Tristan Rogers - Product Led CEO & Advisor

Welcome to another CEO Wisdom Pod. We have Tristan Rogers for round 2 or 3, I can't remember. He's a product-led CEO and advisor in London. His company is called Huck Industries. When we talked, Tristan, I think you had just started Huck. Now you're almost 2 years in, so I want to catch some news here and know what's up. So welcome for round 2. Tell us a bit more about what's been happening in your life and in Huck.

Yeah, um, time flies when you're having fun. So, uh, we did a fundraise last year, June last year, that went well, um, gave us the, uh, ammunition, I suppose, to kind of complete a bunch of changes I wanted to make to the business. So we ended 2025 on target. 2026 was more ambitious, uh, we're not there yet, we've got a quarter to go. Q4, but looking good. Launched our new software platform, which is a data-as-a-service platform for location intelligence. Picked up some big fat global customers and very much being product-led, so listening to where the customer wants us to go next. So yeah, I think we're on track for year 2 of our plan. And then next year we launch in the US, which is a big step for us.

Interesting. And what does— how did product-led change in the age of agentic AI?

Going faster. Yeah, I think it's not just going faster. I think we are definitely going faster as a result of different teams using tooling to do more with fewer people, which is, I think, is a common theme. Where it becomes beyond just going faster and perhaps more interesting is where we— so just a bit of background. So we are a location intelligence platform. So the simple premise is that if you're a business that heavily relies on the performance of a physical location, so think retailer, So you're Nike or you're, you know, you're a big brand that has stores around the world. Where should you put your store? Why should you put it there? What are your competition doing? How is your existing portfolio performing? So we are gathering different sets of data that give the customer an insight into that location performance. Now this could be number of people coming, the spend that they bring, the demographics of those people. It could be crime rates. It could be the number of Ubers in the area. It could be the price of parking. The list goes on. Where an LLM could potentially bring something entirely new to Huck is when, if we take all of those different and fairly disparate signals, could there be patterns in that that could help us do predictive modeling? So could it actually start to say these things look like this, therefore in 6, 12 months' time, this could be a location that today would not be on your radar? That's what we're starting to think about. We're just starting to do some tests on that. We're not quite there yet, but I think that's potentially a whole different area of value.

You said you landed these big fat client accounts. How did that happen? Was that you? Was that Some of your sales team, how did you get in touch with them? Tell us a bit more.

So back to being product-led, so it always comes down to being visible. So we did a lot of work on, again, that's the other thing with LLMs, it's to some degree destroyed the outbound market in sales because it's a bit like Terminator, it does not stop until you are dead. And so the amount of content going outwards towards customers, prospects, is unrelenting. And if you or I are indicative of being a consumer, if somebody phones you on your cell phone and you don't recognize the number, you don't pick up. And now Apple has a concierge service that filters all of that anyway. LinkedIn, pretty dead. Email's definitely dead. It's just really tricky to do outbound because people are being bombarded with outbound generated by AI. So we had to turn it on its head. We had to rely on the fact that what we're doing is real, what we're doing is needed, and that those that were responsible for location performance were going to have to go and find something that could help them. So as long as our marketing, our was genuinely good content and it was available for people to easily find, whether it be via an LLM or Google search or if they're browsing on LinkedIn. And in time we start putting more content out on popular publishing sites, customers will find us. Then it was about making sure that as a data company, the way I look at this is that as a data company, We have the answer to a brand's questions around location performance, whether they buy it or not. So our investment is no longer in traditional ad spend or even physical events necessarily. It's about proving to the customer without any investment or risk on their side that our answers are useful to them, our data is useful to them. So We're offering customers free trials. We genuinely don't bother them. And that in itself is what has led us to those big clients have been able to test us, test our data, get comfortable, and then make that commitment.

What's different with your platform? Like, let's say I'm a retail shop, I have a franchise or whatever. I take pictures with my iPhone and send them the plan of the of the floors in the commercial center, and I send them the Google Maps. Um, I mean, I send them— I send Claude, right? Like, them. It's interesting I call it them. Um, and I send it all this data. What's the difference in between that and using HUC?

So there's 2 things. One is a lot of the data which is more to do with— so a part of our data panel is GPS data. So consented mobile apps that require someone to have location services enabled. So if you had a parking app, for example, you need location services enabled to perhaps get the parking code for the area you're trying to park in. If it's a fitness app and you want to map your run, you need location services on. So it's about selecting partners where there's a legitimate reason for location services to be on. We are then able to take anonymized GPS readings from that app in huge quantities across a huge number of apps globally and build a panel of anonymous pings, GPS pings. We're at the opposite end of the telescope to things like targeted advertising. We don't care about the individual. It's the opposite. We care about how many people are in this area. Now, we're not talking about significant percentages of the populace, and that's where some of our modelling is involved in modelling it towards an estimate which is accurate in terms of the number of people. But that data is proprietary. It's not something you're going to find on the internet. Then we have to test that data against things which are available. Some of it behind a paywall, some of it not. So in terms of coming up with the raw number, how many people are here, how long were they here for, where did they come from, where did they go, that's pretty proprietary stuff. But the second area of value is that, as I mentioned earlier, with lots of different sorts of data, parking data, crime data, weather data, Businesses that are serious about location performance do spend a lot of time and money just trying to find all of the different bits of data that might build the fuller picture of the location. We're doing it for them. So an LLM is great at pointing at stuff and getting intelligible answers, but as we know, it will always attempt to build an answer even if the dataset is incomplete. So Our idea is even before we get onto LLM constructs, we're building the most complete location dataset available to the customer so they just don't have to bother going and finding it and then working out how to build a rounded answer from all of these different datasets.

Right. And if you have multiple locations, if you're a giant, then I guess that it starts to be useful, right? Because you want data to lots of potential locations to your new store, right? And you may have one launch per week, right? You do 50 stores a year or something, so then it makes sense to not have to do the work again and again internally. Plus, yeah, the cost of doing that work doesn't make sense. Plus, if you mess up one location, then it might mean losing millions, right? Because I guess in retail, still location, location, location, and the cost of opening a location is, uh, is quite hefty, right? So that's the ROI method using HAC.

Certainly in terms of expansion, but we also have to think about the businesses that are static, uh, and, and possibly the businesses that are consolidating. So another use case would be, you know, we've got 3 locations, they're reasonably close to each other. Do we need those 3 locations? Could we actually get away with 2 locations because the audience is shared? So it's what we call cross-visitation. So we need to be able to analyze the impact of removing one of the locations and measuring whether you keep the footfall And the associated revenue, even though you've now cut your costs by maybe a third.

What's the footfall, Tristan? What's footfall?

Footfall as in, sorry, the number of people coming into that area, into that store, into that shopping mall.

Okay, got it. Is this problem like the traveling salesman problem, like infinitely complex to crack, like to know who's going to step in your store at what time? Or is it predictable at this point with data and LLMs?

It's fairly— well, yeah, it's a good question. If it's an established location and all things being equal, you would expect to see fairly similar seasonal trends and weekly trends. daily trends. So, um, a Saturday morning is typically going to be busier than a Tuesday morning. The weekend is going to be busier than the week. Uh, depending on who you are, uh, Christmas trading might be busier than February. So you've got these broader trends, but of course this is assuming all other things are equal. So if a a new shopping mall had been built down the road from an existing shopping mall, are you now going to see a decline in people coming to the old mall? That's very much a reality. Equally, if a new housing development was built near to an existing shopping mall, could that potentially be bringing new people to that shopping mall? Or a road closure or a new road, you know, so that— This is why there are lots of different variables that need to be considered all of the time to understand cause and effect. And

you've expanded in the US, right?

We are about to.

Okay. Hence the footfall. I think the US is like foot traffic, right? Like foot traffic.

Okay, I'll adjust my language.

Still, still, I'm intrigued, right? Like you said, no email, no outreach. Sure, the product does the work, but like, how do you get the customers? Like, how do your clients hear from you? You're obsessed with like data of how people find them on foot, but are you obsessed also with the data of how they find you? Like, is there some AEO, for example, nowadays? I see the website looks sharp, right? Like, do you know where they enter your funnel? I'm guessing it's not just word of mouth.

Yeah, increasingly. You know, it's a fairly new endeavor. for us, so we're learning on the job. In fact, we've even built what we call our pipeline engine. So I almost regard CRM as being dead. I don't really care to pay for tooling, I just care to pay for pipeline. So we want to understand where the universe of customers are, and we do care about how they find us, so traceability of the customer. So we can talk about our total addressable market and our serviceable addressable market in Europe and in North America and in Asia, APAC. You can buy lists, and then it's really about making sure that your content is very available and findable, simply on the basis that we know the problem exists, we know people are looking for a solution. So as long as we put our way in that search of that search, they'll find us. But the further out into the marketing universe you go, the less visibility you have of where people ultimately started their journey.

And your ticket size, is it about 30 grand a year? Like what should people pay for to use your software? Is it seat-based also? Like what's your—

It's very, very scalable. This again is new for me where we tended to have a sweet spot with my previous business where the average contract value was sort of around $120,000, $150,000 per annum. With Huck, we can service customers from like $5,000 or $6,000 a year up to $500,000 a year. So it's hugely scalable and what makes it scalable is how many different geographies they need access to, and there is a proportional cost to us to bring more data into the platform. And then there's features. So what do you want to search in that geography? Because as I said, different datasets sometimes are paid for behind a paywall, sometimes they're not. But even if they're not, our processing overhead is quite significant as we turn what could be just, to the untrained eye, a bunch of fairly useless data into something that helps a commercially minded person get to the point where they feel comfortable saying, I think we should take a lease, a 20-year lease on this prime real estate, do a full fit-out, stock it, and staff it. Yeah, that's a significant investment and a significant risk. We've got to de-risk those customers' choices as much as possible. So regardless of where they start, the value prop is the same, but it is entirely down to scale. So if you're looking at one location, that's okay, that can be a small ticket. If you want to look at hundreds of locations, thousands of locations, that's okay too, it's just going to be a bigger ticket.

Interesting. And how's that company different than your last wins? For example, Concrete. Yeah, how's it different?

So Concrete was a very traditional software as a service vendor, so still very much product-led. So we identified a market problem. It was in retail as well. So this was how do retailers ensure the highest quality of operational conduct in a store. So a well-run store generates good revenue. And so what we built was a task management and communication tool that helped retailers deliver timely instruction to stores, oversee the completion of those tasks with the sign-off, and do that every day. And that drove compliance, which drove revenue. And the billing model for that was seats or numbers of stores and, and the features within the platform. With Huck, it's a data platform. So as I said earlier, we have the answer to a customer's needs whether they buy it from us or not. We can deliver value to a customer without them paying for it if we choose to, whereas with Concrete, the only way for value to be proven ultimately was for the customer to use the software day in, day out and see the efficiency gains from using the software versus the manual process. So there was a degree of, although we really tried hard to what I used to call the thin end of the wedge, de-risking the choice to give us a chance. So we were able to introduce things like proof of concept. So a fixed price for a fixed period of time, and at the end of that period, if they, for whatever reason, didn't like it, they could bail from the contract. But there was still that time and the people involved and shifting processes from how they used to do it to using concrete. So even though we did everything we could, there was still that jump-off point that potentially delayed people making choices or potentially didn't make a choice at all, just stayed with what they were doing. With Huck as a data company, We can just deliver them the answers and say, there you go, it's free. We've looked at this location, we think this would be a great location. Or did you know that this location you've already got is seeing declining traffic, declining other datasets that could be something that you should be wary of? You might want to be thinking about moving to a different location. So it's quite empowering to have that ability to offer value without the customer committing to anything.

Right. And is building Huck— sort of my question was like, is it way easier than building Concrete? Took you 18 years to, I guess, exit Concrete. Is everything easier now with Huck? Are there new challenges arising, or are you just having fun nowadays?

I think that, yeah, the, the time at Concrete was definitely a journey of, uh, Plenty of table tennis, Gran Turismo, and cappuccinos, along with learning about how to run a company and learning how to be product-led, listening to the customer, and then looking at the opportunity size, and then productionizing that into a platform. So we were building a product. I think having come out of concrete and taken some time off, I thought back on my learnings and my view was there are lots of good ideas in the world. What matters is how well you execute on those ideas. So not being the founder of Huck, I was able to look at it dispassionately, objectively. I thought what it could offer retail was significant and real estate was significant. So it was about bringing a really solid playbook of how to operate a business like this into Huck and driving it forwards. So that's enjoyable to prove whether that's true or not. So far, so good. Ultimately, that's down to my shareholders to make that decision on whether I'm doing a good job or not. But so far, the numbers tell us that this is working. The fun bits is different. I think fun for me is when you know that your company is almost like a family, and that's culture. So you can still, you can have a high-performing company with people that just do their job, but I don't think they'll be as performant as a company where people do their jobs, but they do their jobs well because they enjoy their jobs, and everyone gets on because they understand why they're there. They understand that everyone's doing their best, and when— and they also know that when they come together and work together, they, they are greater than the sum of their parts. And that is something that I've worked really hard on at Huck since I've been there, to try and create a really strong culture. And I think we're there, and certainly speaking speaking to the team, I might ask somebody, how's it going? Or how do you feel? And their tendency will be to talk about the piece of engineering they're working on, or the latest product idea, or a deal they're working on. And I'll say, no, no, I didn't ask that. I just want to know, how are you? How do you feel? And then what they tell me is how they feel about the job they've got, and coming to work. And if I hear that they're really enjoying their job, generally speaking, I'd call that a leading indicator. That means that they enjoy doing their job well. They're performing at a higher standard than just doing their job without feeling good about it.

And what's on your goals list as a CEO this year? I mean, you've already scaled to great extents. You're constantly obsessing on product. Locking in new clients, right? Like, uh, what's, uh, what's the next mountain to climb?

So, you know, geographically, North America represents our largest opportunity, largest market. So, um, you know, that's going to be a significant investment for us next year, um, even though we'll be doing the same thing in a different market. It's a different market, so I'm excited by that. Certainly, you know, with Concrete, we got to 75% North American revenue. So we were a British company that found great success in North America, and I would like to replicate that. That's one of my things on my to-do list is to do it again. But certainly putting personal desires to one side, the need is very much there. So it's just about making sure that we are seen as a contender and maybe even a leader in that market in 2027. Beyond that, it's back to what I mentioned earlier. I think with an accelerated product roadmap, I'm excited about starting to deliver modeled outcomes. So rather than just reporting on actuals and trailing indicators of what's happened so far, get to the point where we've got enough data and enough pattern recognition to start building some modeling on what's going to happen next. That's fascinating. Now, we're going to have to approach that very carefully, and we have to work very closely with customers, but in theory, it's a plausible endeavor.

Where can people find out more about you, Tristan?

Um, yeah, I'm still on LinkedIn. It's still, uh, it's still a, a shop window, I guess, for me and Huck. Come visit the website. Um, the usual, I guess. Yeah.

Huck.io. So that's Tristan's website, and, uh, hopefully Tristan will be on the