BSY - NASDAQ

Good morning, and thank you for joining Bentley Systems Q2 2025 results. I'm Eric Boyer, Bentley's Investor Relations Officer. On the webcast today, we have Bentley Systems Executive Chair, Greg Bentley; Chief Executive Officer, Nicholas Cumins; and Chief Financial Officer, Werner Andre. This webcast includes forward-looking statements made as of August 6, 2025, regarding the future results of operations and financial position, business strategy and plans and objectives for future operations of Bentley Systems, Incorporated. All such statements made in or contained during this webcast other than statements of historical fact are forward-looking statements. This webcast will be available for replay on Bentley Systems' Investor Relations website at investors.bentley.com on August 6, 2025. After our presentation, we will conclude with Q&A. And with that, let me introduce the Executive Chair of Bentley Systems, Greg Bentley.

Good morning, and thanks to each of you for your interest in BSY. Please pardon my voice, which is suffering from a summer cold. CEO, Nicholas; and CFO, Werner, will, as always, report in detail Bentley Systems continued excellent operating and financial results for '25 Q2, and thus, for the first half of 2025 as we track consistently towards our outlook range for this full year. Earlier this year, in reviewing 24 Q4 I looked back over the years since BSY's IPO in 2020 to quantify that our outlook range for 2025 would complete the process of at least doubling over these 5 years each of our key financial metrics of ARR, revenues, adjusted operating income less stock-based compensation and free cash flows while minimizing equity dilution. Then most recently, in reviewing '25 Q1, I likewise looked back 5 years to quantify the respects in which we've purposefully gained further business resilience over the span. It is even more clear this quarter that we're currently benefiting from those improvements. Our excellent operating performance to date in 2025 is in keeping with a primary sustaining long-term growth driver over this period and which will prevail foreseeably. Going digital has become the enduring priority for infrastructure engineering, in particular, because of pervasive resource constraints. To keep up with the world's imperatives for infrastructure performance, resilience and adaptation, each BSY user and account needs each year to achieve step functions in productivity and value generation through enhanced utilization of software, cloud services and AI. To help quantify such progress and software consumption per engineer, I would like now to, again, but for the last time, I think, look back over 5 years but this time with reference to external market data. This slide, which we still use in our intro deck today, shows global counts of engineers and related technicians and their software expenditures as tabulated by U.K. global research firm, Cambashi. Conveniently, for our process of monitoring the long-term trend, this last data that we had on engineering employment and spending is for 2019, immediately preceding BSY's IPO. We originally compared spending per engineer/technician for infrastructure engineering to that of product engineering to glean a data point for the market potential headroom. But now for the purpose of this look-back, let's examine the changes in just infrastructure engineering spending over time as Cambashi has just provided an update, which slightly refines 2019, and which most importantly, introduces the most recent year for which this data, including engineering employment is available, 2023. I do not find it surprising that over these 4 years, the total number of infrastructure engineers and supporting technicians has only increased by about 1% per year on average. While there isn't sufficient granularity in both data sets to establish this, I believe that even this nominal increase is concentrated in less developed countries, while I believe that in countries like the U.S., infrastructure engineering retirements have exceeded new graduates. Most significantly, Cambashi finds that software spending by infrastructure engineer/technician has grown at a compounded annual growth rate of about 10% over this period. And this is in nominal rather than constant currencies. Cambashi's estimate is that the constant currency growth rate was a full percent higher. By virtue of BSY's constant currency revenue growth rate during the period, we somewhat outgained this broader market. Hence, software spending per infrastructure engineer/technician has grown about 9% in nominal currencies or indeed approximately 10% in constant currencies over this period. Even so, the $514 annually per engineer or technician tends to appear low compared to averages for BSY users. Usefully, for this newly available 2023 data, Cambashi has provided refinements for better understanding. To start with, we can now focus on employment and spending just for engineers rather than also including the technician categories, which are rather miscellaneous and less representative of BSY's primary user profile. See here Cambashi's observation that in 2023, 17.6 million infrastructure engineers globally spent $10.2 billion on engineering and GIS software averaging a relatively higher annual spending of $579 per engineer. For this new 2023 data, Cambashi also provides analysis of employment and spending down to the level of countries classified together within the 5 quintile tiers of per capita GDP. The wide variation in average spending is striking to me. The engineers in the most developed Level 1 countries, while nearly the least numerous, spend the significant plurality of the global total, averaging over $1,900 per year per engineer. This is generally consistent with BSY's average and global uniform pricing and utilization for E365. So the numerical bulk of infrastructure engineers in less developed country levels represent a multiple of long-term upside opportunity as they inevitably will tend to catch up in going digital and in expenditures to do so. Now coming back to Bentley Systems and our standing as the market leader in so many infrastructure engineering market segments, as we show in our own assessment of the competitive landscape in this slide from our intro deck, let's examine Cambashi data to understand how so much of this $10 billion in software spending by infrastructure engineers obviously goes elsewhere than to BSY. To do this, let's parse the full $14 billion of now 2024 software revenue, which Cambashi ascribes to BIM. This term originally referred to building information modeling decades ago, but has come to connote as here a catch-all for what could otherwise be described as AEC for architecture, engineering and construction as opposed to product engineering or manufacturing. Within this total, BIM design is not only the largest submarket, but we believe it is the prerequisite to success throughout the others. The greatest ultimate opportunity is for digital twins to support infrastructure, operations and maintenance. Although this submarket is only nascent so far, we think our asset analytics initiative portends its comparatively unlimited potential monetized per asset rather than per engineer. But the potential compounding value of a digital twin depends significantly upon its assimilation of the digital context, digital components and digital chronology from the BIM design. So here, per Cambashi, are the market shares within the $6.5 billion of BIM design revenues. The largest by far is for Autodesk followed significantly by Bentley, Schneider, Hexagon, Nemetschek, Trimble, and in all China, I have grouped together all the Chinese companies identified by Cambashi listed here in the fine print. To help understand the widely varied mix across infrastructure segments per competitor within BIM design, Cambashi analyzes each company's such revenues by product category within plant; architectural, which includes mechanical, electrical and plumbing or MEP; civil; and structural. Looking here at the relative proportions for each company, one sees says that every one of these punitive competitors has a considerably different primary focus than Bentley Systems. At Bentley Systems, we don't break out product revenues in these same categories. and I suspect that Cambashi's estimates here seem to discount Seequent's major share of geotechnical and environmental software used in civil engineering. While we have a comprehensive portfolio across all disciplines required for major infrastructure projects, including their plant and building engineering aspects, our focus is characterized by civil in supporting structural products, primarily for horizontal infrastructure networks. By contrast, Hexagon and Schneider's design products are clearly focused primarily on plant engineering Nemetschek design products are concentrated in architectural/MEP. Trimble's design products are primarily divided between architectural MEP and structural to support such buildings. China has more or less official state-owned software for its country-specific structural analysis code and many competitors for building software. Finally, the primary design product focus for our principal competitor, Autodesk, is architectural/MEP for vertical infrastructure buildings. Within Autodesk's civil product share, its offerings for civil site design likewise associated with buildings loom large. And as you may recall, we at BSY see this as a substantial competitive upside opportunity. Our BIM design software peers and competitors are worthy and resourceful. While our space is desirable and envied, each of these others have a principal focus that is clearly different. I believe that we are favorably differentiated by virtue of our established franchise as the outright market leader in comprehensive infrastructure engineering software, particularly for the horizontal networks of public work/utilities, grids, roads and bridges, rail and transit, water and wastewater and resources and the geo professional disciplines, structural disciplines and project delivery collaboration across all of this. Given the world's prevailing imperatives for infrastructure performance, resilience and adaptation, I wouldn't trade positions with any design software peer or competitor. Notwithstanding this favorable competitive backdrop, our consistent execution should not be taken for granted, and I commend our management and colleagues for '25 Q2. Over now to Nicholas.

Thank you, Greg. We delivered another strong quarter despite ongoing global uncertainties. This performance underscores the resilience of our business model and the strength of our end markets driven by secular infrastructure investment. As demand for better and more resilient infrastructure continues to outpace the available engineering resource capacity, our software plays a crucial role in helping infrastructure engineers achieve more with less. Our strong first half performance reinforces our confidence in meeting our full year outlook based on low double-digit ARR growth, continued margin expansion of approximately 100 basis points and robust free cash flow generation, consistent with our long-term financial framework. Turning to Q2 highlights. ARR grew 11.5% year-over-year and 12% when excluding the impact of China. Growth in the quarter was underpinned by a net revenue retention rate of 109%. E365 continues to be a growth driver with renewals consistently reflecting stronger commitment levels. The willingness of accounts to commit to higher contractual floors in return for corresponding ceilings signals their confidence in the strength and sustainability of their own demand environment. In Q2, we once again added 300 basis points of ARR growth from new logos, primarily within the SMB segment. And for the 14th consecutive quarter, we added more than 600 new SMB logos through online store. Retention within the segment remained high, further signaling confidence in the demand environment this time from smaller accounts that are arguably more sensitive to economic uncertainty. Turning to our tone of business by infrastructure sector. Resources was our fastest-growing sector this quarter with Seequent delivering a particularly strong performance. Notably, growth for Seequent was led by mining, outpacing civil for the first time in 6 quarters. While we are seeing early signs of improvement in mining exploration, it is still too soon to call it a market recovery. Public works, utilities delivered another solid quarter performing in line with the company overall supported by sustained global infrastructure investment. The needed focus on grid resilience is particularly benefiting Power Line Systems. Growth in industrial sector remained modest, while commercial facilities was flat. Performance across regions in Q2 remained largely consistent with prior quarters. Growth in Americas was once again solid with Latin America continuing to stand out. In the U.S., our accounts remain confident in their outlook for the year despite ongoing uncertainty related to tariffs, policy shifts and regulatory changes. There's also increased optimism that Congress will now prioritize meaningful permitting legislation. The combination of executive orders, updated agency guidelines, recent court rulings and bipartisan state-level initiatives all signal momentum towards comprehensive permitting reform, particularly in areas such as transmission, critical minerals and other strategic verticals. Our Power Line Systems and Seequent businesses, both standout growth drivers since their acquisition 3.5 and 4 years ago, respectively, are especially well positioned to benefit from these developments. Also noteworthy, Congress has already begun working on the surface transportation reauthorization of the IIJA 1.5 years before its scheduled expiration, a clear indication that civil infrastructure investment remains a national priority. In EMEA, we delivered another quarter of solid growth with the Middle East continuing to lead the region, followed by the U.K. Investment remained strong across transportation, energy and water infrastructure while momentum continues to build in defense-related projects, data centers and nuclear. Together, these trends reflect a broadening base of demand. Recent announcements in the U.K. and Europe point to a continued supportive funding environment into the foreseeable future. In June, the U.K. government published its 10-year infrastructure strategy, which earmarks GBP 725 billion in long-term funding and aims to attract private investment into its national infrastructure. And in July, the European Commission put forward its EUR 1.8 trillion proposal for the next long-term EU budget for 2028 to 2034. The proposal clearly prioritize investment infrastructure and should be a continued tailwind for funding in years to come. In Asia Pacific, the overall performance remained steady. India continued to stand out with positive sentiment across strategic national programs in water and power. AN

Thank you, Nicholas. We've had a strong first half of the year and are well positioned with respect to our financial outlook range for the year. Total revenues for the second quarter were $364 million, up 10% year-over-year on a reported basis and 9% on a constant currency basis. Year-to-date, total revenues grew 10% on a reported and constant currency basis. For the first quarter and year-to-date, strong growth in subscription revenues was partly offset by a reduction in professional services, and to a lesser extent, in license revenues. Subscription revenues now represent 92% of total revenues, up 2 percentage points from the same period last year, reflecting improvement in the overall quality of our revenues visibility, growth consistency and margin contribution. Subscription revenues grew 12% year-over-year for the quarter in reported and 11% in constant currency. And for the first half, more normalized for mix and timing, subscription revenues grew 12% on a reported and constant currency basis. Our SMB and E365 initiatives continued to be solid contributors. Perpetual license revenues for the quarter were $10 million, down $1 million year-over-year. Perpetual license sales make up only 3% of total revenues and will remain small relative to our recurring revenues. Our less predictable professional services revenues declined 7% for the quarter in reported and 9% in constant currency and now represents 6% of total revenues, down 1 percentage point from the same period last year. It is still the case that the largest portion of these nonrecurring services relate to IBM Maximo implementation and upgrade work. Our last 12 months recurring revenues, which includes subscriptions and a small amount of recurring services, increased by 13% year-over-year in reported and in constant currency and represent 92% of our last 12 months total revenues, up 2 percentage points year-over-year. Our last 12 months constant currency account retention rate remained at 99%, and our constant currency net retention rate rounded down to 109% led in magnitude by accretion within our consumption-based E365 commercial model. We ended Q2 with ARR of $1.379 billion at quarter-end spot rates. On a constant currency basis, our year-over-year ARR growth rate was 11.5%, consistent with our expectations. Excluding China, our year-over-year constant currency ARR growth rate was 12%. China is now 2% of our total ARR. On a quarterly sequential basis, our constant currency ARR growth rate was 2.7%, slightly below our '24 Q2 sequential growth rate of 2.9% and impacted by the timing of programmatic acquisitions and asset analytics deals. With regards to seasonality, we expect '25 Q3 to be our seasonal low quarter for year-over-year ARR growth due to the timing of potential acquisitions and anticipated asset analytics deals, which were particularly strong in '24 Q2 and '24 Q3 to be closer to year-end in 2025. And the favorable impact from onboarding the Cesium acquisition will be dropping off in '25 Q3. Our GAAP operating income was $84 million for the second quarter and $200 million year-to-date. I have previously explained the impact on our GAAP operating results from deferred compensation plan, liability revaluations and acquisition expenses. Moving on to adjusted operating income less stock-based compensation expense, our primary profitability and margin performance measure, adjusted operating income less SBC expense was $105 million for the quarter, up 10% year-over-year with a margin of 28.9%, up 10 basis points. Year-to-date, adjusted operating income less SBC expense was $231 million, up 11% with a margin of 31.5%, up 40 basis points. Our margin performance for Q2 and year-to-date has been strong, particularly when considering that the year ago period benefited from head cost run rate savings associated with the '23 Q4 strategic realignment program. Our '25 Q2 margin benefited from our mix shift towards higher-margin subscription revenues and from certain discretionary OpEx spend being slightly more back half loaded in 2025 when compared to 2024. We remain confident about delivering our 100 basis points full year margin improvement target. As a reminder, our OpEx seasonality is always more heavily weighted towards the second half with our annual raises occurring as of April 1 each year and our larger promotional and event-related costs also concentrated in the second half of the year. Our free cash flow was $57 million for the quarter and $273 million year-to-date. This is generally consistent with our expectation that based on our seasonality of collections and expenditures, free cash flow for the first half would be on the order of 60% of our full year free cash flow outlook of $415 million to $455 million. In addition, we estimate that our second half of 2025 will see cash tax benefits in the range of $15 million from the recently enacted One Big Beautiful Bill, primarily attributable to the restoring of immediate U.S. tax deductions for domestic R&D expenses. As a result of this tax policy change, we are raising our 2025 free cash flow outlook to a range of $430 million to $470 million from previously $415 million to $455 million. With regards to capital allocation, along with providing sufficiently for our growth initiatives, during the first half of the year we deployed free cash flow as follows: $135 million fully paying down our senior debt, $75 million in effective share repurchases to offset dilution from stock-based compensation, $10 million in convertible senior notes repurchases and EUR 42 million on dividends. With our senior debt being fully paid down, our net debt leverage, including all of our 2026 and 2027 convertible notes as debt was 2.4x adjusted EBITDA, down from 2.9x at the end of 2024. Our strong balance sheet and projected free cash flow generation will sufficiently fund our dividend, share repurchases and growth initiatives, including potential programmatic acquisitions. Our 5-year senior secured credit agreement dating from October 2024 provides a currently undrawn $1.3 billion revolving credit facility. Combined with our strong balance sheet and anticipated future free cash flow generation, this affords sufficient flexibility, if needed, to refinance the January 2026 maturity of $678 million in outstanding convertible debt, while keeping our cash interest thereafter at about the same magnitude as in the recent past. Interest rates on our debt are protected for very low coupons on our convertible notes and very favorable terms on our $200 million interest rate swap expiring in 2030. And finally, with regards to our outlook for the year, our financial performance for '25 Q2 put us in a solid position to deliver within our annual outlook range for ARR growth, revenues, profitability and now increased free cash flow. With regards to foreign exchange rates, for the second quarter, the U.S. dollar has weakened relative to the exchange rates assumed in our 2025 annual financial outlook, resulting in approximately $7 million of incremental revenues from currency and the total first half favorable impact of approximately $8 million. Based on more recent rates where the U.S. dollar has further weakened relative to our outlook rates, if end of July exchange rates would prevail throughout the remainder of the year, our second half GAAP revenues would be positively impacted by approximately $17 million relative to the exchange rates assumed in our 2025 financial outlook. And with that, we are ready for Q&A. Over to Eric to moderate.

Thanks, Werner. [Operator Instructions] Our first question comes from Matt Hedberg from RBC.

Great. I hope you can hear me okay.

Yes.

I wanted to ask about the macros. It sounded like in some of the prepared remarks that macros remained strong, and you called out specifically strength in SMB. I'm curious, did you notice an improvement sequentially now that we're past some of the initial tariff uncertainty? Or is this just kind of business as usual from your perspective?

Yes, Matt, I'll take this one. I would say it's just very consistent, yes. It's a very consistent environment from the standpoint of all the investment that is going into infrastructure. And yes, there's been noise with tariffs, there has been noise with maybe a change of priorities, et cetera. But overall, it's just a very consistent environment. What we're hearing from our accounts is that they are just very positive about their own outlook whether they are big or small accounts. And if there's one thing that is also quite consistent is the fact that there's really no problem with the demand, there's a problem with the capacity. They just don't have enough engineers. And that remains the backdrop for our own demand environment, which is we are uniquely positioned to help them be more productive with the capacity that they have in order to cope with that demand.

Great. Thanks, Matt. The next question comes from Joe Vruwink from Robert Baird.

All right. Great. I thought the TAM analysis was good at the start. One thing that comes to mind that your products are typically thought of as the premium offering in the various disciplines you serve. In premium because of functionality, but oftentimes that also comes with a certain price point. When you think about reaching the long tail of engineers that are spending below a $1,000, let's say, do you have the right product set and right reach, right go-to-market model to effectively address?

I think we have, indeed, the full range. And the -- our traditional product, MicroStation, by which we started the company is really the entry point for engineers to start working with our software and it remains one of the main growth drivers in SMB. So I think that's quite telling, which is that's where we start. And then when we have accounts who start to use the MicroStation software, they become great targets for upsell to more sophisticated application that will be more specific to certain assets or to a cross-sell with related engineering applications or collaboration software. So we have the full range. And I mean, a testament to that is just the continued growth that we see in SMB. There's just no slowing down. You heard it in the prepared remarks, more than 600 new logos for the 14th consecutive quarter. And again, very often, those new accounts that we're winning through SMB are actually starting with MicroStation.

Thanks, Joe. Next question from Kristen Owen from Oppenheimer.

I want to double-click on some of the TAM analysis and ask you about the data center opportunity. Is there a way to understand sort of the Bentley Systems total addressable market there, whether it's through energy infrastructure, roads that go out to the data center or even some of the water infrastructure? And beyond that, is there sort of a persona that you're thinking about that you can bundle, any sort of potential external partnerships with hyperscalers, asset owners? Just sort of unpack that data center opportunity for us.

I'm going to let Nicholas respond in the main, but I will say that I'm struck during the quarter by the U.S. AI strategy, which articulates that it needs to include an infrastructure investment strategy for the reasons you described. And of course, this isn't limited merely to the place where the data center sits, but the way it's connected up, especially with the electrical grid, that is a particular opportunity for us. But the other aspects of the portfolio beyond merely the building and its structures, our structural analysis software especially is used in that. But a data center is a small city altogether and includes all the other aspects of the ecosystem you mentioned. But Nicholas, you've been looking at this, I think, more in particular.

Well, I will say, Kristen, the way you're looking at it is the way we're looking at it as well, yes, which is, indeed, there's a lot of infrastructure when it comes to data centers that is for the data center itself and then surrounding the data center, as Greg just explained. These are really more like mini-cities or campuses. And then whenever you have, let's say, high spatial density with engineering complexity, this is typically a great opportunity for us. So we do offer software to design many aspects of the data center and the surroundings of the data center. You mentioned the road, of course, the distribution lines, the water and so on and so forth. But we also offer software that helps put all of this together and understand the interrelationships between all of these infrastructure assets. Now I think what's interesting with the hyperscalers is they're obviously software first. They understand the power of software and they're quite receptive to the notion of let's make sure that when we design, we design well from the start for operations. And then we make sure that there is a digital twin that is being created at the moment of design, which is used to schedule and manage the construction then be used for the operations of data center. I'm not talking about the inner operation of the data center, obviously, but all the surrounding infrastructure. I think a good, let's say, testament to that is the uptake of SYNCHRO, which is used quite systematically for the construction modeling of data centers by hyperscalers.

Thanks, Kristen. The next question comes from Jason Celino from KeyBanc.

Great. Werner, I don't know if I heard you correctly, but did you say that the free cash flow guidance was moving up $15 million because of the benefits from the OBBA. I know there are a couple of ways of recognizing the benefit. I mean, not to get too much in the nitty-gritty, but is this the accelerated adoption? Or should we expect multiple years of benefit?

There will be multiple years of benefits, Jason. The $15 million is for the first year and it considers actually that we had prepayments already in the first half of the year. So the normalized annual benefit that we are seeing for '25 would be approximately $10 million higher than the $15 million. But based on the prepayments that we took, this will roll over into the 2026 year. But we are still evaluating, but we don't plan on an accelerated adoption. So that's just a normalized benefit, if you will. And then you will see more benefits coming over there for the following years.

Thanks, Jason. Next question from Siti Panigrahi from Mizuho. Okay. Maybe we'll move on and come back. Can we get Warren Meyers from Griffin Securities, please?

Yes. This is Warren in for Jay. Quick question, aside from the plan to incorporate Cesium where appropriate in the portfolio, what are some of the other critical development and deliverables for the next 6 to 12 months even if you don't expect an immediate impact to revenues? And related to that, over the past few months, Bentley has exhibited an uptrend in the number of R&D openings while keeping sales openings flat. Could you comment on the thinking behind those recent numbers as well.

All right. So on the first question, let me take a step back and say that our R&D priorities are AI, Cesium and an integration of Cesium integrated across the portfolio and then the deeper integration of Cesium with iTwin into a unique platform for the built and natural environment. So with respect to Cesium and our progress there. So I think in last quarter, we talked about Evo, which is our data and compute platform for geoscience data that is coming out of our Seequent business. And this one has already -- had already embraced Cesium as the main user interface. And then as I commented in the prepared remarks, we brought very important iTwin capabilities into the Cesium platform and we announced that at our very first Cesium Developer Conference. The reason why this one is particularly symbolic is you might remember that one of our strategic objectives with the acquisition of Cesium was to leverage the Cesium ecosystem to accelerate the adoption of our iTwin capabilities. And so there you go now, the iTwin capture capabilities that are used to create reality models, be able to visualize that with [indiscernible], being able to run AI on it to detect some features, et cetera. All of those capabilities are now immediately available from within, from within the Cesium platform. Now what is also a great validation of that strategy is that one of the longer-term partners of Cesium, which is EarthBrain, a JV, which is majority owned by Komatsu in Japan, has agreed with us to expand our partnership from Cesium to iTwin and to adopt iTwin capture and the iTwin platform in order to support their own applications for construction, simulation and management, yes. And then just in the last few days, we released MicroStation 2025. So again, our very core product for generic modeling, which now has built in the Cesium integration so you can actually visualize your designs in full 3D geospatial context, including by the way with 3D photorealistic tiles that are coming from Google, leveraging the same technology of Cesium. So some very exciting progress there, and there's more to come. There is the integration of Cesium and Bentley Infrastructure Cloud. And a lot of that we'll be discussing at our annual conference in a couple of months now in Amsterdam yes. And then to the second question, which is the fact that you see openings in R&D and you see, however, openings in sales quite flat, I will say probably 2 things. One is a clear sign that we are able to scale the business without necessarily throwing people at it on the go-to-market side, right? So it speaks a lot to, again, the strength of our demand environment and the strength of our products. The other thing I will say is we're quite fast and probably a little faster in hiring go-to-market resources than we are hiring in R&D resources. And our big focus when it comes to hiring R&D resources is AI, which again is our #1 priority when it comes to investment this year.

The next question comes from Taylor McGinnis from UBS.

So it seems like you're well on your way to hitting the midpoint or even slightly above the constant currency ARR guide for this year. But in terms of the scenario that could push you to the higher end of the guide to 12.5%, what does that look like? You mentioned the permit reform opportunity earlier on. It sounds like emerging growth opportunities with asset analytics and maybe some of the AI stuff continues to move along nicely. So are there any second half catalysts that you're particularly excited about that could be a source of upside for us to monitor?

Taylor, I think you named them to which I would add that our acquisition strategy this year is focused on asset analytics opportunities rather than broad programmatic acquisitions as has been the case in the past. And we're fussy about those and -- but nonetheless determined about it. And we have not had a significant acquisition since last year's Cesium acquisition, which will roll off here in this third quarter. But we are hopeful about such opportunities during the remainder of the year.

Yes, thanks, Taylor. The next question comes from Matthew Bullock from Bank of America.

This is Matt on for Koji Ikeda. I wanted to ask a little bit more about asset analytics. Maybe if you could just help remind us the growth profile there, the run rate today, the key areas of execution focus for that business outside of the programmatic and potentially larger acquisitions in that business line. And then finally, it sounds like there's some seasonality nuance this year, slightly tougher comps from an ARR perspective this quarter and third quarter. Just maybe help us understand the shift in seasonality in 2025.

Matt, I'll say that the business, the asset analytics business, isn't yet of such magnitude that we break it out and track it. What I can say is it contributes a lot of our volatility. Such volatility as we had is largely related to that because the rest of everything is subscription oriented and renewing floors and ceilings and so forth. And asset analytics, we're chasing the major opportunities in the world because we want to get our elbows out and have those proof points. So they're large and lumpy when they come. We have also discovered that sometimes they last only a year because the asset analytics are during a construction phase of something and then you have to sell again to the owner-operator constituencies. And so we're learning this business as we go. And there was reference to the second and third quarters last year in 2024 being when we won the big deals and we're working on a bunch of them now. You may want to speak about one, Nicholas, that has a particular qualitative aspect that we think is auspicious for the future in the case of Blyncsy.

I will. Before I talk about Blyncsy, which is our road monitoring solution, I will say that with the other asset analytics business we have, which is for cell towers, it has indeed been historically volatile because of the reasons that Greg just explained. I think as that business gets bigger, the volatility will decrease, but also as we are refocusing on slightly different type of accounts with slightly different use cases. Historically, we've been working a lot with equipment manufacturers, for example. They will use our software to inspect towers before and after the installation of that equipment. But more and more, we're targeting the actual owners of the cell towers or the mobile operators, and their use case is different because they actually want to monitor those towers on a continuous basis. Therefore, as we win some of their businesses, we will decrease the volatility. But now back to road monitoring solutions, indeed, a very auspicious opportunity here, as Greg just indicated. Traditionally, we are selling to transportation authorities and those can be long cycles, by the way, but very worthwhile. And I think at some point we'll pass the tipping point where there will be so much great examples of success of transportation authorities with our software that others will follow much quicker. But we now have one very large global engineering services firm, which is leveraging the capabilities to offer their own high-end value services to a U.S. territory. The reason why we like that a lot is because it fits exactly our strategy of asset analytics where we don't want necessarily to go after all the owner-creators under the sun. We do want to leverage our strong relationship with engineering services firms. And as they diversify their own business beyond product delivery going into asset operations and in order to offer these high-value services for asset operations and maintenance, we want to be right there with them with our software, with our technology, with our platform in order to allow them to do that. So we welcome that very positive development, and we hope it's the first of many to come.

Thanks. Let's move on to Wolfe Research for the next question.

This is Ivan here for Josh. Can you elaborate on what drove the slight downtick in NRR? And then when do you sort of expect that to go back to the double-digit range?

Well, I'm going to say first, I think it's teetering between 109% and 110%. If you think of our business generally, half of our NRR is from price escalation, the other half variously from the contributors to consumption otherwise. And then there's another 3%, which is from new names on top of the NRR that makes up our ARR growth as that is 11.5-ish or so percent. So the -- it's all rather consistent, but round up or down, you could say they're from 9% to 10%, but I realize that's from 1 to 2 digits also. I think it's kind of accidentally in that range. But Werner, would you like to comment more quantitatively than that?

Yes. I think you were spot on like this. If I go to the numbers, I think it was like 9.45% rounding down to 9%. It is within that 9% to 10% range and it did drop down to 9%. So there's nothing really significantly driving kind of a shift from 10% to 9%. It's within that range and that's a solid range for us.

I guess something to remember about NRR is it looks back altogether fully 2 years. And there's a lot of China in that -- China pressure in that still. At some point, the China ARR downdraft has to slow down because it's down to only 2% of ARR at the moment. But when you look back fully year ago versus 2 years ago, it's pretty significant.

Thanks. Last question will come from Mizuho Securities.

For the earlier mix up, wrong button. But excellent. So we also look at other vertical names like the Cadence, Synopsys. And one of the common themes we hear across all the verticals these days is shortage of engineers and how AI is helping to kind of like help that problem through agentic AI. That is like the rave in AI world these days. So given all the acquisitions and Cesium and the inroads into the AI world, I'm curious how you're thinking about agentic AI in your products going forward? Where are you in the process right now? And what's your vision going forward? And how does it help elevate the shortage?

Siti, I'm going to ask Nicholas in the main to respond, but that -- this notion of quantifying consumption volume number of engineers, especially, is a reason that I wanted to look back and get new data from Cambashi on the census of engineers. And of course, the past 5 years won't be like the next 5 years with agentic AI. But in effect, we've been experiencing some of what comes about when there are fewer users, I think there are fewer users in the advanced economies in those Level 1 countries. They are fewer, but they're spending more, but you can't assume there's going to be more of them or that they will be spending more time on doing one thing because for one thing, they're going to be having their agents spending some of the time. And I know Nicholas is going to talk perhaps about some of the directions commercially that, that implies. But generally, what I'm confident about is that they'll each spend more. The particularly unique aspect of this for infrastructure engineering is that the bulk of our end markets or half of them, the engineering firms themselves have primarily been charging per hour. And that, AI is going to change the way in which they will need to monetize their services. So we can help them and participate with them in ways of getting compensated for value over time. In other words, we're not on the other side of that. We're on the same side of that with our users and helping them to generate this step function. They need to generate more value because there are fewer of them doing more work, but they need to be paid for that step in valuation. So we'll be measuring things in different ways and incremental ways and so forth. But I know Nicholas and team have already been exploring as we're about to bring one AI native -- that's extreme term, but one AI-based product to market, as you know or might know. So Nicholas, perhaps, over to you on that.

Siti, thank you for asking that question and it's great to end this conversation with this question because it is so much about the future of infrastructure overall. AI comes up in every CEO conversation I have with engineering services firms. It's really top of mind. Because of this backdrop, we've been discussing quite a lot on so much demand yet so few engineers to do the work. And engineering service firms are absolutely leaning in to adopt AI. It starts with simply using AI for normal business tasks like I'm sure many of you here in the call are doing already. But it's also more and more leveraging AI for engineering tasks. It will start there by leveraging whenever they can capacities that are coming from software vendors. We talked about one very large global engineering services firm leveraging our own solution for road monitoring, which is Blyncsy for asset analytics. But of course, they're also very, very interested in leveraging the capacities that we are building for design. So coming to agentic AI now, yes? I think this is where we're going to see, indeed, a step function in productivity from those firms. Now we're talking about the core market. We're talking about the core engineering design work where, through AI, we can automate all the mundane tasks that are sucking time away from engineers to do what they really like to do, what they're really supposed to do, for example, automating drawing production, et cetera. So we announced early access to our site engineering application called OpenSite+ at YII last year, and then we'll talk about the progress we've made at YII this year. The engagement with accounts with this product is truly impressive. I think everybody sees the potential with this product itself, but everybody sees the potential beyond for other engineering disciplines, for other engineering applications. So while we are partnering that product to make it available for a more general availability, we are also working actively in expanding those capabilities to other engineering applications, right? And then last thing I will say is we see many engineering service firms who are building up their own teams with AI specialists, with data scientists. They're creating their own AI agents. And here, I think we are very well positioned to support because -- many of them and we would like all of them are leveraging Bentley Infrastructure Cloud where they bring all their engineering files that are coming from also sort of vendors, very often Bentley, sometimes not Bentley. Now our platform is completely open. It can ingest all sort of file formats, but the more important thing is when we ingest those files, we basically map the data in those files to a schema, so that we make that data available for AI. So we're beautifully positioned to help them tap into past designs in order to provide context for the AI agents to make recommendations going forward, yes. Now we're really very much at the beginning of this, but we are all very encouraged to see those developments. We're very excited about it. We'll discuss it more at our annual conference later this year. What I will say is that the -- and I will end there, the potential productivity gains for everyone involved in the ecosystem of infrastructure through AI are such that we're going to see value redistribution. We're going to see business models changing and potentially our own business model changing in order to support it.

Thanks. We have one more. Dylan popped on from William Blair.

Maybe, Nicholas, to that point in particular as well, too, it sounds like the idea of permitting reform kind of coming through incremental productivity and maybe conviction that you're seeing from customers with higher floors on kind of the E365 collars, if you will. Wonder how you think about kind of that evolution playing out from a pricing perspective as they are may be able to work through those backlogs a bit faster, draw down on that spend faster. Does that mean accelerated growth for Bentley? Maybe is it into the durability of that growth? Like how should we be thinking of the implications within the pricing structure?

I mean, so the confidence that we're seeing from our accounts, whether large or small, is what's really playing out at the moment of renewals and with our larger accounts. This is where we agreed together on both higher floors and higher ceilings. Of course, they come together in the low double-digit percentage, let's say, around 10%. So just a sign of the confidence, right? I think from a pricing standpoint, we always make sure that we capture our fair share of the value that our software provides. When it comes to our applications overall, we've been quite consistent with our price escalations in the mid-single digits. We were quite reasonable at the time of very high inflation not to go too high. And therefore, we are able to continue to increase our prices in the mid-single digits. Now when it comes to AI, we're going to see if the way we -- I'm not talking about the price point now, but the way we price, the metrics we use are adapted. For the most part, our applications are based on users. They are either subscription when it comes to SMB business or they are application-based usage when it comes to our enterprise accounts on E365. And of course, we have a number of applications targeted for asset operations that use different metrics like assets, number of assets, definitely the case of AssetWise historically and the case of the new applications we have for asset analytics that we talked about, road monitoring and cell towers. But now when it comes to AI for design, the feedback that we're getting for the first accounts that are testing OpenSite+ is that maybe, indeed, right now, it's a bridge too far to change completely the pricing metric. We still need a user component. But maybe application-based usage is not the most appropriate pricing metric because the value of the software cannot really be necessarily translated in the actual usage in number of days. And the productivity gains are so much that we may need something else. So with them, at least the agreement is we'll start with a subscription, term-based subscription. And then we will have an add-on or additional charges that kick in as soon as the usage of AI is extensive and goes way beyond what a normal engineer will be able to do in a given time. But that's going to be an evolution, right? So we're doing this not in isolation. We're doing this constantly discussing with our accounts, so what is appropriate going forward. We're seeing openness for alternative pricing metrics going forward, but this is going to take a little while. We're very much early stage.

I will make sure that we don't fail to do a commercial for our Year In Infrastructure conference in Amsterdam in October. Many of you have attended and found it very worthwhile. Just to tease a bit this year, the Going Digital Awards, the finalists are all there and we'll be presenting their -- presenting their projects. We ask them each this year how they were using AI. And I was very impressed with the things that we didn't anticipate that we heard in terms of the ways in which they are using AI and they'll be talking about that in their presentations. And we announced the finalists earlier this week. And of note, I think I have this right, there's 9 from the U.S., 10 from Europe. And then that number put together, 19, I think it is or 18 from Asia. So it's just a picture of where in the world innovation is occurring. And I hope you'll try to join us there in Amsterdam. One last thing as we have time here about the -- beginning of your question was about permitting reform. And here in the U.S., there's not going to be apparently One Big Beautiful permitting reform bill, but it's everything that's going on is encouraging and enabling and accelerating infrastructure investment generally, but especially for the electric grid. And I'd like to say that data center isn't the main reason for that. Data center just being added at the margin to what is already an overtaxed electric grid for reasons of data center it's having to run at even greater capacity beyond design capacity with issues of resilience and so forth that we've seen elsewhere in the world. So those are what it feels like. And Nicholas mentioned in addition to Power Line Systems, Seequent seeing green shoots as well with new mining. We hope -- we think it's too soon to call a change there, but we may be reporting further about that at Year in Infrastructure as well, I hope, and I hope to see you there. I'm not sure I'll be there in favor of Nicholas, but it will be well worth hearing more about the subjects we've been talking about today in October in Amsterdam.

Thanks, Greg. That concludes our call today. We thank you for your interest and time in Bentley Systems. Please reach out to Investor Relations with further questions and follow-ups, and we look forward to updating you on our performance in the coming quarters. Thanks.