64 minutes | Oct 9th 2020

EP239 - Retail Sales Data with US Census Bureau

Play
Like
Play Next
Mark Played

EP239 - Retail Sales Data with US Census Bureau 

Paul Bucchioni is Branch Chief and Scott Scheleur is a Supervisory Survey Statistician, both with the Retail Indicator Branch, Economic Indicators Division of the U.S. Census Bureau. In this interview, Paul and Scott walk us through the real sales data products that the US Census publishes and gives us advance about how to interpret the data.

Don’t forget to like our facebook page, and if you enjoyed this episode please write us a review on itunes.

Episode 239 of the Jason & Scot show was recorded live on Wednesday, October 7th, 2020.

http://jasonandscot.com

Join your hosts Jason "Retailgeek" Goldberg, Chief Commerce Strategy Officer at Publicis, and Scot Wingo, CEO of GetSpiffy and Co-Founder of ChannelAdvisor as they discuss the latest news and trends in the world of e-commerce and digital shopper marketing.

Transcript

Jason: [0:24] Welcome to the Jason and Scott show this is episode 239 being recorded on Wednesday October 7th 2020 I’m your host Jason retailgeek Goldberg and as usual I’m here with your co-hosts Scot Wingo.

Scot: [0:39] Hey Jason and welcome back Jason Scott show listeners, Jason if there’s one thing we agree on it’s our love of data and one of our favorite data sets is the e-commerce data published by the US Census Bureau. Both of us get a lot of questions about this data we talked about it a lot I think it was episode 233 we spent a fair amount of time kind of just talking about the data and what was going on there that we saw and, so we get all these questions and we thought what better way to understand this data then to hear the details right from the horse’s mouth. So on tonight’s show we were really excited to have and we just learned that they listen to the podcast that’s even more exciting Paul book kyani and Scott Sheeler on the show from the US Census Bureau welcome to the show guys.

Jason: [1:28] We are doing terrific and Paul I know you guys listen to the show so you know exactly how we always start we like to let the guests introduced themselves a little bit to the audience so maybe we could start with you can you tell us a little bit about yourself how you came to the field and in your current role at the US Census Bureau.

Paul: [1:47] Sure so I started my career at the Census Bureau about 16 years ago and the entire time that I’ve been there I have been working in the monthly retailgeek, I’ll Branch so I’ve always always been working with retail and one way or another my background is in finance and economics, I dealt a little bit with Finance in the in the stockbroker type field before I came to the bureau but since I’ve been at the bureau I’ve been in the monthly retail started as an analyst I’ve worked on all industries that has to do with monthly Retail Sales inventory, e-commerce and I’ve been I’ve been in here ever since and working as a manager now I’m the branch chief of the of the retail branch and yeah that’s pretty much what I’ve been doing the last 16 years.

Jason: [2:31] That is awesome I’m always terrified to talk data and economics with people that actually know the fields like we have these phrases at work there are people that are data fluent and there are people that our data literate and I’m like barely data literate. Thrilled thrilled to have some data fluent people on the show and Scott can you introduce yourself.

Scott S: [2:56] Sure thanks Jason Scott Schiller I’ve been at the bureau will just say longer than Paul and. You know I’ve been involved in basically retail wholesale and services my whole career currently both Paul and I are in the economic indicators division. Where we are kind of responsible for measuring a lot of the economy from the trade balances to the manufacturer shipments inventories orders data to the quarterly Services data to monthly wholesale and of course to retailgeek. So we have a we cover most most of the economy in our division and our specific areas responsible for what, fancy title is consumption and wholesale indicators which basically means we can cover consumer spending, retail and services as well as the wholesale piece of the economy so that’s what we’re currently doing prior to the bureau I actually started your right out of college.

Jason: [4:05] That that is certainly impressive and, I just want to clarify because some before we even get going people talk about us Department of Commerce data and they talk about Census Data you guys both worked for the US Census and this, is in fact US Census Data right so I’m assuming for the most part people are confused when they give other departments credit.

Paul: [4:29] Yeah so you’re correct I mean like the US Census Bureau is part of the Department of Commerce so you know and it’s it’s always been one of those things where every time a release comes out you’ll notice when you look on the news or something it says the US Department of Commerce released today, three tail numbers right but it’s comes from the US Census Bureau.

Jason: [4:47] Got it and then this is for sure the coolest part of the US Census Bureau is that also true.

Paul: [4:54] Well absolutely of course. There’s no doubt about that everyone just thinks that you know you know I’ve always said oh is that the clarifying Scott probably has to do the same thing whenever we’re somewhere and someone says hey what do you do for a living and because I work at the US Census Bureau what do you think they say oh so you can’t people right so it’s not nope there’s there’s more to it than that so.

Jason: [5:12] Yeah I was going to say like I assume everyone’s first reaction is wait you only have to do anything like every ten.

Paul: [5:17] Every 10 years right what do you do the rest of the time yeah I have to come up with some so many answer but it’s not hard.

Scot: [5:22] Yeah I like to even reduce that you’re from the bureau it makes you sound like the FBI I bet you at cocktail parties are like we’re from the bureau.

Paul: [5:27] Yes we’re from the bureau you just slam it down yeah.

Scot: [5:34] Let’s start it at kind of the the super 30,000 foot level you guys put out a bunch of I guess you call them products so there’s like the advanced the monthly the quarterly, maybe talk listeners through those products and then you know how do you get this data and roll it up into the products.

Paul: [5:54] Sure I’ll take that I’ll start with that so right so you really can’t talk about all the monthly retail programs without kind of seeing like where it starts from its senses and so you know but besides that our monthly surveys we have an economic census and then we have also an annual retail trade survey so you know the economic census is basically the foundation of the retail programs it’s collected every five years, and it’s similar to like what we were just talking about that the 10-year population counterpart that the senio what it does is it collects information for all retail stores in the US including sales by product business, characteristics employment and payroll and then there’s the annual survey which comes out every year and that’s roughly about 20,000 companies, okay and that collects more than just sales and inventory it also collects things like purchases and accounts receivable and things along those lines and then you get to the stuff that we do in our branch, so the monthly retail trade survey and the advanced month of retail trade survey which I know can sometimes give people a little hard to understand, the the monthly retail trade survey is roughly about 13,000, sample size okay and what we do there is that collects that actually collects sales data inventory data and e-commerce data, and it comes out approximately six weeks after the close of the reference month it’s been around since 1951, um and about eight you know week it’s we’ll talk about the data in a little bit about like at the types of data that’s easily adjusted seasonally adjusted.

[7:23] And then the advanced one which is the advanced monthly retail trade survey which I believe is the one that most people use in the one that everybody gets excited about on release day on CNN and Wall Street that is a sample size of approximately 5,500 companies, it’s a subsample of that monthly retail trade survey just like the monthly retail trade survey is a subsample of the annual survey that has been in existence since 1953. And it goes it basically what the only the only.

[7:57] Did data that we collect on that survey is sales data because it’s such a quick turnaround because it’s a it’s basically it’s released approximately two weeks after the close of the reference month so as Jason probably knows, it will be the D September data will be released next Friday, on October 16th and that will release the September data and that will also then release the monthly retail trade the preliminary data for August, so and then in addition to that one last piece is the is the quarterly retail e-commerce product which as you can imagine comes out quarterly and that collects just the e-commerce data so it takes data from, just the just the e-commerce forms the data that has already been collected on the monthly retail trade survey but it’s broken down into just e-commerce, and then we also have a supplemental product that came out last year which gives you even more granular data and takes that Top Line e-commerce number and breaks it out into internet’s codes which will get into different Industries so you can kind of see more and more of the granularity which people have been asking for for some time though so. That’s kind of like the quick overview without spending too much time on that I hope that hope that hit and what you.

Scot: [9:09] Yeah it’s super helpful can you give us a sneak preview of the data I’m just kidding. The, when you said some of these things go back to the 50s that immediately popped in my head was someone had to go you know I’m envisioning that being on like some kind of a punch card and then like having to convert it into some format that you can work with now is how does that data like. How did you guys have to go convert that data back to something you can you can now make more.

Paul: [9:39] So that’s very great very very good question so when I first started here a long time ago one of my first tasks was to we had all these on paper. And so we scan them in basically there and some of them because you can imagine we’re quite discolored but we had them and so what we did is we stand them and we were able to put them on the web as PDF documents and they are on their from so if you actually go to our website which we will talk about later census dot-gov backslash retail I’ll try to plug that as much as I can you will see on there we have all of our historic data for releases so users can go in and then they can actually pull up like the May 1964 release and see what the sales were back then it’s pretty cool.

Scot: [10:23] Very cool wow.

Scott S: [10:24] Believe it or not I you know what Paul got here I gave him that assignment and somehow somehow he still stuck around he didn’t like go running for the Hills.

Scot: [10:32] That’s The Hazing the new guy.

Scott S: [10:34] It was like the hazing.

Scot: [10:35] Yeah I got a first project for you what do you know about.

Scott S: [10:38] It’s right here.

Paul: [10:38] I’m still with him yeah I was a little worried I said is this is this what I’m going to be doing for when I got here like.

Scot: [10:47] If scanning off.

Paul: [10:49] Yeah right right.

Scot: [10:50] The so there every five year you when you said all retail stores so that’s so like what how many retail stores are there’s are like 50 thousand a hundred thousand like I have no idea.

Scott S: [11:02] So I thought like this one so so basically the economic census does, basically a collection of all firms with employees so so that’s roughly you know I don’t know what the latest number is maybe the million range of locations retail locations of course there’s not that many. Retailers because there’s Raquel from multiple locations but we. You know we do ask retailers that have a number of stores to provide information for each of their individual stores so it is a massive operation, you know not quite to the scale of the 10-year population census but it’s a. It’s a pretty big operation where we do reset kind of like all of the information we have about businesses we have methods to do that in the interim between the five years there’s there’s ways we get structural changes and we get updates from, other administrative sources but that’s a big kind of once every five years begin to reset do an actual physical kind of like basic like a physical count.

Scot: [11:59] Wow you guys like Dread that year like wins then went to the next one I don’t know where we are in the cycle.

Scott S: [12:03] So it’s not every year in the Years ending in 2 and 7 so the next one 2022 and Paul and I defer to our colleagues on that one we let them handle that one we got our hands full with the month.

Paul: [12:12] Yes there’s a there’s another another area that takes care of that just like there’s another area that takes care of the annual break them all up so.

Scott S: [12:21] But we work we work very closely with them as you can.

Scot: [12:24] Yes you guys spend the bulk of your time on the monthly and quarterly just making sure all that that that’s enough Cadence that it’s a full-time gig.

Paul: [12:30] That takes up all of our time it’s a challenge enough to get that data out as quick as we do with with with. As reliable as we’d like it to be is collecting all the data that we do so yes it keeps us very busy and then we have other things that we do which we’ll get into later like when we actually you know, Benchmark to the annual survey every year so we have different things that we do besides that that keeps us busy.

Jason: [12:55] Awesome so I want to double click in the some of this but a couple of quick questions first so 13,000, businesses fill out the the monthly retail trade survey and five point five and a half thousand businesses fill out the advanced monthly is it the same 13 + 5 .5 every month or does it cycle or.

Paul: [13:19] Great question so what we do it is the same every month for a certain period of time so we do a what we call a business sample revision we do that roughly every five years for the monthly retail trade surgery and about every two and a half years for the advance monthly retail trade survey that way it gives its it give it puts less burden on respondents out there you know so that someone’s not reporting to the survey for you know which 40 years straight or something like that and now obviously depending on the type of company some people there are companies that stay on the survey for a longer period of time but but it is the same ones each month and the 5500 that are on the March sample side are just a subsample from that 13,000 in the month of retail trade survey.

Jason: [14:05] Got it so on that cycle you pick your 13,000 respondents in your 5.5 and then it’s actually the law that they have to fill out that survey right like.

Paul: [14:16] So it is the law for them to fill out the annual but the but the monthly retail trade and the advanced retail is not mandatory it is voluntary but we do stress the importance, to the responders from when the when we mail them and talk to them.

Jason: [14:32] Gotcha and then so then we ask for these 13 and then presumably you very reliably get some subset of that and some subset you don’t get like if you if it’s voluntary and some people don’t, provide do you just use like some statistical means to sort of level it up or or.

Paul: [14:55] So Yeah so we there’s lots of documentation on our website as well since it’s not gov backslash retail that has sections about like how data is collected how the survey works, and so you know we talked about ways and they’re how we how we handle and how we take care of companies that did not report in time and and Scott can even give you a little bit more information on this as well.

Scott S: [15:21] Yeah basically if the company doesn’t report then yes there you know as part of the monthly retail trade survey we will. We will develop an estimate for them based on either the reporting history of the company or how the companies the other companies in the same industry are performing for that month. So we do come up with a replacement value for companies if they don’t report.

Jason: [15:46] Got it okay and fun fact I was actually just on the in RF digital council meeting this week and the the in RFC Chief Economist was you know talking about some stuff and he you know not surprisingly who averaged a bunch of your data and people are asking all kinds of questions about your data and like you know these are all the vp’s of e-commerce and a as a joke and this will make more sense later they all called themselves the chief non stores officers that’s the. The Unofficial title that they all gave themselves and they were all talking about how when they were earlier in their career they used to be their job to fill out the census.

Scott S: [16:28] Interesting.

Jason: [16:29] Yeah so there are several people that are like oh man I used to have to do that every month, so so that’s that it just kind of fun to think about so we’re all. Super short attention spans of course and so everyone pays attention in my world to the advanced monthly because we want to talk about last month as soon as possible right and so two weeks after the month clothes you give us this cool Advanced Data. A month later you get a bigger sample does, so I’m assuming if I wanted to look at like a three-year Trend or something I’m probably smarter to use the monthly survey monthly product versus the advanced monthly product because I’m only losing, one one month of data and I’m getting a bigger sample size.

Paul: [17:19] So yes you’re correct so the advanced one like you said is the one that everybody looks for right it’s that first indication of how the how the economy is doing in the number that, that that people see and then what happens is you have time then so you know there’s real if a company cannot report in time to that advanced survey then, can report in time for the possibly for the monthly okay so so you’re going to get more you’re going to get more data in the monthly series then you do in the in the advanced because there’s because over time you’re going to add to it because you can revive at any moment in time you can revise the prior month so like when we put out data next next, Friday for the September March will also be putting out the, August preliminary now that just came out last month so there should be more data that comes in you’ll get a more updated number which is when they start to talk about what are revisions are out there, um so yes if you’re going to develop any type of Time series that would be what people would use and then even more so than that what happens every year as we will in the in the April time frame as we will put out our annual revision our Benchmark report to the annual and could and opens up for us to add even more data at that point so that’s how that would work does that make sense.

Jason: [18:33] It makes total sense and then one more question because Scott’s chomping at the bit to jump in the. When you get that more data like do you actually publish a revision for example to the advanced product or or does it just show up in the the monthly product.

Paul: [18:52] So it’ll show up in the advanced product as the prior month but it will also show up in the monthly product so because only the prior month is getting revised so that’s where you will that’s where you’ll see it, we do not go back and like recreate the report again if that’s what you’re.

Jason: [19:08] Yeah I know but but so to be clear like on October 16th you’re going to publish an advanced monthly product that has the September data in it you’ll publish the monthly product that had or no. That has the August data in it. That right yeah August and we also publish a revision to the August Advanced Data or no.

Paul: [19:32] No there will not be a revision to the August Advance there will be there will be a revision to the August preliminary but you’ll see that in the September advance in the columns where we show the revisions to the August that will be there and you’ll also even at that point in the August, a full report that comes out you’ll also see the final revision that you can make for July so you can still actually correct that as well. And so that’s that’s how that works.

Scott S: [19:59] The other thing to add there is when we release the monthly retail trade survey so like in the example we’re using. When we release the August you know next week when we also released the September Advance the August will also include a bunch of additional detailed levels because the advance is so quick and sample size is smaller, you know the so you know the industries are sub-sectors we come sub-sectors that under retail that we are able to publish you know are somewhat Limited. You know so we are able to expand that and publish a lot more detail. When we do that when we do the larger monthly survey ferns for example like in clothing stores like in the advanced will publish clothing stores butt in when we get to the, monthly were able to break it down in the women’s clothing stores or Family Clothing Stores or shoe stores or, Alex cetera so that’s that’s the advantage of having the larger sample and having a little more time to get it done is that we are able to dive into the data a little deeper.

Scot: [20:55] Brickell I don’t know why I never connected these dots but just does all this roll up into GDP.

Paul: [21:02] Great question again so yes so the the it’s about a third of the retail that is about a third of the personal consumption expenditure component of GDP so and the other two-thirds is made up of services so that’s how, Retail Partners.

Scot: [21:18] Yeah I have a started a new company that’s Digital Services so I have this whole Spiel where I talk about GDP breaks down you know consumer services are much larger than super good so you guys create GDP that’s crazy you measure you measure GDP.

Paul: [21:32] Yeah we actually were and we work with we work with PE a a lot who creates GDP so they’re actually you know when we were in the building they’re in the building were into so it’s a little easier.

Scot: [21:42] Trust me out I don’t know why.

Scott S: [21:44] Yeah there are Sister Sister agency of us under Department of Commerce.

Scot: [21:48] Very cool okay I never I never connected that I don’t maybe I’m the only ones but who knows. The sudden let’s go let’s go back to the data so Jason kind of tease this a non-store definition give us on the e-commerce guy on the show so how does e-commerce come out in these reports store non-store what are the definitions of all those things.

Paul: [22:13] So I’ll give you the high level, it in the in the monthly report so in the advanced report but we have an on store number that comes out at a very high level and I’ve heard Jason hit on this before in previous episodes it includes more than just e-commerce so it’s going to include things like Fuel and like direct Sellers and catalogs for those that are still around things like that so and then once you get to the monthly retail trade survey the the larger one where there’s a little bit more detail There’s an actual e-commerce line there, and so in that line now it’s you’re taking out the fuel and the direct seller so you just have the e-commerce but you’re only have the e-commerce for, companies that, that do not set that do not fulfill from their store so they have a sense they’re big enough that they have a separate Warehouse that they fulfill from and that’s what goes into that number I’ll let Scott now talk a little bit about what goes in if you’re into of what goes into the quarterly e-commerce report and how that breaks it down even a little further.

Scott S: [23:10] Yeah so so so on the cool you know so on on the monthly you know we have that line that says electronic shopping and mail order houses you know so keep in the mail order houses in there. Switch back to what it used to what it started as basically, you know but it but it includes you know anything a company that would be classified as one of those things wouldn’t would have so if it’s all sales whether it’s online whether it’s whether they actually still have some mail order operations if they take a phone call. You don’t process an order that way all of that data would be included so it’s not a hundred percent e-commerce. And it’s not all of the e-commerce so what we try to do in the quarterly basis is we produce two things one is just a high-level report. That basically takes sums up for all of retail and this will be excluding Food Services because technically that’s not retailgeek. What’s the total e-commerce number that’s the number that you know comes out once a quarter. You know that gets quoted and and you know what we released the second quarter e-commerce data like on a seed like Dustin basis and I know we’ll probably talk about seat adjustment later that was roughly about 60% of. Retail sales for the second quarter so that’s so that was the original report that reports been around since the fourth quarter of 99 we started doing that back then.

[24:25] Of course in recent years and you guys have touched on this in you and some of your podcasts. You know it’s a little grayer what companies are now doing with their online fulfillment. And as a result we’ve tried to adapt what we’re able to provide to the users specifically you know that’s Paul said. The goal behind the non-store is that that’s really a non store operation well now you have all these things like you know fulfill from store ship the store. Pompous you know you have all these different components which graze the.

[24:58] You know what are people doing how are they breaking it out how are how are people reporting the data and you know at the end of the day we’re at the mercy of the retailers to basically. Report the data as they see it and Argo and and because like like Paul indicated, you know we’re relying on the businesses to produce the survey we try to align when we asked them with what they have, you know we don’t want to give them all these elaborate instructions that basically maybe at the end of the day would basically label us to put it in all the right buckets. They can’t give it to us and then we just lose the response and that would be worse so what we tried to do is we tried to say okay you know recognizing that there might be some inconsistency and you know how a company either you know ships to your house or you buy online pick up in store and how that’s fulfilled might be different. We basically produced an experimental product about a year ago and actually I think Jason reference this in a podcast not too long ago that.

[25:55] You know where we kind of tried to combine it and say okay you know what regardless of whether a store tells us they fulfill it from their store or they fulfill it from a warehouse you know if the store is say you know clothing store. Let’s just some all that data together so if they fill it for filling from their store for Phillip from Warehouse no matter what the. You know what the origination is you know let’s call that like a digital transaction online transaction everyone call it and let’s kind of group those together so that you could kind of see how things are. You know how that performances you know across the various Industries and how it compares to one another so so that was a big that was a big achievement we were able to do that with the data we already have some companies which was nice we didn’t have to reach out and burn the company’s again. And and it did provide some granularity and of course with with the the data that’s been coming out of the past. You know five or six months I mean it’s really provided some insight into what’s been going on.

Scot: [26:50] Trickle the this is a question to feel free not to answer this so you know so it’s neon e-commerce guy I immediately kind of think well you can’t measure your Commerce without measuring Amazon is there do you capture Amazon in this at all or are and it’s fine if you can’t say.

Scott S: [27:06] We can’t we can’t.

Paul: [27:06] We can’t say you’re correct we cannot say.

Scott S: [27:09] Yeah I mean I’ll retell all retailers are eligible to be selected for the sample but we can’t talk about any specific ones.

Scot: [27:16] Got it okay I have had a feeling that maybe since about that’s fine, um nothing that’s really interesting though is so like Amazon is there’s Amazon themselves and then there’s like all these third-party sellers so it seems like you know and then you know the company I started previously we had thousands of these little sellers so. The interesting. To you know 13,000 could be a bunch of small ones you could have big ones do sir how do you get the mix for that to try to look like the u.s. mix sir is there some specific stats magic you do there.

Scott S: [27:48] Yeah so basically and that’s kind of where the economic census comes in is that by having all that information on all of the businesses in the United States were able. Stratify the sample so that we can you know. Um produce estimates at the levels we want to and make sure we have a good mix of the large businesses small businesses Etc, and we stratified based on size and size of the business and you know and then we can use the economic Census Data as kind of a you know the target population make sure it’s representative.

Scot: [28:17] Got it that makes sense yeah so that gives you you know you have this five-year check in to kind of understand what the what what the statistical relevance of everything is.

Jason: [28:29] Awesome so just a couple more qualifications on the e-commerce and we’ll move on obviously like this is what we get the most questions about so-so a white just to re-emphasize what you’ve already said every every retailer accounts differently in some some Count Their e-commerce separate some don’t some call it e-commerce if they take the money online some call it e-commerce of the products deliver to the hat like there’s a ton of different definitions and you’re trying to ask questions that the retailer is likely to know verses.

Scott S: [29:01] Exactly.

Jason: [29:02] So super hard.

Paul: [29:04] It’s a challenge.

Jason: [29:05] It’s a challenge for sure I can absolutely appreciate that so in the abstract perfect world if, Scot Star Wars memorabilia.com and I don’t have any stores I’m only online I should be reporting a hundred percent of my Revenue as e-commerce and so on the Advanced monthly it’s going to be bundled into the non-store and in the monthly it’s going to say e-commerce right.

Paul: [29:31] That is correct you would you would be what the industry calls a pure play so you would be your entire your entire business, live solely online you would fill that out right and did exactly what you said you would be in the non-store category in the advance and then you would be in the, you would be in the electronic e-commerce portion of the monthly and then you would also be in the quarterly report.

Jason: [29:54] Yeah and then if I later opened a store and I sold half of my stuff out of the store and the other half I shipped to customers homes from my fulfillment center, in an Ideal World I should be reporting half of my sales as store sales which would then show up in the in the appropriate category which I assume Star Wars memorabilia is essential Goods is probably the category that.

Paul: [30:20] Mmm Yeah.

Scott S: [30:21] Obviously.

Jason: [30:23] And then the other half of my sales I would report as e-commerce is that am I thinking about.

Paul: [30:30] That is correct right and if you were so nice as to break that out for us and give us that data like that then yes it would that’s how we would capture correct.

Jason: [30:36] But it’s entirely possible that someone filling out the survey is just going to say we’re predominantly stores or were predominantly e-commerce and therefore we’re going to put all the numbers in.

Paul: [30:46] And this is kind of like what Scott had hit up before you know where we’re always going to be at the mercy of the retailer right as to how they deem what they what they think the definition to them is e-commerce or retail rates oh no matter how we tweak our definition or try to learn more about the definition from doing research and talking to people you know we know what we want and we know what we deem to be what e-commerce is it’s just getting everyone’s opinion to buy in on that as is that as I said before is always good is always going to be the challenge but we’re always we’re always trying to reach out and talk to different agencies and and and establishments to see you know what others are thinking as well and we’ve gotten some good good feedback from people.

Scott S: [31:26] Yeah we spent weeks spend time working with like a lot of the trade associations you mentioned National Retail Federation earlier we you know we work with them you know and and try to you know and others to make sure that we. That our definitions represent kind of how the industry views, you know the definitions the e-commerce of course is changing so fast so rapidly that that’s a hard one.

Jason: [31:51] Are the challenges it’s a moving Target.

Scott S: [31:52] That’s a moving Target.

Jason: [31:54] You’re talking about and I know you can’t talk about specific retards but if you were talking about Target five years ago they had a bunch of e-commerce and they shipped it all from a fulfillment center today they fulfill 80% of their e-commerce from stores so like whatever was right five years ago for them would not be right today.

Scott S: [32:12] Exactly one other one other clarification is that you know what we do for e-commerce in terms of companies that have like a separate e-commerce division, versus their brick-and-mortar stores either whether they started as brick and mortar and went to e-commerce or vice versa it’s so early what we would do for any part of retail so if a company operates furniture stores and they operate clothing stores. You know that’s two different Industries and so we would ask them to split out that data separately and for the most part companies will split the date out, and similarly with the e-commerce move we get good cooperation the companies will split out the data so that we can put them into the different Industries and measure how furniture stores are doing and clothing separately so that we don’t like. Put the furniture Trend in the clothing in the clothing stores industry so, so that’s it’s similar to how we do it for all the industries with you know we would so a complex company that operates four different types of stores you know Grocery and furniture and clothing and department stores or something could be in for could get for different. Request from us to fill out.

Jason: [33:14] Got it and then one more thing I need Commerce and then I’m going to move on I promise, the a cool thing about the new e-commerce products is so if you’re just looking at the monthly data the monthly product you’re going to see, e-commerce aggregated so you’re going to see a total number for e-commerce but you’re not going to know how much of that was a parallel versus home goods for example and so so one of the cool things about the quarterly product is you you then try to categorize, or disaggregate the e-commerce into those categories do I have that right.

Paul: [33:50] You do that has been something that has been asked of us for a pretty pretty long time and as you can imagine to get a new product up and running takes it takes years and years of research and talking to people and Smee and Scott were involved with that as well as other as well as some of our other colleagues meeting with different organizations and stuff and so to get that put out for that granularity for people you know, it was a huge success I think people would gotten some some very good feedback on it because and especially like if you look at the if you look at like the most recent report you know in a time like now we’re a lot more a lot more activity is happening online right you wouldn’t you wouldn’t know from the quarterly table that we put out before when it was just the Top Line number that something like, food and beverage stores were up a hundred and one percent from the previous quarter from second quarter to the first quarter and that and that from the prior year quarter is up 220 percent I mean, that’s that’s huge and we would have never got that before without us having it that supplemental table you want to just got a large e-commerce number and I think a lot of people would have just been like oh yeah we know e-commerce is large right now because everybody’s buying online writing is all this new stuff like buy online pick up in store so to get that granularity with and get to see a lot of these different Industries and what they’re doing you over years I think is a definite value for people.

Jason: [35:08] Awesome so now I want to transition so you mentioned your url census.gov forward slash retail you’ll be happy to know that’s the third favorite on my favorites bar.

Paul: [35:22] That makes me happy probably makes God happy too so.

Scott S: [35:24] Wait what what I want what are number one.

Jason: [35:27] Yeah don’t be too mad it’s Gmail and like my my work salesforce.com login.

Scott S: [35:34] LOL that yeah that’s good.

Jason: [35:37] So you’re right up there I’m not saying I organized it by frequency or anything.

[35:43] You are where you are so but on the morning of August October 16. I and a lot of other people in the industry we’ll get up and we’ll I’ll admit I get up to eight to actually wait for the data it’s already waiting for me, I’ll get up that that data is available I’ll do you know look at see what happened and I will start throwing out some tweets about like what I found interesting about the data and so will a bunch of other people and, to my annoyance none of us will say the same thing or quote the same data so I. Three people will all say oh the the US Department of Commerce or I will say the US Census Bureau. Released this data and x and it was by 5.2 percent or whatever right and then someone else will say a different thing and it’ll be four point three percent and another and so I want to kind of and I think I know why that’s happening and I know you guys know why that’s happening but I want to our listeners to understand um what’s happening there’s a bunch of different cuts of the data in different ways to look at it in different ways to talk about it and unfortunately most of us.

Paul: [37:07] Good you don’t.

Jason: [37:08] So so the first thing that that, like so they’re like I want to unpack that a little bit there’s a couple of parts of that but the first thing is you do, you have like top lines but then you have a bunch of categories and, most of us are interested in a particular subset of categories and if unless I’m mistaken you do us a couple of subtotals right so if we just wanted to talk about Automotive there’s an automotive line if we just wanted to talk about the top line there’s a top line but for example the top line is going to include new car sales it’s going to include fuel sales and it’s going to include restaurants which you guys call Food Services is that.

Paul: [38:00] That is correct yep I was going to I was if your I was going to tell you like the different breakouts we have if that’s what.

Jason: [38:08] Yeah yeah yeah.

Paul: [38:09] Sure so we are we so like on the advanced report for example will have the retail and Food Services total that’s everything with retail and like you just said the right the food service is right you what we will what we will provide and publish is a not seasonally adjusted number and a seasonally adjusted number so and then we’ll you’ll have the level and then you’ll also have the percent change from the prior month. And the percent change from the prior year right and then you also have like the rolling quarter, so like what you were just saying is you will get a lot of different the matter with depending on where you go you might see one that says retail sales are up like you know to tensor and another one says retail sales are up 5/10 right they one could be looking at I had not seen easily adjusted other one could be looking at seasonally adjusted one could be looking at year-over-year what we tend to focus on and what a lot of people mostly quote is the seasonally adjusted month to month, Trend and then whatever the revision was along with that total like you were just saying we also then provide a total that excludes, motor vehicle and parts dealers because that is a category that a lot of people like to see they want to know but since Otto is so big and make such a large portion roughly like twenty percent of the retail total people want to see what would it be.

[39:20] Without, without auto brightness so we provide that number and then we provide a number that excludes guests just gasoline since that’s a big number and you know with everything that’s been going on that goes on with gas and how that’s so price driven, people want to see what retail would be like without that number and then we have a total excluding motor vehicle and parts and gasoline so it excludes bolted oh so you could just get the number retail without those in it and then there’s the retail number so that number is retail excluding Food Services if you do not want to see Food Service it’ll still have the motor vehicle and parts in gasoline in there but it will remove the, the food services and then one last one that you’ll see and it does not come from us but a lot of people in what you might hear something to the effect of the be a control group, and so what that is is that’s retail and food services and they take out Otto, building gas and food services so that’s a number that sometimes you’ll see quoted in various news outlets but it does not come from us we did we don’t have that control group that we use. That kind of clarify some of the categories and the different okay good.

Jason: [40:27] Yeah so the first thing is like a bunch of us could be taking a different segment that we’re looking at and it could be one that you gave or it could be one on calculated myself right guy I can aggregate it up a bunch of the categories and just get gas for example if I want it right and side note.

[40:46] Part of the reason people would take categories out is particularly if you’re interested in e-commerce historically there’s some categories that weren’t very e-commerce friendly like almost no gas is sold via e-commerce so I today there are some, some edge cases where maybe it is but for the most part you know gas wouldn’t be eligible to be, any Commerce sale in the old days a lot of people don’t consider restaurants part of retail I disagree but I get it and in the old days there was no e-commerce for restaurants now there’s a lot of e-commerce for restaurants with door – and whatnot in the old days nobody ever bought groceries via e-commerce, now a bunch of people are buying groceries versus e-commerce so in the old days if you wanted to say like the core categories that were, common for e-commerce you might have take it pulled grocery out and food services and gas in Auto right but today you know Tesla sells online you know grocery is like 12% e-commerce right now or something like that four for each of those things that the definition than a that someone like me might use like to be fair like Hat Wag we have to put more of those categories back in because they are increasingly falling into the e-commerce bucket. But so besides that you hit on some other things that we all tend to use differently you give two sets of numbers seasonally adjusted and non seasonally adjusted numbers.

Scott S: [42:13] Yes so I’ll try to tackle this one. This is this is a popular question as you can imagine so so generally speaking you know the data we get and we aggregate up goes into producing we call are not seal adjusted number. With a couple of caveats one you know the industry much of the retail sector follows like the. The retail reporting calendar right they opt they don’t operate on a calendar month they try to keep four and five week periods together to maintain the times of years and the kind of comparable basis year-over-year so they so they report, you know not for the month we’re measuring so for September or something I think the.

[42:55] You know they’ll report appeared at the moment either going to return home and he’d acre for September was but if sometime in early October right so it’s it was a period that ended in early October wasn’t really September first, XXX. So we have to adjust that to put that on the same basis because not everybody follows that we asked them for their September sales and you know something like the Auto industry doesn’t typically follow the Reach Out calendar the grocery industry has their own calendar, in many cases so so we put all of that on the same basis well once you once you do that and whether you have to do it or not, there’s differences of course in months there’s differences in numbers of days you know from the extreme of 28 February so for this year when it’s leap year the 31 in a number of months, you know if we just kept the data Nazis they just at you would have increases just based on the different number of days in a month. It also have differences in the month or increases or decreases based on the fact of what those days are in the month you know if you have a month that has five Saturdays in it that can be a huge. Increase in certain things like you know just think of a grocery store on a Saturday versus like a Tuesday so if you had a. You know if you had 5 Saturdays in a month it could overly inflate those numbers so that they may not be comprable you know to what you’re comparing it to last month or when you’re even comparing it to last year because the calendar can shift a little bit. So we do we do make an adjustment for assent to that trading day difference we do have some adjustments for things we call like moving holidays this.

[44:22] Both the most obvious example of this is like Easter and the change in spending patterns around the Easter so when it’s in March, the spending is going to fall more in March when it’s in April the spending is going to fall more, an April depending on the timing when it’s in the middle kind of overlaps two months so we make an adjustment for that then of course we also make an adjustment for really what the what we call it which is seasonals the seasonality of the industries this is the fact that you know. In the springtime people are planning a bunch of flowers so they’re going to a lot of lawn and a lot of nurseries and you know so so will we have to do a seasonal adjustment to account for the seasonality, um so so we do all three of those things we do a seasonal adjustment, we do a holiday adjustment were appropriate not every month of course and we do a training day Jasmine so they basically you could put all of the months on the same basis any Theory you can compare any any period you want across the whole time series you can do a month the month a year to year, um you know you can take any quarter and compared to any quarter and a prior year the goal of that is to basically remove. Remove the extraneous factors that might cause the data to be different to hopefully. You know uncover the actual underlying you know growth or decline that might be going on in retail.

Jason: [45:36] Yeah okay and then you also highlighted one other variable that comes up a lot like people will quote a percentage and it’s super important to know whether they mean percentage change from last month or change from the same month last year.

Scott S: [45:54] Yes that’s and you know and I know a lot of the industry follows you know year-over-year as the metric that is tracked and of course A lot of it is same store your every year we don’t do same store we do. Whole store so basically if you if you had 500 stores last year and you have 400 stores this year you’re going to show a drop you’re not going to show a sing store for just a 400 you have, if you and vice versa if you have 600 instead of 500 you’re now going to show an increase, you know rather than just the performance of those 500 stores one important thing to clarify actually that I meant to mention earlier was, we don’t press the dust this data so it is nominal and nature so things like gasoline where it which can be volatile for prices you know that can be, and that’s one of the reasons like people try to exclude gas sometimes because it may be a little misleading, because there is there is no price adjusted done our friends at me yeah they do that when they put it in the GDP they do a price adjustment.

Jason: [46:51] Yeah and I was going to say less frequently but occasionally you will see some US Census retail data that’s, attributed as inflation-adjusted and my assumption is that something that someone did with the data themselves after they got it from you you got you guys don’t, don’t do that right now.

Paul: [47:14] Right we don’t do that it’s just that’s other people you know this thing as you can imagine it’s probably just there’s a lot of people sitting there waiting for the data on that day right and so they get data set they downloaded into whatever program they want to use and they run their own regressions or models and apply things and they come up with their own and you know they might fail to say and whatever article whatever that you see are their tweet or something they might put something there that is of like fact that it came from us but that’s not the truth.

Jason: [47:37] Yeah totally got that okay and then now I’m going to throw out my premise for which data I prefer to use you can react if you want right, so in general if I’m talking about. In absolute number for a particular month so the dollar value of last month’s sales in a particular category or whatever I’m going to use the not seasonally adjusted number, because that’s that’s the actual number that happened in that given month, but as soon as I start talking about Trends over time or comparing two different times I’m going to quickly switch to seasonally adjusted numbers because it’s going to be more accurately reflect. The true true changes from between those two months.

Scott S: [48:31] I think that’s I would probably agree.

Paul: [48:33] Yeah I was going to say the same thing I think would you say makes complete sense using over the time for the season just to take out as everything Scott was just talking about, like the moving holidays and seasonal seasonality of stuff you want to see what the true Trend would be I think that makes sense but from a one-month perspective if you want to see what they’re real, data is and you want to look at the Nazis Lee Johnston I think that makes sense as well.

Jason: [48:54] Yep, and then in terms of year-over-year or month over month I’m for retail and there’s look all this data is there because there’s a legitimate reason you’d want all of it I’m a big fan of year-over-year data because they’re like you know even with the seasonal adjustments the. There are too many things that are just intrinsically different from from month to month in in retail or in most categories of retail so it just it feels to me more apples to apples, to say you know what happened versus the same period of time the previous year.

Paul: [49:36] So what you say makes sense and I think you’re a lot as Scott would probably agree with a lot of a lot of out what you will see from from news outlets out there and and companies and research firms they usually report year-over-year and even when we’re looking at like when we look at our data and we compared it to outside things like what we receive from like black box intelligence for like for Food Services or for things like that a lot of those, places and a lot of those organizations will quote year to year so, yeah I mean it definitely makes sense I think we’re we definitely show that’s why we show both I mean and what we focus in on is I think people want to see on the day of is always they want to see like what happened from last month but I guess you know it’s whatever serves your purpose.

Scott S: [50:19] And I think and I think they’re in the current time if they deal with the with the pandemic in the past few months I think the. You know it’s important to use the month the month and the year to year in conjunction with each other right because some of the trends are so extreme. Bennett kind of it’s easy to get lost in like the size of the numbers going down and coming up. So using it you know in conjunction with the year-over-year gives kind of a checkpoint against Okay so. This is saying for this month it’s up or down X but like what does that mean like where we at compared to pre-pandemic levels or or you know where we at compared to where we were last year and I think I think it’s important to be able to use both. And I think one of the event is one of the you know advantage of having seasonally adjusted data is in theory you could compare you know August of this year to February even though maybe that’s a not comparison you normally would do, you could do it because you know if you wanted to kind of see you know a before-and-after kind of perspective.

Paul: [51:17] And that’s why in our report we’ll all we will supply the month or month and the year to year and then that’s easy Johnson this easily Justin and we’ll put out tweets because like a sky was just alluding to you know you see something interesting like like from last month in August like food services and drinking places were up like four point seven percent from July server was like okay that’s been up like three months in a row now but it’s still down 15.4 percent from last year and throughout the whole year, it’s down when you look through the first eight months of the year so when you look at a lot of these different categories they look good on a month-to-month basis not so good on a year-over-year basis but then some of it is the other way around, where things look great on a year-to-year basis like what we’ve seen with non-store retailers through the through the last 6 months or building materials right or Grocery and I think when people see that. In those categories that makes a lot of sense.

Jason: [52:06] Yeah and so then two other little things just popped up I forgot to mention earlier but but when your, thinking about what categories you want to look at when one kind of gotcha to know is there are two categories that have the word food in them right like theirs Food Service which is a restaurant and and then there’s there’s food retail which might be like a grocery store. Do I have that right.

Paul: [52:31] Yeah so the Naik the industry 445 that is food and beverage stores so that’s going to be your grocery stores your liquor stores your your all those types of all the other grocery stores they’re not restaurants so what’s in the 722 industry which is Food Services, and restaurants is exactly what it is you get things, like all your takeouts and your food your restaurants and your fine dining your limited all anything that has to do with that it’s not a not a grocery store so sometimes people get that a little a little confused.

Jason: [53:03] Yep know exactly and then you just reference something that I’m ashamed to say I didn’t know but it suddenly occurred to me of course you do you guys Tweet stuff yourself like is there a Twitter account we should be following.

Paul: [53:14] We do so we actually have a Twitter account and an Instagram and a Facebook it’s just that US Census Bureau and actually on the so if you follow that on the day of the release we haven’t we call up with our tweets they get reviewed and then they are released on the day that we we put out our report will put out a number of different things will show a different flavor of like our month you know a month-to-month trends that look interesting year over year Trend so we’ll put anywhere but sometimes between eight eight sometimes 10 will do it on our quarterly reports also and we’re going to start to try to get into the business now of, you know not necessarily just on the day of the release but if we have some interesting things that not to overload everybody all at one time to maybe like later in the week or sometimes throughout the month is just to also put on some interesting facts.

Jason: [53:55] Nice and I’m assuming it’s super easy for just anyone in your office to just send out tweets under the government accounts.

Paul: [54:04] It’s the easiest thing you’ve ever seen I mean we can do it so we never get any backlash or anything it’s love you me and Scott do it or sub knocking yeah. As you can imagine yes is it is difficult but it’s something that we’ve been you know it’s gotten a lot of good feedback you know our department the area that runs that it’s nice to see like when we’re like we’re actually think one month we’re actually trending and there was like some some cool stuff going on there and that it’s really it’s really neat to see that stuff.

Scot: [54:31] The so I know where you guys were we’re recording this at kind of a late time so we’ll go through these quickly you kind of talked about how you look at some other third party data do you guys kind of triangulate on that at all or our feed it in so we have people on the show that can they look at credit card debt there’s tons of data out there do you guys look at that at all.

Scott S: [54:54] So yes we’ve actually well first of all we use a lot of third-party data just in the normal analysis and validation as we you know like a sanity check against what we’re saying, but then we also in recent years have been working with a number of companies to try to kind of blend the data between like the third-party data survey data other administrative data that we might have internal to the bureau, to try to produce some some estimates and actually just last week we released a new product that is actually it’s another one of these experimental products that we did for the. For the e-commerce breakouts but it actually is breaking out our monthly data by state. So we’re doing modeled estimates where we are blending data from a third party Source from survey data and from internal administrative data and producing from January 2019 to the present. Um monthly data by State for the non non non store the non e-commerce basically the brick-and-mortar component of.

[55:56] Because allocating the revenue to the e-commerce component that we’re still working out of figure out that methodology so. But and and you know and so we partner with some third-party companies investigate how we could do this the accuracy of their data and we’ve been working on it for a couple years and we were able to produce something and we’re always looking to like, see what other data is out there there’s data is much more available than it’s ever been before and you know we. We’re trying to say when we can leverage without having the burden the retailers which already have enough on their plates.

Scot: [56:31] Cool random question to does is the services that’s not you guys but our do they have Geographic breakout Gina.

Scott S: [56:38] Services does not does not get servers with the services data is on a quarterly basis and of course it covers mean it’s like you mentioned earlier it’s a huge part of the economy so it’s very Broad in nature but no we do not have, Jagger got Geographic Baker breakouts yet but we do that is on our list.

Scot: [56:59] Tell me you know a guy that would love Geographic break out of auto services data.

Scott S: [57:04] I will make sure I went and when when we need it when we need somebody to support us I will come back to you in a little while and you can give a good statement.

Jason: [57:12] So a you need to commit to fill out some good surveys.

Scot: [57:15] I will I’ll be happy.

Paul: [57:15] Yes yes yes.

Scott S: [57:16] That’s right we expect your cooperation if course we couldn’t tell anybody if you were we’re not on the same.

Scot: [57:24] Absolutely and I wouldn’t divulge that either.

Jason: [57:26] Would be it would be totally obvious because Scott is actually the biggest part of the service economy so if it’s a big number you know he’s in there. So in one one side note on the third party data there are ton of other people that like have their own panels or surveys or things and they publish data, what listener should know is almost all of those companies index their data to your data.

[57:54] Yeah so you know if they’re hearing from 10,000 people and they want to scale it up to. They do a correlation and say hey how does our estimates map with the the US Census Data and it follows a closely then they feel like it’s it’s it’s credible. So so when you’re looking at almost any data set in our space there’s a good chance it’s at least influenced by this US Census census, so it’s very fundamental I do want to wrap up we’re almost out of time but just briefly talking about tools we’ve talked about your website a couple of times and I would just plug on your website, that it’s surprisingly functional so listeners should know you can download Excel data you can download a PDF and you know you can load that Excel stuff into or a CSV into whatever Analytics tool you want but you also have like a. Pretty functional kind of click and an answer query engine so you can pick which of these categories you want and what date range and whether you want to adjust it or, or not and you can grab almost any data set that you might want pretty easily straight from the the website.

[59:09] That being said there are other tools that are pretty integrated with your data. So I I know Google has some experi

Play
Like
Play Next
Mark Played