Edge has many different stakeholders, applications, and needs, and this is especially true in distributed retail environments. This episode of Utilizing Edge features Simon Gamble of Mako Networks talking with Brian Chambers and Stephen Foskett about the complexities of technology at the retail edge. The key according to Gamble is segmentation of brands, franchisees, and technical applications. In some cases a single location might even include multiple separate companies or tenants under the same roof. Video, sensors, IoT, and location-based services are coming to retail locations as well, and some of these leverage outside service providers as well. Although sharing infrastructure is desirable, segmentation and security is key. Retail edge environments are increasingly complicated, but there are many ways to consolidate, converge, and standardize to make them practical to implement.
Hosts and Guest:
Stephen Foskett: Welcome to Utilizing Tech, the podcast about emerging technology from Gestalt IT. This season of Utilizing Tech focuses on edge computing, which demands a new approach to compute storage, networking and more. I’m your host, Stephen Foskett, organizer of Tech Field Day and publisher of Gestalt IT. Joining me today as my co-host is Brian Chambers.
Brian Chambers: I’m Brian Chambers. I am the head of enterprise architecture practice for Chick-fil-A. We do a lot of work at the edge, and I write about technology at BrianChambers.substack.com.
Stephen Foskett: And as for me, you can find me as a host, of course, of Utilizing Tech, but also of the On-Premise IT podcast as well as the weekly Gestalt IT News Rundown. So, Brian, you and I participated in Edge Field Day recently. We also, of course, have been talking about planning for this particular podcast series here, Utilizing Edge. One of the things that came up is the fact that the edge is but one of the things that makes the edge, especially retail edge, very different. It’s not that sort of regimented top-down traditional corporate architecture. In many cases, you get different people, different demands, different constituencies, different applications, all sorts of differences in terms of who’s doing what and why. We’ve got to figure out how we deal with all these people, right?
Brian Chambers: Absolutely. I mean, I think the edge has a whole different group of stakeholders who are dependent on its services. On top of that, there are a number of people that we were used to having available in an IT environment in a data center. Even people who work on services in the cloud were usually more developer-accessible, that we just don’t have at the edge. So, in my environment at Chick-fil-A, the people we have in our environment where the edge lives are people who work on prepping food and cooking things and serving customers. But we don’t have anyone there who has that traditional IT skills that we don’t have system administrators, and it’s not really viable for us to do a model where we have to deploy them across 2800+ restaurants across our chain in order to resolve issues. So, it’s very different and brings a whole new set of challenges, and that’s different as a result.
Stephen Foskett: One of the things I think that came up at Edge Field Day when we were talking to Mako Networks was not the necessary specifics around the networking, which is great by all means, but the fact that you need different connections, different solutions, different ways of connecting with different applications back to, I guess, different motherships. And then it’s all very, very confusing. That’s why we wanted to bring in Simon Gamble from Mako Networks here to talk to us on Utilizing Edge. Simon, this is one of the things you brought up in the Edge Field Day presentation. Welcome to the show.
Simon Gamble: Thanks for having me, and thanks, everybody, for joining us today.
Stephen Foskett: So, as I said, you talked about this and the interesting angle here is that it wasn’t really a technology angle, so much as an organizational thing – a unique aspect of retail at the edge. Could you tell us what you mentioned earlier and then let’s discuss it?
Simon Gamble: Distributed retail enterprises have their own unique challenges. They are often very large organizations and frequently franchised into franchise groups. These groups themselves can sometimes be as big, if not bigger than the main retail brand. They have their own desires and wants for the technologies being deployed at their edge, which may conflict with the desires that the brand has for the technology at the edge. Being able to accommodate all parties in these pretty complicated environments is really important to have a successful technology deployment. It is necessary to give people the flexibility to use the technology for the purposes they wish while still adhering to a brand standard that is primarily aimed at predicting customer data at the edge.
Brian Chambers: I completely agree with you. I think it’s a very interesting dynamic, and it probably depends a lot on the business that we’re talking about. If you have a big franchisee in the retail world that is very particular about what they want to spend on technology, what their footprint is going to look like, and what they are willing to do, then you have a new set of capabilities you want to introduce. But it’s across a very heterogeneous environment that’s going to be challenging. How do you bring an appropriate degree of standardization to that? Are there any solutions that you’ve seen that start to solve that problem? Maybe by bringing a good balance between giving autonomy decision-making to a franchisee or a location owner while still bringing enough standardization to actually build some sort of platform or something that’s useful that doesn’t have that “wild west” feel, that doesn’t have all the concerns and challenges of almost everything being different in every single footprint. What are you guys seeing that is doing that in places where they’ve been successful applying that sort of pattern?
Simon Gamble: I’m gonna come from a pretty one-sided perspective, and that’s obviously the Mako one, because we do deploy technology like that night I think I think the key is segmentation. You know another issue that we that we see at these big distributed retail enterprises that I didn’t already mention is that often these franchisees own different brands within the same industry as well. If you’re in oil and gas, these big franchise groups will own a different brands of gas stations. If you’re in quickserve restaurants they’ll own different QSI brands across their own estate and we believe the way to do it is through segmentation. So probably segmenting the areas of the network that are important to the franchisee from the areas that are important to the brand. That allows the brand to get their standard deployed properly in the parts of the network that made it to them, and it leaves the rest of the network that matters to the franchisee to be under their control and if I would overly simplify things I would say typically it’s point-of-sale in front of house networks that the brand wants to keep under control and it’s typically back office networks but the franchise groups want to keep under control.
Stephen Foskett: It’s interesting that you mention to Simon because it really is almost a matrix here where you have multiple owners, multiple brands, multiple franchisees, and of course, multiple applications because part of your segmentation discussion as well was the fact that you know your payment processing system really needs to be segmented from your inventory control system or your HR system for legal technical reasons. But also, as you mentioned, I think that this is really key for organizational reasons because it’s very possible that the franchisee might have their own payment processor that they need to use across their various locations whether it’s restaurants or retail. I think that this is something that is very different about edge, and those of us who are listening to those who come from more of an enterprise background may be kind of looking at this in terms of “oh, well so there’s different application on or is there different groups within the companies different stakeholders within the companies.” So we can kind of see where this is coming from. But to those people I’m just gonna say hang on because you cannot imagine, I think, how different it is when, you know, at the end of the day all those people have the same boss, and at the end of the day all these people don’t have the same boss and I think that’s a very very big change.
Simon Gamble: Most of those people often have multiple bosses, I mean, if you’re if you’re a franchisee. In some ways, the brand is your boss, but also your own organization is your boss. So you are reporting to multiple people and you’re trying to achieve multiple things. As you said Stephen, enterprise is typically segmenting networks out, primarily based around security. I wanna keep my credit card data off my void network for example, but in a distributed retail enterprise, You’re also segmenting because of different organizations, different applications, so the decision and the choice of how to how to segment is much more complicated and you need to leave yourself the flexibility to make modifications over time as new technologies come along, as you’re wanting to do different things in the retail environment, if you wanted to connect the store to different applications, either in the cloud or in physical data centers, you can’t paint yourself into a corner on day one because technology delivers amazing things to retail and retail customers, and brands and franchisees need to leave themselves open to be able to adapt to new technologies and bring them into their environment when they want to. Also to keep them out. In some markets, there’s a lot of third-party app providers that will try and go around the brand and put their own technology into a retail location to cater to things like home delivery maybe advertising menus online and the brand might not want that. So giving the brand the power to stop that stuff from happening to stop them from being circumvented by a third-party is also important. Point being, distributed retail is really really tough because there’s so many different people involved.
Stephen Foskett: It occurs to me that some of that stuff they may actually want, and again, this is one of those unusual aspects of retail stores, especially a lot of them have sort of sub tenants where they’ll have you know maybe it’s a different company that owns the pharmacy or maybe it’s a different company that owns, well certainly owns the Starbucks, but even things like display and I’ve even heard of situations where it’s sort of like a different company supplies the dairy and frozen foods than supplies or boxed foods or things like that mean so you as a consumer think ‘oh I’m going to such and such store,’ but instead you’re walking in and encountering almost a mall of different of different things with different agreements like you said, certainly of the brand may not want that, but they also might want that, and they might want to enable that. So that’s kind of flipside of this. Do you see that as well where brands are coming to you and saying that we have to allow this completely different company access to the same connectivity, the same physical locations. We have to be able to make that happen but we don’t want it to touch us.
Simon Gamble: I mean the example you gave a pharmacy is a great one, so yeah. Most major pharmacy chains, the pharmacy is different from the rest of the store and the pharmacy might only be a really small part of the store, so there’s two separate businesses under one roof. I see that all the time. Gas stations are the same. They might have a quickserve restaurant or sandwich shop or a coffee shop in there, and that’s a completely separate business run by different organizations with different IT needs. Sometimes with that like with the pharmacies that that separation is decided ahead of time and it was part of the business. Other times it’s it’s just something that a franchisee puts in there and the main branch has to accommodate it, so yeah, it’s even more complicated. But yeah, if you choose technologies that that can accommodate that type of scenario then, I don’t want to repeat myself, but you’re not painting yourself into a corner so for me and most of these brands know this and they’re so used to it. I’ve been working in it all day every day. It’s just hard to find technologies that can accommodate them because technologies that are typically aimed at an enterprise often try and squeeze themselves into a distributed enterprise, and in particular, a distributed retail enterprise and they don’t fit the different products for different needs.
Brian Chambers: I think we’re starting to touch on something I think is interesting about the edge that we maybe haven’t gotten into as much which is the different degrees to which the edge requires a multi tenancy right? And obviously we talk about segmentation a bit that’s when it becomes valuable, but multi tendency is very interesting because in kind of one extreme, say the far left, you’ve got cases are there actually similar to ours were we kind of owner environment holistically in a restaurant we can do kind of whatever we want with it and we have multiple tenants, but they’re all internal teams that are building applications specifically for our business in four hour restaurants as we have a lot of control over that and we have a certain things that we don’t have to be quite as concerned about from the security perspective because we own that environment out right, its in our space, etc. Now if we flip over to kind of far right at the spectrum you could have these environments where you have two, three, four plus different businesses, all sharing some degree of infrastructure, possibly for a compute for these workloads, an now this is a whole new set of challenges, which is really interesting. In addition to segmentation, what are you guys seeing in that space and what are you finding interesting in terms of challenges and solutions that exist?
Simon Gamble: In segmentation, it is key. Sending data to different locations for processing as well as doing it at the edge, we see as something that’s growing. Data processing for both information purposes and monetization are things we are seeing in different industries. The point-of-sale data that traditionally is really just being used for accounting purposes in big retail is now being used for monetization. If you’re thinking, I think of things grocery or convenient stores, for example, there’s limited shelf space and they are supplied by these big consumer package goods companies that want sales data because it allows them to do things like modify packaging and see what sort of impact it has at locations, getting locations to put product on different shelf areas and seeing what impact that has but can’t get real time information without paying the retailer for that data out of a POS systems, and from a retailer’s perspective, they can get quite good revenue from selling that data. There’s obviously concerns and cautions that need to be taken around sending too much information and sharing too much information but that’s an example I see that I think it’s quite interesting for a lot of retailers and for sending that away from the edge, and it needs to be done in a secure manner. And there are other really smart things that I see being done at the edge by analyzing video so most retail locations have lots of video cameras around and that information is really heavy. It’s a big data cost of sending that into the cloud for analysis. So you can do the compute at the edge and you can then send just the data streams into the cloud for combining information across multiple stores and businesses can get a lot of really useful information about that. If you think of the gas station how many people never come in to the store, so how many people are in the full court, how many people come in the store and all those people that come in where do they walk in the store and what do they buy? Retailers can use information from the video processing to see if I change my shelf layout what happens? If I move things around what happens? They can get that information for themselves as well. So it’s two different types of data being processed in different places to the retailer and the brands benefit. So I see those things really starting to accelerate and a lot of those because the price of computers come down so retailers can afford to do that sort of stuff now where is five years ago it sounded like a great idea but when priced up it was, it was just too expensive. And I think as the price of compute continues to lower, businesses will be able to do a lot of really interesting things with that or that they own that’s theirs, both monetization wise and in terms of getting information that helps the business.
Brian Chambers: I think you had a really great point, Simon, about the big data payloads and the potential to do things that are just not possible to do otherwise by shifting that computer the edge. I think you had camera vision which is a huge use case, but it just seems like we’re gonna live in a world where more and more of the things that are involved in every single business, be it retail, manufacturing, oil fields, whatever else the case may be, it’s gonna be a more and more rich data environment and we’re probably gonna see more and more different types of data sets that are very large, very big in your video image and intensive things and I just think that is gonna be a critical thing that we’re able to leverage that edge capability to do a lot of processing. We share the things that we need back to the cloud and all is well there, but we can also use those things operationally, which I think is a critical differentiation. We use them in our business even if we’re off-line when we have that computer at the edge so I think it’s really big and important there that I’m seeing of emerging a lot in the edge as well.
Simon Gamble: We see a lot of that video compute not only being used for analyzing customer flow in things but also reducing in-house shrinkage. So you can overlay pause data on top of the video stream you can see if someone at a bar is pocketing money, you can see if someone’s handing out too much change on a regular basis. There’s a lot of stuff that you can do with it that makes it affordable to keep an eye on things and automate at scale. These businesses tend to be pretty large and a lot of the industries we work in, the mom and dad stores are really going away. For a lot of these big, distributed retail enterprises they used to be made up of a lot of little mom and dad companies. Now we see in all types of verticals that, um, they’re changing and a lot of his mom and dad businesses are either acquiring the other mom and dad businesses around them or be being acquired themselves. And so big data pieces are getting bigger and I have to say like ten-fifteen years ago when they turn big data started coming around, it kind of scared me a bit. I thought how do people deal with this and WellStar just said hey these guys are starting to combine and these companies are getting bigger. The price of compute coming down means that stuff is also open to a smaller business. It’s small companies that are starting out but want to one day get big, that’s no longer off-limits to them. You don’t have to be a colossus organization to get benefit from the stuff. But it makes it easy. The price of compute makes it easy for smaller businesses to get the benefits of this analysis and it’s also making it more efficient for large organizations to make sense of it, because the problem the bigger organizations head is so much doubt it was going into places that you couldn’t analyze it, that you couldn’t look at it, couldn’t understand it, couldn’t make sense of it, too much is falling out of the bucket and the big organizations, they’re now getting really clear pictures about what’s going on. There’s third-party companies starting to appear that will do it for you, so you don’t have to do it all yourself. If your business is saving high quality fried chicken meals to customers that they really love, you don’t necessarily have to be a data analysis organization as well. In fact, you probably shouldn’t. It’s probably more efficient for you to do some of our house, but you know, outsourcing stuff is really important too. Organizations and analyzing data at scale can often do it cheaper, faster, better, than you can do things in the house. Sitting in my opinion.
Stephen Foskett: I want to bring in another application that we see a lot at retail and that is location-based services and basically to the collision of Wi-Fi for customer use, apps, and memberships, and hyper location services. And we have seen this at our Mobility Field Day events quite a lot where companies are preparing solutions, for retail especially, that are designed to basically, as you were just saying Simon with this whole big data and customer customization and custom experiences, where businesses are rolling out these solutions that would basically track people throughout the store. I guess you could say that it’s to improve customer experience but we know that it’s actually to improve retail results, Let’s say that. A lot of this stuff is either handled, as you said, by a third-party provider or is handled through basically “free” Wi-Fi. Is that something that you think is sharing the same infrastructure, and Brian, this is this is a legit question for you too, tell me do you think that stuff would be sharing the same infrastructure as point-of-sale, credit card processing, etc. or do you think that this will be entirely separate infrastructure?
Brian Chambers: I tend to think that segmentation is key again. I think it probably depends. In some cases, customer Wi-Fi might be sharing certain elements of network architecture, but then have some really clear separation from things are operational and in business critical. In other cases, it could be a completely separate solution. I know in the United States we have a ton of companies that they have services that are just offering customer Wi-Fi so you can partner with them and they’ll take care of it. It might be a separate circuit with all that traffic is going over it. I’m curious to see what Simon is seeing and thinking about, but I think no matter what the solution, whether it’s separate or they’re sharing will come back to our phrase of the show so far which is segmentation is key. What do you think Simon?
Simon Gamble: I think at scale if you want to hear as much as possible, it drives cost down significantly, but you need to segment as appropriate. So I guess Wi-Fi shouldn’t be anywhere near point-of-sale for example. But if there’s no reason I can’t share the same equipment, as long as it’s secure and you know that the bigger you get the the more saving a dollar a month at each store across 10,000 locations makes a difference you know? So we see a lot of shade infrastructure and we deploy a lot of shared infrastructure. I also want to point out you’re getting that data I see a little demos from from new companies that are saying, “hey we can help you analyze customer flow through a store and we might do that via heading three Wi-Fi access points in triangulating way people are moving,” etc. etc. and a lot of the time was showing and what people bite on is watching their data live in in real time because it’s kind of intriguing, but in reality no one’s gonna be sitting there watching that stuff in real time so you need to have an executive summary that comes out once a day, week, month for it to actually be valuable to an organization. Just like people need to be careful about looking at the very big difference between looking at a single site in real time for 10 minutes for while someone’s demoing technology to you versus looking at your entire estate in a way that you can actually make sense of.
Brian Chambers: What sells and what works operationally are sometimes two very different things for today.
Simon Gamble: Exactly. But you know if you can, it’s an old story, but if you can get your budget out a marketing to pay for an IT spend that can be good too. That’s why a lot of these things exist because if IT is struggling for money and marketing has a pile sitting there then yeah you can get some money out of them. A lot of the marketing bells and whistles can’t be utilized properly and so people need to think about it.
Stephen Foskett: It is pretty scary to think about basically crossing the streams and having guest Wi-Fi able to access, you know, important behind the scenes retail applications, but that being said, I think that it’s also inevitable that that’s going to be happening, that they will share infrastructure. So I think that it requires extreme levels of security. There was a highly publicized situation of sort of a sitting in the parking lot hacking over the Wi-Fi at a retail company. And that is not at all uncommon. Anyone who’s been to DEFCON knows about some of that stuff. But yeah, I think that it is pretty scary from a technical perspective to think about crossing those streams. It’s also scary I think from a consumer perspective to think about all this data that’s being collected, but I don’t think that we’re going able to restuff that genie in the bottle because of what you just said Simon, which is it kind of comes down to budgets and the way that these things are sold. I think a lot of companies look at, you know, localization and location-based services and customized ads and things like that as a way to basically make a lot of this edge IT pay for itself or even more, become a profit center. In many cases that’s working. I think that if you look at some of these large retailers, many of them are making money, like real money, real profit, off of the IT infrastructure they have in their stores nowadays, whether it’s through partner-based advertising, multiple tenants, as we said the loyalty program/customer, location, services, etc., but it’s not all bad. I think that also some companies are also using these to improve customer service and to make customers happy. I know that increasingly were seeing distributed purchasing where you can basically buy something from anywhere in the store and restaurants as well. We’re definitely seeing a quick serve restaurants of proliferation of self-service ordering, roving ordering, as I know Brian that you guys have have implemented as well, where people are ordering from places that you wouldn’t think, you know, from the driveway, from outside, and that actually improves customer service. I think that customers really appreciate those applications. So maybe maybe we can shift gears a little bit. Yes, some of the stuff is scary, but some of the stuff is helpful too, right?
Simon Gamble: I think it ultimately it all has to be helpful. When everybody in business, no matter what you’re in business to make money. That’s what business is about. But you’ll most important asset are people, whether it’s internal people or customers to come back to you, so if you if you scare of your employees all your customers, you’re not gonna be making money in business for very long. So these things have a way of self-repair. There are some dodgy people around that that come up with technologies that may be a more of a scam than now, but they don’t last very long. And anyone that’s been in business for a decent period of time knows you have to look after your customers and your employees. So yeah, everybody needs to be aware of it, but as a consumer, speaking as a consumer, I’m not particularly worried about a lot of the stuff. I have some cautious skepticism about a lot of technologies but I think by-and-large, these things are for the best.
Brian Chambers: I don’t have the answer to this one but I think we’re in an interesting intersection where everybody wants the best possible customer experience, which look across all businesses and there’s almost always a digital, you know, mobile application or person using technology. There’s some technology in the mix of that like that value chain, that customer experience, whatever it looks like. So at the intersection of that, people actually demanding that, it’s okay. What people seem to want is convenience and you know, fast service and all those kinds of things. Then we got on the other side privacy and being able to respect the wishes of people from a privacy perspective there’s a ton of laws about that obviously both you know in Europe, in United States and some other places, so we got these two sort of like different ends of the spectrum, and then ultimately have to figure out how do we match those together on people’s desires on both sides, and still do what you’re saying Simon, which is still run a business and hopefully still have a degree to where people can connect with other people in the service aspect. So it’s a very interesting intersection. And those two spectrums I think we’re gonna keep kind of butting heads with each other and running into each other a little bit for a while. But I do share your optimism that I think these are things that can be worked out every time and I think there’s ways to find solutions to respect the privacy and meet that customer service and customer experience expectation.
Simon Gamble: I agree, Brian. And it really sums up our conversation today, um, complication for these big distributed retail enterprises, the networks are complicated, the solutions that are being delivered are complicated, the considerations that everyone has to take about how this impacts internal and external people are coming into an out of the business. And I think if I had to summarize our conversation today and I’ve said this before on a couple of other conversations like this, it’s think. Think about what you want to do, think about of the inputs and outputs that you’re wanting to do and you’ll find a solution. Work with partners that work with organizations like yours and have some experience. Talk to your own people that have done this before and think about what you want to achieve and you’ll have a great technology in your business. You’ll have great edge, you’ll have happy customers, and hopefully you make more money.
Stephen Foskett: I think that’s a great way to to wrap it up there, Simon, and thank you so much for this conversation and frankly you’ve given us a lot to think about and this to me again was my take away from the Mako presentation from Edge Field Day. Yeah, it was great to learn about what the product does, but it was also, and I think may be more valuable, to think about the unique aspects of these environments and how that demands a different solution. And it kind of comes down to it, you know, I guess if I had sort of a thing that I’m always looking for with products, it’s not about how cool the technology is it’s about how well the technology solves the the specific challenges of a specific environment. And so that’s why I think we need special purpose products, special purpose solutions for edge and we can’t just throw in whatever worked in the data center, whatever worked in the campus, whatever worked in the cloud, whatever works previously. We have to try to find a solution that meets the specific needs whether it’s multiple tenants under one roof, whether it’s the conflicting demands of the brand versus a franchisee, whether it’s the kind of data collection happening with consumers, whether it’s offering new solutions, new at levels of customer service to those people. All of these things require specific new solutions and so it’s it’s really been interesting conversation. Before we go, Simon, where can we connect with you and continue this conversation on edge?
Simon Gamble: You can find and connect with me and Mako at LinkedIn, also makonetworks.com and there’s another discussions and things that you can find in the Gestalt IT Tech Notes series as well.
Stephen Foskett: Thanks and will link to those in the show notes too. Brian, is there anything new going on with you?
Brian Chambers: I continue to do my normal weekly newsletter at [email protected] and it’s called the Chamber of Tech Secrets, I am having a lot of fun with that and getting some good engagement from the community so would love to have you follow me there. I’m also around on Twitter, @BriChamb, and then on LinkedIn as well Brian Chambers. I got that name at least. So I’d love to connect with folks and talk more about this kind of stuff. I love all the things that are going on in the edge environment and all the unique problems that people are working on solving there.
Stephen Foskett: Thanks Brian and Simon and thank you for listening to Utilizing Edge. This is part of the Utilizing Tech podcast series. If you enjoyed this discussion, please do subscribe you’ll find us in all of your favorite podcast locations. You’ll also find video of these on YouTube just got a YouTube.com/GestaltITvideo and you can see those. Also, you might want to check out Tech Field Day and specifically Edge Field Day where we heard from Mako and we learned a little bit more about their specifics along with a bunch of other companies, and will be returning in July with another Edge Field Day event. This episode is brought to you by GestaltIT.com, your home for IT coverage from across the enterprise. For show notes and more episodes go to utilizingtech.com or find us on Twitter or Mastodon at UtilizingTech. Thanks for listening and will see you next week.