All right, I think we're live here. Morning everybody. Thanks for joining us today. We are here for the Trimble session. I'm Quinn Fredrickson, senior research analyst at Baird, covering advanced industrial technology. Trimble is a great example of the blend of physical and digital in the application of AI, with that blend primarily occurring in the construction, geospatial, and transportation markets. We're glad to have Phil Sawarynski, who's Trimble's CFO, here today for our discussion. Phil is going to make some opening remarks and go through some slides, and then we will come back for Q&A.
Great. Thanks, Quinn. Thanks for having me. It's great to be back to the conference again this year. Let me start off, quick overview of Trimble, and we've made a lot of transformations over the years. How would I describe Trimble today? We're a platform company, and our main areas of focus in the industries, as you mentioned, are construction, transportation, logistics platforms. Our strategy is connect and scale. Connect, we connect our stakeholders across the industry workflows. In construction, that looks like concept to completion to maintenance for a construction project. It means connecting the architects, the engineers, the contractors, the owners. In transportation, it looks like connecting the stakeholders and the workflows from shipping all the way through settlement. What is the opportunity?
As we look at construction, there's over $10 trillion being spent in construction, and in any given project, up to a third of that needs to be reworked and wasted, and a big part of that is due to having poor data and miscommunication. 90% of construction projects are either over time or overrun on cost. How does Trimble differentiate? We're on both sides of the equation. Trimble's in the field and the office, the physical and the digital, the hardware and the software. We take the ground truth from the field, we connect it to our best-in-class digital capabilities via our platforms, and the result is having the right information from the field going to the office, and the right information from the office going to the field. Our stakeholders are connected, and they have a single source of truth seamlessly.
That intersection of the physical and digital that I just mentioned, it provides us, Trimble, with a structural data advantage along with that trust of our customers. That manifests itself into petabytes of data within our platforms and our ecosystem, and the trust from our customers that that data is secure. The connection across those workflows means that we can actually solve higher-order problems in extremely complex industries. This is where we can provide a much higher ROI for our customers than just point solutions. How do we think about this in the AI world? That becomes then the foundation for AI, and it allows AI to accelerate that productivity that we're already giving to our customers.
Think about an agent that is grounded in the physical reality, that can reach across the depth and breadth of the capability of Trimble within the platforms, connects not only Trimble solutions, but third-party solutions, along with the customer's unique data to solve these high-order and complex problems. Let's think about the transformation that we've done over the years from a financial model. You can see here, going back to 2020, what's structurally different? We've simplified, we focused, and we strengthened the business and the company. That looks like almost 80% of our revenue being software services are recurring. We have the most visibility into our cash flow and into our revenue that we've ever had as a company as we go forward. Our operating margins and our EBITDA margins have grown almost about 500 basis points since 2020. We're expanding the margins.
Our CapEx is light as a company. Our net working capital is negative. Think about capital expenditures, or sorry, excuse me, capital allocation. Year to date, we've had over $550 million of share repurchases. We have about $770 million left on our share purchase reauthorization. If I look back to 2025, the beginning of 2025 through today, we've actually repurchased about $1.4 billion of shares. We're returning capital to our shareholders. For our guide this year on an EPS growth, we're expecting approximately 17% EPS growth. On this slide, here's a summary of the business, or by business geography and revenue type. You can see between AECO and Field Systems, the construction represents about 85% of our revenue. We are a global business. We have a little bit more than half of the revenue in North America. I'm really excited about the trajectory of the business.
I'm even more excited about the opportunity and what we're doing around the world of AI, and being able to leverage those capability, the data that we have, and that we continue to grow and expand upon. We see a significant opportunity and potential to not only increase the penetration of the market as it exists today, but also expand the addressable market, leveraging AI and delivering value to our shareholders. One thing I'll say before I close and then go to the Q&A, we just recently, or actually going on right now, is our Dimensions, which is our user conference. I just came from there. That's in Las Vegas. We're expecting over 8,000 of our customers. And I can say the excitement of what we're presenting, where we're going, and what we're doing, I've never seen it higher in my 17 years at Trimble.
I'm really excited about where we're going. I'm excited about the opportunity set in front of us, and we're well-positioned with our competitive moats widening and deepening, and our customers are reflecting the excitement that we see as well. Turn it over to Q&A.
Great. Thanks, Phil. For the audience, if you have any questions, you can send them up to the iPad, session3@rwbaird.com. I think it would be helpful just to start off with a little bit of an update on the third quarter print in terms of what you saw. What were some of the key higher-level takeaways for you, Phil, in terms of demand, margin performance, and ARR growth?
Yeah. Thanks for the question. Let me sort of go around the company. Starting with AECO, which is our construction software segment, I characterize that business as fundamentally sound, and we see that in a couple of metrics. Our net retention rates, excluding SketchUp, which has a little bit of a consumer element to it, is around 110%. That has been very stable. Our gross retention rates have been stable as well in that business. Our net new ARR is growing throughout the year. We expect to end the year or exit the year at the highest level for the year. I like the trajectory there. When I look at our remaining performance obligations, it is about, for our recurring business at the company level, it is about 13%, and AECO has the majority of that.
If I look at actually the subscription portion of that business, the pure subscription, it is actually higher than the 13%. When I look at the business delivering the results, we actually increased our guidance for the year for the operating margins to 36%, which is about 100 basis points we had earlier. I would say fundamentally, the AECO business continues to perform. When I look across the Field Systems business, the civil construction continues to perform as well. Within the ARR profile, we had a couple of discrete events that we talked about on the call, but I will just reiterate here. If I look at a year-over-year basis as we exit the year, I will put it into sort of three buckets. One is a white-label product that we talked about, I think, last call or last quarter, which is a product we were not making any money on.
We were buying from a third party. We decided to do that in-house, so build it in-house. That ends up winding down by the end of the year, so that ends up putting a headwind on our ARR for that product. It is the right decision economically for the company, and we just announced at Dimensions that capability, so we have now come out with our version of that that is embedded in our WorksManager product. The other two categories I put it in are macros. Our T&L business is a transaction business through Transporeon in Europe. About two-thirds of that business is transaction. The market we see, both spot prices and transactions have come down relative to the first half of the year, and so that ends up impacting the ARR calculation.
Now, what I also look at is the gross customer retention, and that's close to 100%, which tells me that it's not about customer churn. It's more about just the overall macros in the market. The other part of the macros is part of our Field Systems business. We sell correction services into the agriculture market, and in particular, into Brazil and South America, and there's been more challenges as we continue throughout the year within the agriculture market as we go throughout the year. There's one more discrete item on the ARR, which is we were attaching these correction services to some of our hardware and putting that into our construction and our geospatial markets. This was a little over a year ago, we started doing this. We had some expectations on the take rates on the other side of that.
Those end up coming in a little bit lower. To me, economically, it still made sense to do because I think we're finding incremental customers, just not at the rates that we were anticipating at the beginning of the year. So those are sort of the three items. Fundamentally, I think when I look at the businesses, they're all continuing to perform, and even with the macro headwinds in Europe with transportation, we're still growing that business in a challenging market.
Okay. That's helpful. I wanted to touch on just those discrete headwinds in field. First, just the intentional exit you mentioned of the white-label data processing service. Can you frame up why that was the right decision? Why exit now? You mentioned the internally developed replacement product. Any thoughts on how we think about the cadence to the ramp of ARR for that product throughout 2027, recognizing it's going to be growing off a zero-dollar base?
Sure. Yeah. Great question. Yeah, so again, the margins on that product were very low for us, so we made the economic decision to again, build it in-house. We knew the product was going to be available. I think it's going to be available in January, but we actually did beta release at Dimensions. So it made sense for us long term and, well, not even long term, short term, as we start to build these, where it's going to be a much shorter time to get to accretive profitability. I would expect, again, you made the point, we're effectively growing from a zero base, exiting the end of this year for that product line, and we expect to grow it throughout next year. Now, where I take comfort, if you will, on the product and the launch is, these are customers we already know.
These are products we already know. We sell through dealers. Our dealers already know this. This isn't a product-market fit concern. It's really around that ramp throughout the year, as I expect for 2027.
Okay. T hen the ag market weakness in your positioning services, is that purely a macro issue, or was there any internal execution issues? What portion of Field Systems ARR is represented by positioning services and ag specifically within that?
Yeah. Good question. Look, the majority of that is really the macros and the challenges around that business. South America tends to have a lot of the correction services, just given the size of the farms and the fields there, where the satellite-based corrections just are a bigger portion of that, and that geography has just been more challenged than the rest. I mean, the overall ag market has been challenged, but even more so in South America this year in that business. So I'd say that largely is just a macro impact.
Okay. And sizing up the exposure there within ARR?
Yeah. So call it roughly, so we have about $400 million of ARR that is the Field Systems business. About half of that is what we call our positioning services, and there are other products within there. Call it rough numbers, just half of that is related to the correction services, and then a large portion of that is actually in the ag, focused on the ag market.
Okay. The third discrete item there, the positioning services, the corrections in the construction market. Could you just explain the dynamics of why those renewals after that first year kind of trial came in a little bit light of your expectations and caused that reduction in the ARR guidance?
Yeah. We have had experience in this, so a lot of where we have grown the business is what we do is we include the services for the first year when you buy the hardware. These are corrections that if you do not need a terrestrial network, you can use the satellite to get high-accuracy points on the Earth. We have done it in ag, we have done it in other places. We have experience doing that, but we now decided to take that into the construction and geospatial market. There is a lot of value for our customers to be able to do that. If the network goes down or if there are other issues, they can continue to work, which is really important. We had some expectations on the modelling. We knew it was not going to be like the ag market because the dynamics are different.
We understand the construction and the geospatial market. But when we actually created the models, we were a little bit more optimistic on those than what we are seeing. Some lessons learned that we have had out of that. We think we have got some good ideas how to better educate the market on the value of the products as we go forward. But yeah, we were under the expectations that we thought coming out of the gate.
Yeah. Have you reset those conversion assumptions on the remaining pieces of that business as it comes up for after the one-year renewal, and just how should investors get comfortable that we've now de-risked the hit to ARR from this bundle conversion dynamic?
Yeah, that's exactly right. I think there, again, there are some lessons learned here as we went through the modelling. We have a much better handle on our expectations for these products going forward. A big portion of when we took the Field Systems ARR, we said the low single digits now exiting this year was to exactly that, is applying those lessons learned and de-risking the ARR as we go into 2027.
Got it. Okay. Maybe switch back over to AECO since that is 65% of your ARR. As you said, the business seems to be healthy and performing as you would expect with 13% ARR growth in the quarter. Could you just discuss what you're seeing from an underlying market health perspective, which verticals are currently seeing strength that's enabling you to drive that level of ARR growth, and any color on your visibility on this segment into 2027?
Yeah. Let me sort of start with the high level. If I go back to Investor Day, we talked about at the company level $72 billion of addressable market and about 25% penetrated. The industries we're in and why it's attractive for us is these are large global industries that are underserved, under-penetrated with technology, which is what we do. We see that opportunity as more secular than cyclical to be able to continue to penetrate in these industries. So, when we talk about the opportunity set in AECO in particular, we said about a $50 billion addressable market that's only 20% penetrated. We see the more and more digitization in that industry, and that shows itself in not only Trimble's performance, but other competitors in our space.
We're all growing, and that tells us that there's more digitization that's being adopted with our customer sets. We like the opportunity set. We like our setup in particular, having the field-to-office and the hardware, software connectivity. We're seeing, we still see the same opportunity. We look at our customers' backlogs. That's something important to us as we look at those opportunity sets. As I mentioned earlier, when I look at the RPOs as we go into 2027, I think that, again, 13% on the recurring side of the business, that gives us some good visibility as we go into 2027.
Great. Wonder if you could break down just the growth algorithm a bit for this business. I think you've talked about new logos, cross-sell, upsell. Just how should we think about how that algorithm breaks down, which are contributing most to growth right now, and kind of the remaining runway?
Yeah, I think it again, it dovetails into the prior question. We continue to see, again, the net retention, the net at the 110%, excluding SketchUp. So we're selling, continue to sell and upsell into our existing customer base. Roughly a third of our bookings are in new logos, which obviously we want to land more and more customers, new customers, and that gives us more of a base to do the cross-sell and the upsell opportunity. Investor Day for the AECO business, we talked about over $1 billion of opportunity for that cross-sell, and we continue to look at moving our customers via bundles from one product to two products to three products to four products.
What we see is the more and more we can add more products to our customers and they buy more, our whole bundle and all of the offerings become stickier with those customers. So we really like that play in the cross-sell. I also want to characterize the cross-sell as not $1 billion as a static number. As we continue to do more and more cross-selling within that existing customer base, we're also adding more capabilities, both organically with our development at Dimensions that I was just at. We just announced scheduling, which is a new offering. That's another part of the portfolio to be able to connect the when to the where, why.
When we talk about our platform, that is another part of the platform where you can take the scheduling and the timing element, you tie that to the field, to the office, and then to the project management, and that is something that you cannot get via an application because you are connecting the full workflows, you are looking across the entire platform, and you are solving higher-order problems. So that cross-sell opportunity, we continue to add on the other side, and I believe that on the other side, that continues to grow as well.
Okay. It is related, but I often get the question from investors just how do we think about Trimble's ability to continue to grow ARR at this low-to-mid-teens rate that you have been growing the business at, given you are kind of bumping up against that law of large numbers. So just any thoughts on that topic?
Yeah, and it is a great point. I mean, we are a scaled business. The AECO business is approaching $1.7 billion of ARR this year, so we do have a denominator effect, and that is one of the reasons why I mentioned earlier the net new ARR, because I see that still growing, which says the business is still stable, we are still healthy, we continue to grow, but there is a law of large numbers as you get into that customer base. But then I take a step back and I also talk about that addressable market for us, and it is only 20% penetrated. So I think that gives us the ability to continue to grow. The other piece of this that also we believe is that the addressable market can also increase.
The more and more AI adoption that we have, we think that can also expand the addressable market as we go forward. We like the setup and we like the opportunities in front of us.
And I think now might be a good time to bring in the AI discussion, which I think no software discussion would be complete without touching on that. So just within your AECO portfolio of the major buckets, the architectural, engineering, construction, owner pieces, just how would you rank the defensibility? Which capabilities are easier to replicate with AI and which depend on engineering accuracy, embedded workflows, or physical world execution?
Yeah, so I think ultimately, so we do have AI embedded in the products and the applications, but I think what's important is where Trimble is positioned as a platform company, and more importantly, that we're positioned with both the digital and physical, that connection from the field to the office. Because ultimately, when you look at a project and you look across all of the stakeholders of that project, these are really complex issues. When you're connecting the architect to the engineer, to the contractor, to the owner, you need to have a source of truth. As the projects continue to change, if you're building a building, for example, and there's an architectural change, the engineer needs to know that. They need to know if they need to re-engineer something in a building. The contractor needs to know what they're building and if something has changed.
This is where that inefficiency comes in that I talked about earlier, is you got to have a single source of truth operating with the same information. If you don't, this is where all that rework, all of the waste happens. When you think about Trimble and you look across our platform and the fact that we have all of those capabilities and have all of those workflows, we connect them to the field. So those are the people digging the dirt, putting up the drywall, putting up the beams, manufacturing the steel in the steel fab shops. Having that connectivity across there is something that I think is unique to Trimble. You can't do that just by having a single application. You have to have the platform, you have to have the connected workflows, you have to have the stakeholders in the environment.
When we look at some of the numbers of our customers working in there, in Q3, we had over 30 billion API hits. We had over 1 million projects that were created in Connect. We're bringing the customers in, the customers are doing the work there, but they're looking across, and when we think about, you mentioned, you were asking about the AI agents. The AI agents, and in our case we call Trimble Assist, is the capability of having the AI agents not just looking within the application, but looking across the applications, looking across the workflows, across the platform, and being able to connect that easily for our customers and be able to connect the dots through all those life cycles to be able to solve those higher-order problems. This goes back to what I was talking about with the connectivity.
The schedules intersect with the field-to-office . Something changes, how do I know what that means from a scheduling standpoint? What does that look like from a project? How do I move my assets around? We acquired a company called Document Crunch earlier this year, which is around contract risk. What are my obligations in the contract? By having all that connectivity, all those applications, all those workflows, all those capabilities sitting in Trimble's platform, we are in a unique situation that is really hard to replicate.
How do you monetize all this value that you are providing to your customers? Maybe if there is any details you can share with us in terms of what the mix of the portfolio looks like in seat-sensitive versus some sort of hybrid consumption or outcome-based pricing, and just how quickly you can move to more of that hybrid monetization model?
Yeah. It is a great question. We look at it as an and or a hybrid. Right now, the majority of what we sell are named user licenses. What we do is we separate, we will have good, better, best type offerings within there where you can have different capabilities, different features, and functionality. The more capabilities you want to have, we would bring them up the stack as an upsell into a higher tier within that. Now, we also see the ability on a consumption basis to augment those named user licenses, and this is where I was talking about the Trimble agents and the platform.
When we look at our agent layer on top of our platform, there are agents that Trimble can design for customers to use, and the more and more they use that, we can see that as an opportunity for more consumption. What is also part of that platform is the customers being able to not only connect the Trimble offerings, but third-party offerings and their own information to create their own agents within our platform. Why is that important? Well, within our platform, what our customers want is also the trust, and we can create the security that they can use and they can go in, create the agents within there versus in their own environments and having to develop all the security and everything that goes around that.
We see that whether it is a Trimble agent or that it is your agent, you are actually operating within our platform and you can apply consumption to that. It is early, so I am not going to call a needle mover yet. We see some anecdotes I will give you in our SketchUp offering. For example, what we did there is we had the good, better, best offering. Within each of those, we include a certain amount of what we call Trimble Credits, which are our tokens. Once you exhaust those, you have to buy another block of tokens. What we have seen is quarter- over- quarter, a doubling of that. The base is small, it is not material at this point, but it is at least an indication of, number one, where we are going, and number two, that to some extent, our customers are willing to pay for more tokens.
It is early and it is on a small base, but it is at least one data point.
How about internally within your own four walls? We have been asking every company at the conference just how the internal implementation of AI has changed at your company versus last year in terms of use cases, outcomes, and spending.
Yeah, I will put it into two categories. One is really the external, and this is the revenue drivers. What we are seeing there, and we have reorganized how we operate in product development, what we are seeing is our ability to accelerate the time to market for our organic offerings. At Dimensions, we have a lot of new products that are coming out. A lot of that is a manifestation of us just being able to move faster, deliver more products to the market, and I see that as a revenue side. When I look also from an efficiency side internally, I see more and more adoption of AI automation, and we are seeing efficiency for, let us say, the non-front-facing or customer-facing.
I look a lot at the back office and what we're doing there, and where we see some of that manifest is we talked on the call earlier around roughly $80 million of run rate cost we're getting out this year. And a big part of that is because we're changing how we actually operate, how we work throughout the company, and that's leading some to efficiencies.
Great. Last one I wanted to ask about transportation and logistics. You did announce a strategic review following some inbound interest in the second quarter without discussing the process or potential terms. What determines whether this business creates more value inside of Trimble versus under another owner?
Yeah. It's a great question, and I'll say that we continue on the process. From a timeline, there's nothing new. We're as expected internally. How we think about this, I look at it through sort of three lenses, and that's the shareholders, the customers, the employees. Where's the best fit? What makes sense for all of those? As we go through the review and we come to a conclusion, I mean, ultimately it's going to be where can we provide the highest value for our shareholders? And that can manifest in a sale company, it can manifest in keeping the company. We need to go through the process, but fundamentally, we're going to be anchored on where we can provide the best shareholder value to our owners.
Great. I think we're out of time, so thanks, Phil. Appreciate it.
Yeah. Thank you.