Spatial intelligence is supercharging merchandise and marketing teams. How? By transforming data into dollars. In-store teams are still at a disadvantage when it comes to knowing how their shoppers behave, compared to online retailers who have unprecedented visibility into customer behavior.
Join special guest George Shaw, CEO and Founder of Pathr.ai, and Shelley as they discuss how spatial Intelligence captures data to reveal a deeper understanding of how customers shop in-store. AI-powered technologies can drive store layout optimizations, product placement, and marketing promotion effectiveness.
Find out how Pathr uses a retailer’s existing cameras to provide merch and marketing teams with real-time, actionable insights that help drive key business decisions.
Special Guests
George Shaw: CEO and Founder of Pathr.ai
Transcript
Transcript by Descript:
I’ve always seen spatial intelligence as a foundational technology that will someday be ubiquitous. Everybody who manages a physical space wants to know how the, the users of that space, what, what they do, what, what’s their behavior. You have an office building. If you have a cruise ship, if you have a retail store, whatever it is, you want to know what the people that are showing up at your space are actually doing.
Retail Unwrapped is a weekly podcast hosted by Robin Lewis and Shelley Kohan from The Robin Report. Port. Each episode dives into the latest trends and developments in the retail industry. Join them as they discuss interesting topics and interview industry leaders. Keeping you in the loop with everything retail.
I’m so excited to have George Shaw here. He’s the CEO and founder of Pathr.ai, and I want to just get a little bit of information from you outside the podcast. I’d love for you to tell us. About your background, where you came from, and how you ended up really founding your own business. Thanks, Shelley. Um, yeah, it’s a really, um, kind of a long and circuitous story how I got here, but I’ll try and shorten it up a little bit.
I was always into technology all the way back in the 80s. I was kind of a math nerd. I was into computers like before a lot of people were into computers and then made a left turn when I was 18 and went to art school and started studying graphic design. Right. designing posters, I designed a bunch of movie websites, I was kind of got really into um, into online web design, which taught me a lot about creativity and innovation and how to like create new stuff but within constraints, within business constraints.
Eventually I got sick of that, I just kind of got bored with what I was doing and went back to school, found myself at the MIT Media Lab where I was studying language acquisition, trying to understand how little kids learn to speak. From what we know, it’s mathematically impossible, but every kid does it.
So somehow they all, we, there’s something we don’t know. They all managed to learn how to speak, started studying that. And then from there started learning about tracking and spatial intelligence and how we can understand the movements of people and how that came to relate to. In that case, kid learning to speak, eventually went to retail, eventually kind of found myself here, uh, That’s amazing.
And what I’m most fascinated about you is the length of time you’ve actually spent working the field of AI and specifically AI technology. So you talked about being a math nerd, like where did this passion for tech come from? It’s way too far back for me to really remember. I was really into math and, and you know, kind of just computers and technology and algorithms and, and those kind of things from when I was a really little kid.
I was probably three or four years old when I really started what, on that kind of stuff. And my mom started really young doing arithmetic on my fingers and, and stuff like that. I started really young with that. That’s great. Kudos to mom for that. That’s awesome. And, uh, tell me just one more quick question is, what’s the best thing about your job today?
Oh, I love creating. I love making new stuff. Anything I’ve ever done in my life, anything I could do in my personal life is usually about making something. I don’t like rearranging things or managing things or organizing things, I like making new things. And that’s what we do. We make new software that does new things that have never been done before.
So that’s what gets me up in the morning. That’s awesome. Love it. Excellent job. Okay, now we’re going to start the podcast. And it might sound like you might be repeating some things, but don’t worry about it. I’ll let you edit it. Great. Hi everybody. Thanks for joining our weekly podcast. I’m Shelley Kohan, and we are so excited to have a great guest with us today.
Who’s really leading the AI and spatial intelligence retail technology. And. George Shaw has also been here previously on our podcast. So welcome George Shaw. He’s the founder and CEO of Pathr.ai. Uh, and obviously it’s great to have you on the show again. Thanks Shelley. Really happy to be here. I love joining you guys on the Robin Report podcast.
So we worked together many years ago. I, oh my God, I think it’s coming up on Could it be 20 years or close to that? It’s been a long time. I don’t think we’re that old yet, Shelley, but it’s probably pretty close. I know it’s over 10. So we’ve known each other a long time. And, uh, one thing I learned so much about from you is this transformative role of technology, specifically in retail.
And you were always a visionary, like you were like 50 steps ahead of everybody else. And I always appreciated it. Yeah. And so now all of a sudden the pandemic and all of a sudden all of these retailers are now being more innovative about what’s happening in the industry, especially across different functions from store operations, customer service, even loss prevention.
So, AI and spatial intelligence can really be a key to unlocking some of this success for retailers. But before we dive into that, maybe even give our listeners who may be joining us for the first time a little breakdown, because I don’t think everyone fully understands what spatial intelligence means or what it’s all about.
Sure. Yeah. Happy to give a little intro. Spatial intelligence really at a high level. Spatial intelligence is about understanding the movements of people and things and using digital tools to evaluate, analyze and act on those movements. So in a retail store, that’s typically tracking the locations of people as they move through the space, identifying who’s a customer and who’s a staff member and then creating analytics based on that, where people go, where do they spend their time, so on and so forth.
That’s at a high level. That’s what spatial intelligence is. And then what we’re trying to bring to market with Pathr.ai is a scalable version of that. So how can we do that across thousands and thousands of locations? We have to do it very efficiently. It can’t be can’t cost a lot. It has to be a low capex for retailers.
So the scalability is important, and then how we understand that movement is really important. We’re innovating in that space too, and that’s work that really grew out of sports analytics. Where in sports analytics, what they do is they create machine learning to understand every play in the game that would matter to the coach.
Coaches of NBA basketball teams are not data scientists, but they know how to coach a basketball team. So the machine learning does the heavy lifting and tells the coach, these are all the plays that you might run, These are the ones that were more successful, less successful, given this scenario or that scenario.
Here’s what you should do. Machine learning helps the coaches figure that out. That’s been pretty mature in sports analytics. It’s been going on for a decade. So we’re taking that same model and applying it to retail, where we can find the plays that are happening in a retail store. I love that analogy.
It’s one of my favorite analogies. And you know, it is, you do have to really look at the details of those plays, right? And today we’re talking about something that’s really important and that something that really been missing in a retailer’s toolkit. And if we look at marketers, they really have relied on rating and online methods to gouge, you know, the effectiveness of digital campaigns.
So understanding how digital campaigns are doing through data and analyzing that data. This then allows them to optimize their strategies with their target audience and really have more precision in how they’re directing some of those marketing initiatives. And so if you look at an example like Nielsen, who’s been kind of the gold standard for measuring TV audiences as an example, and online retailers have an abundance of data at their fingertips to understand the customer behavior, where they go, where they don’t go, what they shop, what they put in the cart, on and on and on.
But when it comes to the physical world. Even today, I keep saying this every year, it’s going to get better. But even today, physical world is flying blind. They don’t know what’s happening inside the store. And so let me just add one more point here. There is a renewed interest in physical retail. It’s been happening and everyone is looking at those, uh, creating these environments in the store that engage the shopper, uh, and try to make sure that shopper journey is really been.
More relevant. Yeah, it’s been really exciting these last couple of years to see the resurgence of physical retail. I mean, one of the things the pandemic taught us is that people really want to go to physical stores when nobody could go to physical stores. They were riding in the streets because they wanted to go shopping.
So now that we have this resurgence and this renewed interest, people are going back to stores. Retailers need to know what those people are doing when they’re in the stores. And specifically marketers, like you said, marketers have always enjoyed a ton of data. They, they’ve known what people are doing around their campaigns for a really long time.
On TV with Nielsen, online with Google Analytics and platforms like that, you know. every single move somebody makes relative to the marketing messaging you’re putting in front of them. But if you look at the physical world, like billboards is a good example, but also signage in a retail store, different, different messaging in malls, digital signs are a big one.
People don’t know what’s happening with those. Are folks engaging? Are consumers engaging with my marketing messaging? Is it affecting their behavior? Is it changing their path through a store? Is the signage actually working? Retailers have never known that stuff before. So what we’re doing and lots of other companies like ours, there’s a big movement in this space now to, to build physical digital analytics for the physical world, a tagline that was coined a long time ago, probably when you and I were working together and we might’ve coined that tagline, but we’re finally starting to bring that out into the world.
You know, these digital style analytics, the kind of things that Amazon has been enjoying. For a really long time now, physical retailers are starting to get some of those same kind of analytics. And it’s really exciting. It’s changing the way they think about their store layouts, their merchandising, their marketing, every aspect of how they do business.
Yeah, it’s exciting because spatial intelligence can really impact merchandisers and marketers both. And that’s what I think is really exciting. And retail analytics obviously has been a real game changer, uh, for the physical retail stores. As these retailers are really focused on driving top line growth are also really in working on driving that deeper loyalty with the consumer.
So understanding the behavior in the store while it’s happening are really key for merchandisers and marketers. So really understanding what are the influences along that You know, purchasing journey that they have. So how can retailers leverage the power of AI and spatial intelligence to gain even deeper insights into this consumer behavior?
Well, there, there are ways to understand that behavior using kind of machine learning, the plays that are happening and so on, like we’ve been talking about. There’s also a B testing, try a thing here and try. different thing over there and measure the results. Retailers have done that for a really long time, but they’ve measured the results based on just raw sales stats, and you don’t really know why.
So I did a thing over here and I did a different thing over there and I sold more here, but I don’t know why I sold more here. What went wrong over there? And now with more information, they can kind of fine tune what it is that they’re doing. Like if you’re thinking about the, the, the merchandise mix, different products will have different effects on other products that are adjacency effects.
Retailers have known about this for a long time, but we can help them measure that kind of stuff and see when you put this product next to another product, does it help to sell? Specifically, what’s the consumer behavior around product A and product B and does product A cause them to shop more around product B?
I think things like that, I think are really exciting, you know, to understand that behavior more deeply and then again around around signage around like the route that people take through the store, different way finding and so on is all really, really interesting. Yeah. And I’m sure in addition to the merchandising and the marketing folks is also operations.
So understanding, you know, the zone traffic, where people are going at peak times, that type of thing. Um, the other thing that’s really interesting is that, so you mentioned AB testing and retailers have been doing AB testing for years and years, but AB testing back in the day was, you know, a six month project.
So I think what you are doing, which is a bit differently is AB testing is. easier and it’s scalable. So before we would just have to pick like a handful of locations and run the AB tests. But now with Pathr.ai, you can really scale this to do AB testing across a whole fleet of stores. So tell us a little bit about that, or maybe you can give us some examples of that.
Yeah, that that to me is one of the most exciting things is you deploy a platform like Pathr.ai across an entire fleet of stores. And now maybe you have 500 1000 2000 locations to choose from when you want to do an A B test and they’re collecting data all the time. So you don’t have to choose them in advance.
You can wake up on Monday morning and say, I want to see how the Midwest is performing against California and run that A B test and be done by Monday afternoon. That to me is just absolutely incredible when you’re collecting this much data. Furthermore, you can A B test anything too, and you don’t have to know that going in.
You can wake up on that Monday morning and say, I want to see Midwest against California, how they perform in terms of staff behavior, or is my staff interacting well with my customers, and how does that compare from, you know, the Midwest to California or whatever. Way you want to slice up the fleet point is you can make up any test you want on the fly, run it across in a huge swath of locations or choose pick and choose any locations from your entire portfolio and compare and contrast, which is just absolutely game changing for physical retail.
But if we go back to online retail, Amazon’s been doing those kinds of AB tests since there’s been an Amazon. One of the very first things they started doing was running virtual AB tests. You as a consumer wouldn’t even know if you were part of one of these A B tests, but they’d be collecting the data behind the scenes to see if the red button or the blue button worked better to get you to buy the product, right?
So it’s the same kind of thing. Online retailers have been doing forever now. Physical retailers can enjoy that that same power. It’s really a lot of power that they have. Yeah, it’s power and it’s also being very nimble. The agility is what is key to be able to do these quick real time testing and see what’s working.
And the other area is really how do you monetize this? It’s great to get the information to have the data, but you know, how do you actually Build this into a strategy that either optimizes campaigns, drives higher engagement, or helps the retailer kind of monetize the information that they’re getting.
So I know a big conversation, for example, store layouts and optimizing store layouts, and what’s the best store layout. Can you talk a little bit about, are you able to kind of solve that issue of how do you look at data to help you optimize the store layout? Absolutely. Yes. Store layouts is a is one that we’ve been working on for a while on understanding the flow through a store and how you can optimize that flow in order to sell more products.
Ultimately, that’s your that’s your goal in a retail store. Unlike on a basketball court, you’re you’re analyzing the flow and you have a very different goal, right? But it’s the same concept. It’s the same. The same principles apply understanding the flow through the store and where there might be blockages and where there might be opportunities to reroute.
Mhm. people and, and so on and so forth. So there, there are a whole category of things. There’s a, there’s a, there’s a big list of things that, that we can do in order to help retailers make more money in their own business. How do you optimize your business? We can help you save money. We can help you increase your top line and do all those things.
But, but you mentioned monetizing the data too. And there’s a, there’s another category of, of use cases that we can do that helps retailers or others. Directly monetize the data, which I think is also really, really interesting where they have information about how people are looking at advertisements, for example.
So if you think of a retail media network, how do you monetize the retail media network? You really want to do it the same way online retailers do, or the same way, uh, TV advertising works. You know, how many people saw my ad? Well, that’s how much I’m going to pay for that ad, right? Just the way Google monetizes AdWords and stuff online.
And the only way to do that is by measuring the movement of people around the store. So if a retailer has a retail media network, they want to go back and charge the advertisers. They’re able to now charge by click rather than charging by how many times the ad runs. They can charge by how many times people see the ad.
Which is brand new in the physical world. It’s super exciting. It’s so exciting. And of course you did bring up one of the hottest topics out there, retail media networks. It is just growing exponentially. Um, it’s crazy how many, how much retailers are spending and trying to drive revenue with retail media networks.
And it makes perfect sense, right? They have the ecosystem at their fingertips to do that. So I really appreciate that example. And something I should have mentioned earlier, cause I know you use the NBA example, like, but you actually work. With the NBA to develop some of the spatial intelligence you worked in banking.
Um, and now you’re taking all this knowledge and applying it to, uh, retail technologies and what’s happening in the physical retail space. So, can you tell us more about Pathr.ai and maybe how AI can play a role in the analytic solution? Yeah, so we are we are taking a lot of cues from the NBA. We’re taking cues from soccer sports analytics, you know, sort of across the board has been just a massive influence on on every every other type of analytics.
All these sorts of movement analytics. We’re taking a lot of cues from them and applying some of the same AI. principles. Some of the tracking engines that they use were applying some of the same principles there, kind of building up the technology in that way. But the biggest area where we take cues from them is in understanding the tracking.
You do the tracking, you know, people have probably seen visualizations of tracking dots moving around a map. Once you have the dots moving around a map, and what do you do? And that’s where we’re really taking cues from other industries and understanding how to evaluate Okay. how those dots move. The most exciting example I have right now is in loss prevention.
We’ve actually built a shoplifting detector. So just based on dots, picture dots moving around a map, and you as a person looking at those dots, you have no idea what the heck the dots are actually doing. But the AI can go in and understand this one particular dot is not like the others. It’s doing something wrong.
It’s doing something suspicious. It’s not actually shopping. Raise an alert. And that’s the same kind of machine learning that that’s the same kind of machine learning that would be able to say, you know, this is LeBron James here and he’s about to score a three pointer. It’s the same machine learning that applies that says this is a suspicious customer.
They’re probably about to shoplift. So we’ve taken those principles from sports as well that that kind of elevate the AI to understand human behavior a little bit more deeply. And then there’s some really cool use cases that come out of it. Wow, that’s so interesting. I think you call the, uh, bad actors, right?
Bad actors. Is that the right terminology? Yeah. Okay. So when you see, when you have what you just described, so if a retailer has a bad actor, you know, can they actually action on that? Like, what is the action of that being notified of a bad actor? Like, This is this has been one of the one of the most challenging questions, especially for a nerd like me.
I can I can make computers do all kinds of things. But this is ultimately a human problem. It’s not really a technology problem. Once you know that there’s somebody shoplifting, what is it that you do? And yet, so you have to staff up accordingly, you have to have the right people on the ground in order to act on that you have to have systems and processes in place.
Typically, our alerts will go to a remote operations center. And then they’ll take action on it from there, either to notify local law enforcement, notify local loss prevention, notify store staff, or just capture the analytic that that’s really common too, because a lot of retailers are just not seeing all of the shoplifting that’s taking place.
And so just us pointing it out. And saying 5 times shoplifting happened this week. And here are the examples that’s already valuable to them. So in a lot of cases, they’re just kind of capturing that and understanding the shoplifting and then using that as an analytic rather than actually taking real time action on it.
Yeah. And you know, what’s interesting about that is, um, so you’re making. You know, shoplifting, you’re being more proactive about it as a retailer. So instead of waiting six months to a year to get the inventory results that say, Hey, a bunch of stuff was stolen six months ago or a year ago, and here are the top stores, uh, I believe you can have real time visibility and oh my gosh, there’s a lot of activity in these stores and then they can create strategies around that is that I say that in a fair way.
Absolutely correctly. Absolutely right. We, we, we can even do one better. In a lot of cases, we can identify the suspicious behavior, the bad actor, if you will, before they’ve actually done anything wrong. So we can watch their movement pattern and say, well, that movement pattern implies that they’re, they’re, they’re, yeah.
They’re casing the joint, so to speak. They’re getting ready to do something wrong. And so the retailer can take action on that. But really importantly, the retailers have the comfort of knowing that our system has no sort of PII in it. There’s no personally identifiable information in there. There’s no demographic information in there.
Age, gender, ethnicity, we don’t even collect those. So it doesn’t exist in our system. So when we say that this, the system has identified a bad actor, we, they know for sure that it’s unbiased. If there’s no chance that there’s anything personal that’s gone into that, no demographic information that’s gone into that assessment.
So that can give retailers a lot of comfort and allow them to kind of take, take more action on those things. That’s very exciting, very exciting. So I’m going to ask you to pull out your crystal ball and looking ahead, you know, where is spatial intelligence going and what kind of impact do we see on merchandisers or marketers?
I mean, there’s all kinds of things happening in our retail environment, new technologies, consumer behavior changes. They just changing daily now, practically, you know, what’s your crystal ball telling you? I’ve always seen spatial intelligence as a foundational technology that will someday be ubiquitous.
Everybody who manages a physical space wants to know how the users of that space, what they do, what’s their behavior. You have an office building, if you have a cruise ship, if you have a retail store, whatever it is, you want to know what the people that are showing up at your space are actually doing.
So I think spatial intelligence becoming ubiquitous is really, we’re on that road, we’re seeing it more and more, and we’re going to eventually just see it everywhere. It’ll be obvious that people are being tracked all over the place. So I I think that’s happening. You know, we’re helping to push that.
There’s lots of other companies that are helping to drive that figured out ways to efficiently do this operationally sound methods of of tracking and so on, um, deepening the understanding of different behaviors is another area of big development. I talked a little bit about that, but that’s that’s really exciting.
You know, I imagine One up from shoplifting. The spatial intelligence is able to figure out this is, this is the person who’s bound to buy something if we show them the right offer, I mean, you could conceive that kind of stuff out using spatial intelligence, but the key, and this is really for all of my, my peers in the spatial intelligence space is that we build all of this technology in a way that protects consumer privacy, that we’re going to shoot ourselves in the face.
foot if we don’t. And so it’s really the responsibility is on us and it’s in our own best interest to make sure that we’re being very, very sensitive to consumer privacy and that we don’t collect P. I. I. For example, that’s why Panther doesn’t do that. And I think those things are really key to protecting, protecting the general public’s trust.
Absolutely. And I’m glad you brought that point up. What I find so fascinating about you, George, is that all this hype of AI that hit the industry last year, you’ve been literally working on using testing AI, I’m going to say probably for more than a decade. Right. You were a pioneer in the space of spatial intelligence, and you’re just so far ahead of everybody else in the field in applying this type of AI to the retail industry, and specifically to physical spaces.
Thanks, Shelley. I try really hard. Innovation is really what gets me up in the morning, so it’s nice to hear that I’m still out there in front. Oh my gosh, I think so. And I’ve learned so much from you from the day I met you, you were always like way over my head on this. Um, and I had to bring it down to my level of, okay, now how does this work in retail?
So I would have never learned anything about retail if I hadn’t worked with, I know a little bit about algorithms, but I’ve learned a lot about retail from working with you and folks like you. That’s so fun. That’s why you make a good team. So such a pleasure to have you on Retail Unwrapped. Thank you so much for coming.
Thanks, Shelley. Always a pleasure to join you. I appreciate the conversation. And for our listeners, uh, next week’s podcast is going to be a hot, hot, hot topic. I have one of our writers, Eric Weirson, who is an Emmy award winning television producer, former CMR media advisor. To New York mayor, Michael Bloomberg, he advises corporate clients on communication strategies in the United States, Latin America, Africa.
And also he writes for CNN on the political beat. And our topic is going to be the elections impact on retail. They’re not going to want to miss it. So also one last thing for our listeners. Now, if you go on to the ROM Report web website under Retail Unwrapped, you can actually give us feedback on our episodes.
So don’t be shy, tell us, you know, what you’re thinking. And thanks so much for listening. Thank you for listening to Retail Unwrapped. We’ll be back in one week with another podcast. Please subscribe on Apple Podcasts, Spotify, or any podcast service. If you have questions, ideas for a podcast, or anything else, please contact us via the RobinReport.com.