I'm part of the Business and Information Services team here at BofA. This session will be on NIQ Global Intelligence, and I'm pleased to have the Chief Financial Officer, Mike Burwell, with us. We're going to structure this session as a fireside conversation, and we'll open the floor for questions if time permits. Thanks, Mike, for joining us.
Glad to be here.
It's been almost eight months since NIQ has gone public. It seems like the first year of a company going public is really the time that it sets the building block for the years to come. Now, with several quarters of public performance, what has been the biggest misconception to how the investment community perceives NIQ Full View compared to your internal reality?
Yeah. I think the biggest misconception is that NIQ is a static data provider. Internally, we see the business as becoming more embedded, more differentiated, and frankly more valuable as our clients continue to operationalize AI. In fact, our client behavior tells really, to me, a real compelling story, and that is that 105% net dollar retention. We have 105% net dollar retention. We have 98% gross dollar retention. Over 30% increase year-over-year in terms of our data being used by our clients. When I think about that's not about commoditization. Really, frankly, that's deeper embedment that we're seeing actually in our clients. We're also seeing roughly two-thirds of our top clients or our top 50 clients using one of our AI-native products.
We've launched three in the marketplace to date, and that's not AI-enabled. This is just in terms of AI being used as standalone products. I guess it just, you know, when I think about our AI BASES Screener, our AI Product Developer, and now AI Analyst, which we launched last weekend, these are AI-native products that we are seeing adopted already in the marketplace. When I think about our Full View, it's more than just data in that it's, you know, it's governed, it's permissioned, it's harmonized system of record across 90 markets. As we see AI move, you know, in more and more into high-stakes decisions, we cannot have hallucinations.
Sure.
It's important for us to be in that role to be able to help our clients be successful. You know, when I think about it, what AI is doing for us is continue to drive durable revenue growth and ultimate value for all our stakeholders, including our investors.
Perfect. The IPO and the subsequent debt refinancing has fundamentally transformed the capital structure. From a CFO's perspective, how do you balance the need for short-term public market consistency with the long-term technology investments required to maintain your leading market share?
Yeah, I think, you know, it's not an either/or question that you're putting forth, which is, you know, how do you make the short-term, you know, consistency commitments that we've made to our investors, but equally make sure we're investing for the long term? You know, the good news is we spent well over $400 million investing in our platform, and we did that before we went public, in terms of really bringing that platform to life. But equally, when I look at the fourth quarter of last year, you know, we did 410 basis points improvement in our margins. At the same time, we continue to invest, and we continue to invest in AI. We continue to expand our panels. We continue to invest in our Full View capabilities.
Look, we set and look to set guidance that we can hit, so that's what we set up in terms of the short-term view. Over the long-term view, making sure we continue to invest in this platform for the future. You know, when you look at just in a CapEx standpoint, we're looking to spend 6.5%-7% consistently going forward, of which 70% of that's growth capital, the other 30% really being maintenance capital in terms of what it is that we're looking to spend on. I think we're balancing both that short-term and long-term, and that's why I use that word and as opposed to or.
Got it. Great, great to clarify. Thank you. Another AI question. A common theme across the intel services space has been AI and the replicability of data. For those of us in the room who are a bit newer to the story, can you walk us through how you ensure NIQ's data remains proprietary? For parts of the data assets that may not be as proprietary, how does the embeddedness of the data make it difficult for customers to switch?
Our advantage is scale, breadth, and granularity of our data engine, and the complexity and governance of unifying thousands of data sources into a permission system clients can rely on daily for making major operational decisions and run their business. Now, said another way, what we're teeing up is we cannot have hallucinations. People are making multimillion-dollar decisions based on our data, and we can't be wrong. You know, it's something we take very seriously. When we think about scale, you know, we're operating in 90 countries in terms of what we're operating in. When I think about breadth, we're bringing in retail data, consumer panel data, e-receipt data, e-commerce data. We're bringing that all together and harmonizing it in a proprietary product that really builds a reference data layer that people are able to access.
That's what creates this, what I call, decision-grade intelligence for our clients. Look, we have strong governance, so it's our data. You just can't use it in any way you want. We've made those commitments to the various data sources that we've entered into to be able to get that information. You know, look, we see our clients have embedded it operationally into their business practices, whether it's pricing, whether it's promotions, whether it's supply chain. You know, all those workflows are absolutely what we're seeing happen, you know, across our client base. Look, we're very well embedded on changing us out or switching us or simply saying you're gonna hook an LLM model to this is not an easy feat and not something that we see happening.
Building on that prior question, in an era where data scraping is common, why is direct from retailer data still the gold standard for enterprise demand?
Yeah. Look, when I think about high price decisions or very important decisions, call them high stakes decisions, if you will, you know, look, it requires trusted permissions and reconciled data and, you know, not just scraping information. Now we do web scraping too, just to be clear, but it's in the context of putting it all together. It's not just taking some fragmented information, and you're gonna summarize that and project it to the future and have some client making decisions on it. We just don't see that to be the case. When you said that gold standard around retail data, you know, I think it's a gold standard around all the data that we bring into the house and how it is that we look at to make sure it's accurate. How can you know, harmonize it?
How can I put it into metadata? It's not just data. It's how do I take that data and turn it into advice and help our clients make better decisions. You know, look, as AI increases, the cost of being wrong continues to rise. You know, look, the value of what I would call enterprise-grade data only goes up, and, you know, it plays directly to our strengths overall.
Got it. Earlier you mentioned that of the top 50 clients that you have, I think two-thirds you said was adopting AI-native products. Does this suggest the moat is widening as your data becomes more deeply embedded in your clients' own proprietary AI models?
Yes. Yes, it does. I mean, AI widens the moat by increasing embedment as we see it in being embedded and uses intensity. You, as you quoted those numbers, as we look at the clients, our top 50 clients are adopting over, you know, 2/3 Of our AI-native products. We also see those clients then are the same clients that are increasing their spend with us as opposed to non-adopters. You know, data consumption's up 30% on a year-over-year basis. Just to put it in context, you know, we process 4 trillion transactions a week in terms of data sources that we bring in into our platform, up from 3.4 trillion 12 months earlier. You know, tremendous amount of data.
The role that we play, you know, is an important role in terms of being able to harmonize that data, bring that data to life. It's only increasing that moat when you think about people using that 30%, and you're seeing that increase happening with the AI-native products.
Great. Just wanna switch over to your medium-term outlook of mid-single digit organic revenue growth. It's built on a science of retention, cross-sell, upsell, renovation, pricing, and penetration to new markets. Can you walk us through what parts you are most excited about in terms of upside opportunity, and where in the algorithm can you see sustained level of outperformance?
Yeah. I'm excited about our business overall, but I'll break it down in a few points. One, when I look at our intelligence business, you know, it's remained durable and, you know, with our pricing that we have in place and our renewals, you know, we've seen that for the last, you know, almost seven quarters at greater than 6% revenue growth. We feel good about what we're seeing, as it relates to our core intelligence business. Now, our activation demand is intact. You know, we've seen certain clients, you know, defer a little bit of their timing of those projects, but we don't see any difference in terms of demand in the marketplace overall.
As I said earlier, as you look at our top 50 clients, you know, two-thirds of them are adopting our AI-native products. You know, we feel good about our activation business and continue to see that grow into the future. What I'm most excited about, and you'd asked that in your question, is our new capabilities. When I think about AI BASES Screener and how I can develop new products and get them to market, you know, one of our clients in the market has touted it that they can develop new products 65% faster by using AI BASES Screener because they're able to look at synthetic consumers. Where before it would take you weeks or months to evaluate that product in the marketplace, now you can do it in a matter of minutes.
When I look at our AI development in terms of being able to develop that new product and bring it to market and managing those supply chains, is something that we're able to do even faster for our clients and helping them. Equally, the new product offering that we have coming out as it relates to AI Analyst is really looking at personas, over 40 personas that are out there. Say you're a brand manager, we're gonna help you ask the right questions to be able to do this in a democratized way, where before you may need to be an NIQ specialist to be able to do it. Now you're able to ask natural language questions and get responses based on your persona.
Now, we haven't launched all those, all 40 personas yet, but we just launched a few in the marketplace, but we're gonna see those overall. Look, I'm not counting on upside in our model.
Mm-hmm.
Just to be clear, this as you think about it, but I think the building blocks are in place. When I look back and say, "Look at our intelligence business, look at our activation business, and look what I really see in terms of our new capabilities overall," and probably last but not least, is think about our new verticals. I feel good about what our growth algorithm looks like going forward.
Awesome. A lot to be excited about. I think while intelligence growth has been robust, activation, like you said, has been fairly flat over the year. You're taking decisive actions to return activation to growth in 2026. I know you mentioned that you're not seeing, you're seeing a deferment, not really a delay in customers there. Can you just walk us through what is needed in terms of client behavior for them to see confidence and for you guys to see confidence in this recovery?
Yeah. I mean, look, activation softness to date has been driven by project timing and uneven client conditions, not demand erosion, or some people have put forth to us that it's AI disintermediation. You know, again, I go back to that 60% of our top 50 clients using an AI-native product today. I mean, the demand is there. It's increasing engagement and repeat usage. To be fair, we have sharpened our focus on a fewer number of offerings we've had. We've had over 64 offerings in this space. We're looking at narrowing down that. We're also improving our discipline in terms of how we convert those leads into revenue and wins, which I think directly, you know, deals with some of the past friction points.
Look, in a nutshell, I look at it, our demand is intact, execution's improving, activation is well positioned to grow in 2026, and that's our expectation.
Great. Let's switch really quickly to margins. It seemed like during the era where NIQ was private, among the transformation steps was a reduction in cash data costs, which is now about 15% of revenue from 21% back in 2021. From a CFO's perspective, is there a theoretical floor for these data costs, or will continued AI scaling drive that percentage even lower?
Yeah. Yeah, as a CFO, I can't say I probably should never say there's a floor, to be fair. I mean, I just go back to the fact that you quoted on, you know, our data costs at, back in 2021 were 21% of revenue. Today, they're, for the last year-end, they were 15%. The biggest impact of that has been the value proposition that we're bringing back to clients. When you think about data, we get data roughly four ways. We buy the data, we rev share the data, we get the data for free, or we trade and barter the data. Many of our retailers want feedback back on their operations, and we enter into those arrangements.
Because of the value proposition that we're giving to them with our platform and their ability to be able to leverage that capability that we have, we're able to negotiate different deals than we've had in the past. To me, it speaks to, you know, that data going down is something that's just enhanced capabilities, the quality of the product, and what our clients are thinking about it from a value standpoint, overall. To me, it continues to support our margin expansion. You know, some people have asked me and maybe just now, "How do you compare yourself to other clients?
What's the right margin for the business overall? You know, we've looked at and said, if you look at many of our peers, if you will, in a broader case, that margin looks more like 40%. We said in the midterm that we would get to the mid-20%s, and we're spending 15% on data costs. That stitches you right in that bucket. We're not willing to stop there. We're looking to continue to drive with our revenue growth algorithm and 80% fixed cost base to continue to drive margin improvement.
In fact, we've put forth a 200 basis point improvement in margins in our guidance for 2026, of which half is gonna come from our AI productivity actions, and the other half is gonna come from just revenue growth, and with that fixed cost base to continue to see that improve.
It just leads into my next question about your 2026 margin.
All right.
I know you talked, you just touched on it briefly, but just as you think about not just 2026 margin, but the longer-term cadence in 2027 and 2028, how should investors think about the cadence there? You know, what are some of the factors that will unlock margin expansion?
We look back to 2025, you know, we improved margins by 320 basis points for 2025. We improved them by 410 basis points in the fourth quarter of 2025, and we put forth a 200 basis point improvement for 2026. Then beyond that, you know, what I think is we ought to generate 50 to 100 basis points improvement with our revenue growth algorithm and our fixed cost base. Right now, that's what we continue to see. I mean, I won't say forever, but nonetheless, in our foreseeable future, in the years that you talked about, we should continue to see that kind of margin improvement.
Got it. Part of the IPO story has been to use the proceeds to reduce your leverage with the target set to the sub 3x by the end of 2026. Once you hit those targets, how do you see the appetite for capital allocation adjust, if any? How would you manage interesting data assets that come up on the M&A block during that debt paydown period?
Yeah. You know, we had set a target to get below 3.5x by the end of 2025. We ended the year at 3.25x. As you said, we've put forth to get below 3 by the end of 2026, and we were free cash flow positive at the end of 2025, then we'll see the further inflection as it relates to 2026 and beyond. We were looking probably 2-4 M&A deals a year, but we don't see them as big deals. We don't see any place that we've got a void in terms of geographic, you know, point we don't have or data that we don't have to be in place. But we are seeing opportunistic deals coming our way. We did two deals last year.
When we look at these deals, that they're gonna be accretive in year one. They're gonna be tuck-ins when you bring them into our distribution channel, that we're gonna be able to really drive them, you know, pretty quickly overall. When we look at that, you know, we'll look to continue to generate that free cash flow, use that cash flow to pay down debt, and continue to drive it below three. We will continue to look at that capital or cash flow to look at these tuck-in M&A deals overall. Look, the goal is flexibility at the end of the day, is to allocate capital where we see the most value and we'll continue to do that going forward. That's kind of our philosophy in thinking about it at this point.
Got it. 2025 was described as a inflection point and like you mentioned, achieved positive free cash flow ahead of schedule. How does that early inflection change your day-to-day strategic flexibility and ability to pursue opportunistic growth in the current environment?
Yeah, I mean, positive cash flow gives me more optionality for sure. So I'm not looking to change our discipline around making sure we're continuing to drive free cash flow into the future. You know, as I looked at the second half of 2025, we'd put forth, you know, what kind of free cash flow we're going to drive, and we were going to be positive over the second half of 2025 when we delivered $315 million of free cash flow over the second half of 2025. As a result of that, you know, what's that position? So it's lower interest expense significantly, both from the de-leveraging of the IPO as well as the re-financings that we had done. Equally, the one-time items that we've had, we're structurally continuing to drive those down.
You know, we know that that's important for our investors, and it's important for us going forward to make sure the earnings line up with cash and there's not a lot of adjustments associated with that. But at the end of the day, you know, this free cash flow gives us the ultimate ability to invest, to de-lever and respond opportunistically to the things that are actually happening in the marketplace. Look, I don't think we want to stretch the balance sheet, but we want to put the balance sheet in the right shape, and I think we're heading in that direction overall.
Great. Before we end our conversation here and open up for questions, I'd like to do a quick word association. This is something we're doing with the presenters this year at our conference. Tell me the first thing that comes to your mind when I say the following. Going public.
A lot of work.
Data costs.
Going down.
AI.
Misconstrued by the market.
Got it. The last one, your CEO, Jim Peck.
Awesome.
Awesome.
Best CEO I've worked with, and I've worked with several.
Perfect. Let's open the floor up for questions, if any.
Since everyone's talking about AI risk. What have you seen anybody even try to use whatever tool? If they were to try to do that, could you charge like some sort of access fee? Because they're not going to just go cold turkey, like completely shut you and then start. You know what I mean? There's gonna be. Because I'm hearing some of the firms, some of the software companies, that's what they're doing. You know, they're like, "Okay, fine. You want to go start using AI tools, but you want to get our data and you're going to have to pay for the data then.
Yeah.
I think.
Yeah. What I see, you know, is kind of a pushback. People view that they can get access to our data. It's our data, it's permissioned only by us, and we spend a lot of time, you know, cultivating those various data sources over time. We're not just going to give them to someone's LLM model and let them go roll with those things. We have looked at segmenting our clients into kind of three buckets. I call it AI builders, which are the more sophisticated people that want to use it. We call it AI buyers, and then AI beginners. You got different flavors of clients and how they're using AI. Ultimately, you know, it's permissioned data.
It's not just about data, it's also about advice that we're bringing into it. What I mean by that is you're looking at what was purchased, but why was it purchased? What else was in the basket? How am I giving you those insights to make those decision points? You know, look, we're continuing to work with our clients and want to be adaptable, but we're not just giving our data away. Frankly, without our permission, you can't use it. You know, we have no one client that's greater than 3% of our revenue. You know, that's the reality of what's happening. How are you going to get the data? You know, you just can't get it without going through us.
The how we get that data is, you know, we've got bodegas where we're going into India and, you know, beginning inventory, ending inventory, give me your purchases, and I back into your sales. We have a whole field force that's going out in terms of doing that stuff. How you compare e-commerce receipts between various suppliers to match them up is not easy work. That's been going on for years. This proprietary data we have is unique. We're not giving it up easily without making sure we're getting paid for it.
Yeah. The flip question is, if you have your proprietary, but then you're telling me your data costs are going down. Like, how are you getting it for cheaper if?
Yeah
Like if it's proprietary, then people should be asking more for it.
Yeah. It's going down because we're giving out greater value proposition to it. They're able to use it in an easier way. Assume you're Walmart, just as an example, and we're getting those detail or POS reads every single day from them, but also they want impacts of what's going on in their particular stores. We're giving them access to our Discover platform, and they're able to ask in a natural language processing that allows that value creation for that store manager or that individual that's responsible for that area to be able to get that value creation a lot higher. That trade-off is changing every single day. I'm not sure that impacts the proprietary data as much as it, you know, it's just a value exchange that's going on. Yeah.
Thank you.
Thank you.
I think we have time for one more question.
It seems like the CPG marketing spending environment is kind of weak. What impact does that have on you? Kind of how big you gather from.
Yeah. The CPG, you know, environment has been weak for a period of time. What we've seen is, you need more data to be able to. If you're gonna reduce your advertising spend or your trade promotion spend, you need our data. You're not gonna fly blind to make those decisions. We're a small portion in the grand scheme of that spend. As a result of it, you know, we're seeing that demand happen as they're thinking about cuts or alternatives or can I, you know, delay this for a year? They're not flying blind. They want our insight to be able to make those decisions. Yeah.
All right.
Yeah.
That's all the time we have. Thank you so much, Mike-
Thank you.
for joining.
Thanks.