21:47] Will: So, so what you're telling me is the best place to go at it is from here. So if I want to go compare any random month in the past, I have to go hit that button and download it and do man, that sounds like a lot of work. If you're not a cellular apps customer.
[00:22:00] Brandon: Yeah. that's, so that's what we do in the background for you. So we download that report every single day, save it to our database and then keep it for years. So beautiful.
[00:22:09] Will: Do you mind stopping the, sharing there? So, obviously there was a piece of this episode that was about helping you understand who we are at Seller Labs and why we believe we have some really powerful data, helping you understand that we've done some of that dirty work and looked at it.
[00:22:25] Will: and some of you may have been like, wow, what a sales pitch. What we're really trying to help you understand is. We understand this data and we want you to learn how to use it in the best way possible. And we've done the dirty work for you. We've compared the different pieces. We've had hundreds of people point out the different discrepancies and we've chosen the best source possible for every data point you're seeing inside cellular labs.
[00:22:50] Will: that's, what it boils down to.
[00:22:52] Brandon: we've gone back and forth and made changes on all this stuff. And this is the result of years of, fine tuning and adjusting. So pretty good. Beautiful.
[00:23:01] Will: awesome. That is it for this episode of, Amazon data dudes, Alexi.
[00:23:05] Will: Thank you so much for coming today. We really appreciate having you here. And Brandon, thanks for your expertise on the data side. And Jordan, thanks for your insight as a seller. Yeah, good to be here. thanks guys.
