Video Deep Dive: Attribution Models for Enterprise Local Businesses

By March 13, 2019 May 24th, 2019 No Comments

localu-deep-dive-1200px

This is our Deep Dive Into Local from March 11, 2019. In our Deep Dive series, we take a closer look at one thing in local that caught our attention and deserves a longer discussion.Today we're lucky enough to have Brian Smith from RIOSEO join us to discuss the different ways to create attribution and attribution models, focusing mainly on enterprise businesses, but a few tips for SMBs are discussed as well!

If you have a special topic you would like us to discuss for the Deep Dive in Local, please reach out to us. If you would like to be on one or the other of our segments, reach out and send us the topic and your availability.

If you are interested in sponsoring this weekly show also please let us know.
We'd love to hear your feedback - so chime in on Social Media or by commenting below!

Mike: Hi, welcome to Deep Dive. This week we have the pleasure of having Brian Smith, who's the Director of Local Strategy & Data Partnerships at Rio SEO. And we're going to be talking about local attribution beyond the world of ads. So Brian, welcome. And maybe you could give people a little sense of who you are and what you do at Rio, and the role you play in developing attribution models for your clients.

Brian: Absolutely. Thanks again for having me. I do appreciate the invitation and the time we're taking today. So myself, a little bit about. So I've been in the search marketing space for about 13 years, found my way into it via market research, went into mainly usability studies, and then found my way...or at least a home into a local agency here in Denver for several years, worked for another startup company, which was specifically around location management for enterprises here in Denver. They were purchased and acquired by another software company and I landed at Rio, a little over a year ago, leading up their local strategy, bringing individual solutions to help our consumer base out to understand...you know, search the consumer journey, and then ultimately, the path to purchase. And as well as to expand our data partnership programs beyond just the main typical folks of like Google, and Bing, Facebook, and Foursquare to make sure that the NAP accuracy and partnerships prevail and provide our clients success.

Mike: Cool. So could you define local attribution, both broadly and then narrow in on what it means to enterprise, and really to any business in the local space?

Brian: Yeah, absolutely. I think the holy grail as we as search marketers and as well as what we would consider the offline or online to offline attribution is any way for us to be able to track the effort or medium for us to be able to get in front of those consumers to be attracted to our physical product or services, so forth to drive them in store and then that trackability in between, much easier to be solved by a lot of programmatic solutions that are out there. And I think we'll cover some of them today. Not so easy when you talk about not having some of those solutions. You know, you had mentioned on the earlier upfront portion, oh, this is really outside of the ad space. And I wanna make sure that that point is actually understood is that this is... You know, there's a lot of other technologies that are actually currently implemented if you're paying for ads. I wanted to take a specific approach on what solutions might be outside of there and specifically around that.

Mike: So, I mean, I assume we're going to move towards some sort of model of attribution. I mean, I can't imagine that even with specific geolocation of individuals that we actually know specifically down to the individual level with a high degree of confidence. So I assume there's a model of attribution here, where we take location data and merge it, look at it in the context of query data, and GMB data, and other signs, right?

Brian: Well, I think the model comes from kind of what the metrics are that you're using as a specific organization and what's acceptable for you, right?

Mike: So what are the types of metrics? I mean, store visits, purchases? Where do you draw the line and what is typically for your clients? What you see as the...

Brian: Yeah, so we have clients that kind of run the gamut between having any type of store visit from a geo-targeted fenced perspective to redemption of coupons that are internal for them, that are store specific or user specific. We have the availability that are just to turn around and look at post transactional, and look between cookie-ing of that individual, and seeing if that individual showed up, right? So it runs the gamut among enterprises. I think mostly for SMBs, the easiest thing that one might be able to do would be couponing. And we can certainly talk about that here in a little bit.

Mike: So broadly, the types of attribution then are coupons, visits, and then matching up users from searches to purchase?

Brian: Yeah, absolutely. So if I was to really categorize them into four individual bigger chunks, I'd put them mainly around attribution, kind of purchasing online and picking up in-store. Second would be...

Mike: So that's a relatively easy one to track, you know, obviously.

Brian: Absolutely. Yeah, you know, you have that availability to look at the transaction, the consumer when they picked it up, what they specifically bought. And then there's certainly the targeting of online couponing, which would be either store specific or user specific. You have the opportunity to look at post device tracking or geo-tracking.

Mike: So the couponing, where would the coupon exist? At their website or in Google Offers or where Google post offers or somewhere else it exists?

Brian: It could exist at either, right? So, Google Offers are certainly one way for you to be able to post it there. One of the things that I would say about the negative thing about couponing is, is that it can be easily picked up and shared, right? So that level of attribution, if it's a store specific, or if it's a general enough where you're posting it to one place, could be easily picked up at like RetailMeNot and shared. So that's where you maybe wanna get a little bit more specific and look at the availability of posting it on a specific store that would be user-specific and generated, right?

Mike: So we got those two and what are the other two?

Brian: Yep. So the next one would be, I'll put in a classification of post device tracking or more appropriately a dub geo-targeting, right? Tagging a specific individual, grabbing their device ID and then using a third party vendor to understand the tracking of that specific individual and the movement between the individual places. So you can conquest a specific area, see if that specific user went into that individual location, give some parameters of thing, they have to have a minimum number of visit of, let's say, 15 minutes, 30 minutes, 45 minutes and say, all of those users then back out, give me those individual users. And then you can use some attribution, things that are widely accepted within the industry. I know Google's posted several attribution things around individual vertical specific, like retailers, anybody, you know... I think they say 20% of all individual visits to a specific store ended a transaction, right? It's certainly vertical specific, but then you can start using those to back in numbers of average orders values for that specific location.

Mike: So this data, this location data I assume comes from apps, not from say, Apple or Google, right?

Brian: Yeah.

Mike: And what percentage of users have apps that share data back with you or give their apps permission to share data back with you that are active? Like in my iPhone, I'm always asked, do I wanna share location data all the time with this app. Right, who wants to do that? Or just when I'm using the app, obviously, I'm not using the app all the time and, you know, I use Lyft and Uber very rarely, so how does the various permissions, the different phones...? I assume location is a lot more accessible on Android than it is on iPhone, I don't know that, is that the case? But then how accurate... What percentage of population can you actually count on having that and are you able to extrapolate from there against the whole population, I guess?

Brian: Yeah, so you're absolutely right. There can be two individual ways to boil that down. It can be post-device tracking in a typical larger user-based app that you can tie into, pull that user data off, and then put it into, you know, some type of geo-targeting platform, or you can actually cookie themselves and use a solution that would be post outside of that app, right? There's individual... You know, I don't wanna name any specific ones, but there's a great company that's out of Seattle, there's another one in...you know, several in New York. They give you that opportunity to be able to pull that device ID and track it, giving the leverage to be able to look at that individual going to that specific store.

Mike: So I guess, how do you deal with the privacy concerns in this? I mean, do most users realize that they've given you this level of understanding about their movements? And then the question is, is it cross-device, is it iPhone and Android? And then it's mostly app-based. So somewhere along the line, I've downloaded an app that I've given too many permissions to, that you're able to build a model that... Again, I'm trying to understand, is it 20% of the population does this, 40% of the population? It's like, how big is the actual data set compared to the model that you build?

Brian: Yep. I think it's dependent upon...you know, the privacy concerns are absolutely out there. And I think California is the state that's taking the largest step towards that. EU is certainly setting standards for us, for how we can act appropriately with that. Certainly obviously concerns, and if you're cookie-ing individuals depending upon where the legal law stands, you need to actually make sure that they're aware of it, either through your privacy terms and conditions, or either by a pop up and letting them know that they're specifically there. That's how you get around that, in terms of cross-platform compatibility, absolutely. So if you're looking at it...you know, I look at it from my clients, we're not mainly doing it from an app-based but we're mainly doing it from a visit base. So if that specific user is going to one of our local landing pages or store pages, we would cookie them, be able to pull that device ID and then utilize that to be able to provide and partner up with another individual company that can look at that geo-questing or geo-targeting area?

Mike: Do many of your clients have some sort of in-store Wi-Fi where they're able to do that themselves, or does it need a third party in this context?

Brian: Typically, in the clients in which that we're working with, they're not quite there. They have internal Wi-Fi, but they don't have the internal tracking to be able to look at a user base, user base, or at least have some type of app that they actually are triggering and targeting a specific user.

Mike: I see. So clearly the motivation for Target or anybody at Walmart is to get the app running so that they can then specifically know and build their own internal logic rather than relying on these brokers. So how accurate is the broker data in terms of real-time or near real-time assessment of movement?

Brian: It's a great question. So, and the partners that we've worked with, they give us estimates of anywhere between 75% to 80% of the market, at least that might be, you know, drastically over-exaggerated. But, you know, they say that they can extrapolate from there out to any attribution model, right? So that's been a pretty, pretty interesting figure for us to be able to go off with and be able to provide those stats and power of recognition to our client base.

Mike: So what kind of conversions do you see in a coupon environment from...what's the funnel look like in terms of percentage of exposed to, downloaded, walked in the store, used it, for example?

Brian: Yeah, you hit on a good point because it's like you look at couponing, right, anywhere between, you know... I've been in the consumer CPG space in my earlier days and you look at couponing, it gave me range anywhere between half a percentage up towards 6%, depending upon the driving motivators. So depending upon what your offer is and how compelling that is, it really drives that from a couponing perspective. From a conversion base, we see anywhere between 18% to 25% of some of the consumers that we're tracking, and on the back-end and being able to showcase that ROI on the back-end. So if they get to the location, we're seeing that that 18% to 26% or 25% are actually converting..

Mike: I see. So 18% of the people who actually downloaded the app if they make it to the location, which is some percentage, 25% of those are converting...

Brian: Yeah, I mean, what Google says, I think it's somewhere in the mid-70s or maybe 80s or maybe misappropriating a stat there, but that go to a specific location from a GMB listing or from a location, right? But then we use that metric to be able to back into what our attribution models are.

Mike: I see.

Mary: For a long time, beacons were like the new crazy, you know, everybody wants one thing, and I haven't heard anything about them lately. What is your take on the future of beacons?

Brian: Yeah, you know, it's a great, great thing. And, you know, it's one of the things that I've, you know, in the post that I put up on Search Engine Land around attribution modeling for enterprises, I explicitly kind of took beacons out while they're still an available technology and incredibly useful for Google or Yelp to send to the small SMB market, it's an incredibly big barrier to entry when you start talking about enterprises and being able to implement them across 2,000, 3,000 or 5,000 stores.

Mike: That each have 50,000 square feet.

Brian: Exactly. So it's very hard.

Mike: So it's just not that many stores, it's that many stores across that much geography.

Brian: Yep. So it's easier for them to go out from a solution base to either utilize their own in-app or to be able to cookie and use a third party provider to provide this transactional information.

Mary: Great. Thanks.

Carrie: I actually...

Mike: So do you see that changing over time though? I mean, do you see beacons as a slow burn into enterprise or is it sort of a dying technology? Because ultimately, the original vision was, you'd be able to know, not just that they were in the store, but that they were in front of X display.

Brian: Yeah, be able to do the full print and be able to look at where their traffic patterns are, etc. But, you know, I don't think it's necessarily a dying technology. I think it's a technology that suits a purpose for a certain market. And I just haven't seen that the enterprise market be satiated by what it actually provides.

Mike: I'm sorry, Carrie, go ahead.

Carrie: I had a question. You had mentioned when we started Brian about the online purchase to offline pickup kind of attribution model. And one question I have about that is, do you have any data or any idea if that tends to lead to an increase in purchase? So like, I bought one thing offline and I went to the store to pick it up and I ended up picking up four more things. Like, is there any data or any information out there that supports that?

Brian: Yeah, that's a great question. You know, we do utilize a partner that helps us out with that ROPUS or BOPUS idea. I don't have...

Mike: Could you define ROPUS or BOPUS for the audience?

Brian: Yeah, thank you for that question. So reserve online and pick up in-store, or buy online in-store pickup, right?

Mike: As opposed to Butkus, which is zilch, nothing. Got it. Thank you.

Brian: But I don't have any stats at my fingertips. I do know that it leads itself to an increased average order value. I don't know what percentage that actually would attribute. There's a lot of great technologies that's out there that provides that local store inventory and as well as provides the insight based on what's being searched on at any specific moment, you know, the hot things that are actually being viewed at, you know, being able to provide recommendations based on weather patterns, etc. So there's a lot of great technology that's out there that gives you advantages to increase that. But sadly, I just don't have the specific lift on what it provides.

Carrie: Sorry. And I have one other question. How much of this process of tracking stats with an email? Like is it all email generated? Like, that's how they know about the app to download it or that's how they know about the coupon to use it? Like, is that like step one for any businesses start collecting emails in this whole model?

Brian: Yeah, so that's a great question. You know, at least with Rio, we aren't heavily intertwined with the email marketing programs. We look at it from a local location-based perspective on any entity pages that we provide, meaning, is it a locator page? Is it a local landing page? Is it specifically attributed to an individual? And sadly, you know, I think we as industry professionals can do much better than this, instead of looking at last click attribution. We all understand there's multiple individual mediums that people go through before they actually transact. But we do utilize the local landing page if it's a new specific visit and utilize that as a last click attribution, and look at that as a point of follow-through from there.

Carrie: Gotcha. Okay.

Mike: So what percentage of your clients use these different models: BOPUS, couponing, geo-fencing, cookie-ing? Those are the four, right?

Brian: Yep. Yep. So I would love to say that there was more that was out there, but we have a relatively small few that actually utilize it. I would say anywhere between 3% to 5% of our client base are actually actively utilizing it or at least in active conversations.

Mike: Any attribution model for...

Brian: Any attribution. Yeah. It's something that I think that is so...

Mike: And do you see the barrier as cost or technology or both?

Brian: I think it's a combination of both, right? You know, one of the things that I like... And you know, we talked about the privacy concerns, but I love the post transactional understanding of consumers, right? You can target, you can track them or more better appropriate, you can cookie them on your local landing page, take that anormalized user ID, share it with what is more appropriately dubbed as a DMP, take that information.

Mike: DMP meaning?

Brian: Data Management Partner, right? It's looking at being able to look at consumer psychographics, demographics, purchase behavior, all the things in which the privacy concerns or privacy individuals that are concerned with privacy are actively out there trying to fight, right? I think the thing about this is, at least when we're talking about privacy, or at least this post transactional is that it's very big brother, but it's very insightful to us as general marketers, right? We can turn around, we can anormalize a consumer ID, give it to a DMP, have them provide us consumer information on the back-end, and as well as that positive confirmation of transaction. You can start to turn a map out, who my user is, what are they like, you know, social attributes. You can use it beyond within the enterprise, beyond your local...

Carrie: Build the persona.

Brian: Yeah, totally. You build your personas on who's actually purchasing your specific product.

Mike: Although, in this case, we can actually go lower than the persona, right. We can have many more personas because we have much more granular data. So if it is persona marketing, it could be a very small persona group.

Brian: It can be, so...

Carrie: Like conversion rate.

Brian: Absolutely. And you can start utilizing that for ad targeting and start utilizing that for content development. If you're sophisticated enough, you can change your website based on the individual users that are coming there if they fold into that specific persona, like, there are so many things that you can do from a technology perspective. But, again, that's a costly solution, you know.

Mike: So there's this question here, is there a preferred solution over another?

Brian: Yeah. I would say if cost was not an issue for them, the two things that I would do would be, one, to geo-targeted, you know, geo-fencing, or at least post-transaction...well, not post transactional, but it would be post-device tracking or geo-tracking combined with what I would also consider post-transactional data. There's a limitation though, right? At that point, we're looking at big data and from what I've seen, and Mike, I know you're an entrepreneur, this would be a great venture for you, is that there is no...

Mike: Oh, just as a note, the only reason that I start businesses is because I am unhireable. It has very little to do with that fact I like to start businesses, it's that I can't work any place else so that's the reality in an entrepreneur. Go ahead, sorry for interrupting.

Brian: All right, so those two paired together are incredibly powerful. But the limitation is that you have to literally have a team of business intelligence folks that are crunching these numbers for you to provide the insights between those that visited, those that transacted, and marrying the data between the two to look at what is the average store visit for that specific consumer, for that specific locality, for that individual regional area. And then start blowing out those psychographic, those demographic financial income, all of the information you're getting from the GNP, [SP] that's the biggest limitation that I see. If you have a BI department and you're willing to turn around and spend the money, those would be the two options that I would say to go with, provide the most level of insight.

Mike: So clearly, money is a big deciding factor. So how should one approach this decision as an enterprise to make decision whether they can do this or they can make at least inroads into this? And then how are they implemented?

Brian: You know, the first thing that I always utilize and when I'm consulting with my clients is, what are the metrics? What are you looking for and what is considered acceptable to your organization? If it's, literally, I'm fine with the lowest cost denominator and understanding, am I getting a redemption rate of X percent of a coupon? Then go with that. Or, you may have a CMO that's saying, "No, I want hard physical transactional data and I need that," that's going to drive your decision to go into whichever direction. Cost is absolutely a factor. But I always say, start with your metrics first, identify and make sure that those are believable and bought in by the organization, and then use that to define your strategy.

Mike: So Mary, and Carrie, do you have any other questions?

Mary: The only thing... You know, I have a soft spot for the actual small local businesses rather than enterprises. And it just seems to me like this whole idea of being able to use data management systems that can anonymize this data for you and give you back personas, it just gives enterprise businesses a huge advantage if they will take advantage of it. Is there anything that you can see small businesses doing that might help them kind of fight back against this?

Brian: In terms of the level of intelligence, no, but in terms of what they can do to provide that granular level of insight, I think, you know, Mike and I had talked about this, about being able to utilize like Google Offers or you know, put an offer on Google Posts, be able to turn around and look at what that consumer redemption rate is. And then also as well, you know, I think what you guys do from an education perspective in your seminars is educating them and understanding, you know, what are the things that you need to do from brass tacks to the more advanced tactics is including that. And then giving them the understanding of like, couponing is a completely acceptable form and low cost barrier for entry to be able to provide some really good, usable, and actionable information for them. Unfortunately, I think a lot of the stuff that we talked about, beyond that, is going to be somewhat price prohibitive. Yeah.

Carrie: They might look to their associations, though. Like, for example, if you're a plumber, but you belong to a larger association of plumbers, the associations might have the data or the business intelligence to at least help them build personas to market to. So it might not be available to them on a small level, but a lot of these really small guys belong to larger associations. Like, my husband's part of the painters and decorators alphabet soupy stuff, but he gets a lot of data from them about who buys and how much they spend on average in those pieces of the metrics. So they might be able to look to that avenue as well, not individual to their business or market, but maybe individual to their niche.

Mike: And I would add that there's two things that a small business has advantage of, Mary, and one is that they know their customers, and two, they can directly survey them. And I think where the breakdown is, and for many small businesses, they don't take the time to ask directly of these customers who they know, how that new customer found them, right? So even if you just do it for one week, once a year, you can build a model of attribution that helps you develop both a persona as well as a path. And I think that, you know, it just requires a discipline to do that in some way. And that's difficult. I get it. But it's not that difficult, you know, to put in place a process to ask people how they found you, right? And from there, you can get as much attribution as brands spending gazillion dollars here.

Carrie: Right. And I think people are getting so used to providing their email address for anything anymore, that even if that's all... Like, if you have a tech that goes out and services something, or whatever, even if that's the only piece of information they bring back to you in the office, that's a huge opportunity for you to not only attribute future sales, because you can put them into a really simple CRM, which might be a spreadsheet. You can use that to ask them how they heard of you and to ask them for that review, and to interact with them, to build that relationship. I think that email address is a huge piece of how the small guy can compete with the big guy, just create that conversation.

Mike: So I'm going to have to... I have another appointment, I apologize. But let me just give Brian a chance to close this with sort of a summary and sort of high level sort of summary of what you've talked about.

Brian: Sure. So I just wanna say one last thing, and I think that this goes for the SMB that might be a bit more digitally savvy, but you can get into the analytics. And you can start to look at those special interests that Google is actually providing you. And you can start to segment those individual users. Now, if you're depending upon, like, if you're a lead gen site or you're a e-commerce, it's much easier for you to be able to look at that transaction. But you can use that to back in and funnel with those specific users to give you those personifications and use that for marketing purposes. With that, there's a lot of technology that's out there. And I think the biggest thing I want the people that are viewing this today is, is be consultative in nature when you're specifically looking at attribution. That is absolutely key. I've seen so many things go awry, or go askew because there hasn't been organizational buy-in on those metrics. And everybody agrees on that, because once that metric is called into question, everything else becomes called into question and it's not actionable. It's not usable. And at that point, it's just going to be that continuous conversation of, is this accurate or is this right?

Mike: Right. As opposed to agreeing that we know that there's this inaccuracy to start but we're roughly in the ballpark, let's go there and get what we can out of it.

Brian: Exactly. So you know the consultative nature is huge and then don't be afraid to ask questions and be bullish around your specific opinions. I've worked in agencies, I've worked in-house, I've worked in technology, if you believe in something, make sure that...you know, articulate the clear points and push for your benefits and what you're seeing for the organization.

Mike: All right. Well, with that, I wanna thank you, Brian. Brian, just a reminder is the Director of Local Strategy & Data Partnerships at Rio SEO. Rio was a sponsor of Local U at the last Local U event, which we're grateful for, but that and this have no relationship. We just saw Brian's article and we thought it was interesting, an interesting topic. So anyway, thank you for joining us, and thank you Rio for being actively involved in the local industry.

Brian: Absolutely. I appreciate the time. Thank you.

Mike: Thanks.

Mary: Thanks Brian. Bye-bye.

Mike: Bye-bye.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.