Good day, and thank you for standing by. Welcome to the Appian third quarter 2026 earnings conference call. At this time, all participants are in a listen-only mode. Please be advised that today's conference is being recorded. After the speaker's presentation, there will be a question-and-answer session. To ask a question, please press star one one on your telephone and wait for your name to be announced. To withdraw your question, please press star one one again. I would now like to hand the conference over to your speaker today, Brian Denyeau from ICR.
Thank you. Good morning, and thank you for joining us. Today, we will review Appian's third quarter 2026 financial results. With me are Matt Calkins, Chairman and Chief Executive Officer, and Serge Tanjga, Chief Financial Officer. After prepared remarks, we will open the call for questions. During this call, we may make statements related to our business that are considered forward-looking. These include comments related to our financial results, trends and guidance for the fourth quarter and full year 2026, the benefits of our platform, industry and market trends, our go-to-market and growth strategy, our market opportunity and ability to expand our leadership position, our ability to maintain and upsell existing customers, and our ability to acquire new customers. These statements reflect our views only as of today and do not represent our views as of any subsequent date.
We will not update these statements as a result of new information unless required by law. Actual results may differ materially from expectations due to the risks and uncertainties described in our SEC filings. Additionally, non-GAAP financial measures will be discussed on this conference call. Reconciliations of GAAP to non-GAAP financial measures are provided in our earnings release. With that, I would like to turn the call over to our CEO, Matt Calkins. Matt?
Thanks, Brian. Appian had an excellent Q3. Our cloud subscriptions revenue grew 20% to $136 million. Our cloud net ARR expansion rate was 120%, plus 5 percentage points above last quarter. Subscriptions revenue grew 28% to $188 million, plus 9 percentage points above last quarter. Total revenue grew 25% to $235 million, plus 6 percentage points up, and adjusted EBITDA was $43 million. If you have been watching Appian, you know how we positioned ourselves to win in the AI era. You have seen this growth coming, and you understand why it is happening. For those new to our story, I will take a few minutes to explain. This month, the LLM revolution is four years old. It is transformative, but also non-typical. Unlike every other tech revolution I can think of, it did not start with the highest value use cases.
As you know, the first people who bought mobile devices were the ones who most needed mobile devices. When they first lit a street with electric light, it was a big street. But AI is different. It is not used for the biggest work. It is used for personal productivity, but not the mission-critical core work of the largest organizations. Repeated studies have shown this. If anyone still argues today against the profound effects of the AI revolution, this is their best argument, that it is not moving the needle on critical work. So the AI revolution exists, but it is haunted. This is Halloween season after all. The AI revolution is haunted by a question: How can AI be used on the most valuable work? That is where Appian comes in. Appian provides the specific capabilities AI requires to work in high-performance situations.
Appian connects the world's most important technology to the world's most important work. Here are the capabilities I am talking about. Bear with me, please. It is a simple list. First, AI is probabilistic and needs a deterministic orchestration layer. Our orchestration layer is ranked number one in the world. Second, AI is only as smart as the data you give it, especially agents that want to roam the enterprise for information.
Our Data Fabric is the ideal connectivity layer, providing access to distant data sources that is lightweight, performant, and secure. Third, AI needs governance, by which I mean tracking, validation, escalations, and guardrails. Governance means safer operation and continuous improvement. Fourth and last, tokens are expensive. Appian can allocate jobs to whichever LLM or digital worker can do them most efficiently. This saves a lot of money versus sending everything to a frontier model, and it also prevents lock-in.
One, two, three, four, that is my quick overview. With these capabilities, AI can participate in the most important work. We make AI reliable enough and efficient and more powerful. The revolution will not be complete until we can deploy AI into the highest value processes, and that is what Appian is doing for our customers. Our customers are responding with great enthusiasm. Their Appian AI usage is up more than 50% over the prior quarter.
We succeed by bringing AI deeply into the enterprise at scale. For example, our customers have deployed Appian AI to conduct audits three times faster, perform underwriting reviews 75% faster, to run investment processes with 34% greater efficiency. Appian provides the most value to the high end of the market, and especially highly regulated industries. As you know, 80% of our ARR comes from our key verticals, financial services, life sciences, and the public sector.
Our platform is used by leading institutions, including two-thirds of the world's largest pharmas, insurers, and non-Chinese banks, plus 20 major governments around the world. In the U.S., Appian is used by all 15 cabinet-level federal agencies and all military branches. These top organizations have one thing in common: they cannot afford to make mistakes. Many are under mandate to adopt AI, and they need the reliability that Appian provides.
I will share two recent examples. First, a large European bank manages its internal audits on Appian. This quarter, it upgraded its licenses to embed our AI into its workflow. Appian agents will identify compliance risks and recommend actions. Downstream, agents will analyze the business to ensure controls are properly implemented for regulatory compliance. The bank chose our AI for our Data Fabric, which will supply agents with the necessary data to reduce review times from 10 days to 30 minutes.
Second, a North American insurer and longtime Appian customer deployed DocCenter, our document processing AI, into its claims appeal operations earlier this year. Our AI automatically classifies and extracts data from 100,000 files annually. It chose to deploy DocCenter to more areas of its business and increased its Appian spend by 86%. This quarter, Gartner named Appian a leader in its recent Magic Quadrant for business orchestration and automation technologies, BOAT for short. The report praises our ability to orchestrate AI alongside humans and digital workers. Gartner ranked us the number 1 vendor for deterministic workflow automation, and also for platform consolidation. Having spoken already about deterministic automation, I'd like to spend a few minutes now on platform consolidation. There is more interest today in writing new apps with AI than I have ever seen. The reasons are simple.
AI is exceptionally good at writing applications quickly and accurately, and post-Mythos, every old application is likely hackable, a security vulnerability. So, AI lowers the cost of creating a new app and raises the cost of not doing so. When you write an application, you have a choice, whether to build it on a platform like Appian or to write it from scratch. Homegrown applications have always been the top alternative to using Appian. In the era of AI, you still have that choice. Sometimes companies will choose to write an app on a platform, and other times they will write it in code. They should choose a platform when the benefits of the support exceed the cost of the platform. The support, of course, includes consultants and experts, tech support, upgrades, patches, and a corporation that stands behind the application.
Platforms also help unify applications across the enterprise and provide advanced capabilities consistently and out of the box. For many applications AI builds, a platform is worth the cost. Appian offers a compelling combination, the speed of AI development plus the reliability and power of a platform. AI is our preferred development method, and it saves a lot of time. It's even possible now to develop sophisticated Appian applications without ever opening an Appian interface. You can build directly through Claude, Codex, or others. The newly created app can leverage existing Appian objects and can be fine-tuned later in an Appian design interface. Once built, the AI-written app will stay modern through Appian upgrades. It will be secure through Appian certifications, and it will be supported by our community of experts and employees.
For example, a leading insurance provider and seven-figure ARR customer upgraded its Appian licenses this quarter to access our AI and expedite its multi-year modernization effort. The group will replace its incomplete systems with enhanced Appian apps. Our AI will automatically extract requirements from company artifacts, draft new design plans, execute the builds, and iterate on user feedback. Our platform will facilitate a dialogue between developer and AI to validate the apps are secure and reliable and designed as intended. The group chose Appian to modernize its enterprise three times faster than originally planned. This Q3, we exceeded our plans in all major regions, especially in the U.S. public sector. The federal government is more focused on efficiency than ever. It's more willing to contract directly with vendors. Appian is acting on this opportunity by connecting with critical mission areas. We're doing more prime and large deals than ever before.
For example, a U.S. cabinet-level agency and longtime customer expanded its use of Appian this quarter into a new mission area. The acquisitions group decided to automate its procurement process and signed a seven-figure license deal. It directly contracted with Appian to accelerate the purchase cycle. Appian will help facilitate the awarding of over $100 billion in annual budget. My final example is with a U.S. military branch that expanded its use of Appian with a seven-figure software deal. This armed service awarded Appian a prime contract to modernize its fleet management operations after a system integrator's multi-year implementation failed.
The branch will unify dozens of disjointed systems. Service members will schedule maintenance, manage repairs, and procure new parts. Our AI will expedite the process by recommending replacement parts and providing repair documentation. In closing, Appian has claimed a distinctive and valuable place in the AI stack. We connect the world's most important technology to the world's most important work. We provide an ideal platform for AI-written apps, and we are modernizing legacy applications at scale. This latest quarter, by far the strongest in Appian history, is the best validation yet of our essential role in the AI revolution. Over to you, Serge.
Thanks, Matt. I'll begin with a detailed review of our third quarter results and then finish with our outlook for the fourth quarter and full fiscal year 2026. Starting with our results, we had an excellent quarter of new business, driven by continued AI traction. We saw strength across all major regions and industry verticals. Appian exceeded the guidance ranges we provided on our key metrics of cloud revenue, total revenue, and adjusted EBITDA. Cloud subscription revenue was $136 million, an increase of 20% year-over-year. On a constant currency basis, cloud subscription revenue increased 21% year-over-year. Total subscription revenue was $188 million, an increase of 28% year-over-year. On a constant currency basis, total subscription revenue grew 29% year-over-year. In Q3, we saw very strong demand for our self-managed software.
Our customers, who, as you know, are in heavily regulated industries, continue to value deployment flexibility and data sovereignty that self-managed software provides. Providing customers that flexibility to deploy Appian in the cloud or on-prem is a strategic differentiator for us. Professional services revenue was $47 million, up 18% compared to the third quarter of 2025. Total revenue was $235 million, an increase of 25% year-over-year. On a constant currency basis, total revenue grew 26% year-over-year. Our cloud net ARR expansion was 120% in Q3, compared to 112% a year ago and 115% in the prior quarter. As a reminder, we present net ARR expansion on a constant currency basis. This is the strongest net ARR expansion in over five years, and it's another great indicator of new business momentum we had in Q3. Now let's turn to profitability.
I'll be discussing our results on a non-GAAP basis unless otherwise noted. Gross margin was 74%, flat year-over-year and up from 72% in the prior quarter. Our subscription gross margin was 86%, also flat year-over-year and up from 84% in the prior quarter. Professional services gross margin was 24%, compared to 31% in the year-ago period and in the prior quarter. Total operating expenses were $133 million, up from $108 million in the year-ago period. Adjusted EBITDA was $43 million, ahead of our guidance range between $30 million and $33 million, and compared to adjusted EBITDA of $32 million in the year-ago period. This outperformance relative to our guide was driven by greater-than-expected revenue. Net income was $37 million, or $0.51 per diluted share, compared to $24 million, or $0.32 per diluted share for the third quarter of 2025.
This is based on 73 million diluted shares outstanding for the third quarter of 2026 and 75 million diluted shares outstanding for the third quarter of 2025. Our stock-based compensation expense was $13 million in Q3, or $0.17 per diluted share. In the third quarter, we purchased approximately half a million shares for $17 million, bringing our total buyback to $83 million under our current $100 million authorization. Turning to our balance sheet, as of September 30, 2026, cash and cash equivalents and investments were $89 million, compared to $187 million at the end of last year. For the third quarter, cash used by operations was $15 million, compared to cash provided by operations of $19 million for the same period last year. Our cash flow in the quarter was impacted by timing of collections, which can naturally vary quarter to quarter. Turning to guidance.
Based on our strong underlying business momentum and our success year to date, we're meaningfully increasing our outlook for the fourth quarter and the full year. Starting with the fourth quarter of 2026, cloud subscription revenue is expected to be between $143 million and $145 million, representing year-over-year growth of 23% at the midpoint of the range. Total revenue is expected to be between $233 million and $237 million, representing year-over-year growth of 16% at the midpoint. Adjusted EBITDA for the fourth quarter of 2026 is expected to be between $30 million and $33 million. Non-GAAP earnings per share is expected to be between $0.30 and $0.34. This assumes 73 million fully diluted weighted average shares outstanding. For the full year 2026, our cloud subscription revenue is expected to be between $536 million and $538 million, representing year-over-year growth of 23% at the midpoint of the range.
Total revenue is expected to be between $873 million and $877 million, representing year-over-year growth of 20% at the midpoint. Adjusted EBITDA is expected to range between $116 million and $119 million, for an approximately 13% margin and 53% year-over-year growth at the midpoint. Non-GAAP earnings per share is expected to be between $1.20 and $1.24, or approximately 100% growth at the midpoint. This assumes 74 million fully diluted weighted average shares outstanding. Our guidance assumes the following. First, we anticipate our non-cloud subscription revenue to be down slightly in Q4 and up in the high single digits for the full year. Second, we expect professional services revenue to grow in the mid-teens in Q4 and in the high teens for the full year. Third, our net interest expense will be approximately $3 million in Q4 and $10 million for the full year.
Fourth, our guidance assumes FX rates as of early November. Due to continued strengthening of the U.S. dollar since our last call, we now expect FX to represent a 2 to 3 percentage points headwind to our reported revenue growth in Q4 of 2026. Finally, we are now forecasting nearly 3 percentage points of EBITDA margin improvement in 2026, compared to 1 percentage point we expected at the start of the year. In closing, we see our Q3 results and excellent new business performance as clear indications of our AI value proposition. We are excited about the opportunity ahead and will continue to invest responsibly to maximize our long-term value. Now we'll turn the call over for questions. Operator?
Thank you. As a reminder, to ask a question, please press star 11 on your telephone and wait for your name to be announced. To withdraw your question, please press star 11 again. One moment for questions. Our first question comes from Steve Enders with Citi. You may proceed.
All right, great. Thanks for taking the questions this morning. Maybe just to start, I mean, it seems like there's a lot of good momentum out there. Maybe we can just talk a little bit about what you're seeing in the forward pipeline and maybe some of the, I guess, factors that are driving new opportunities here. Maybe as a piece of that, how much are you kind of seeing agent conversations coming into the forefront at this point, and are enterprises ready at this time to start on that agentic journey?
Yeah. Thanks for the question. First of all, demand is robust and continuing to be so. Secondly, I feel that the demand we're addressing is high-level demand and opens opportunities to have conversations at the top of organizations on a larger scale than we might have been able to achieve in the past. The agentic conversation is one that reaches the top of the organization. It's a priority. The effects are expected to be profound, and we can be part of that. There's definitely interest, and we're supplying it with regards to agentic deployment, provided we can make it reliable and performant, and we can. So, they typically want to take short steps, but as we build confidence, we go further. So absolutely, yes, that's happening.
I feel like in this environment, we have an advantage, like a head start, in the way that we've been thinking about the combination of AI and process for a long time. We've been shipping process models with AI embedded within for a decade or so, and over that time, we've gotten to see how they interact and how it was necessary to add functionality to AI in order to round out its capabilities, keep it in line, monitor it, improve it. We're familiar with these problems.
When the LLM revolution comes along, we've already had a lot of time to think about what it means to put AI into a process, to regulate it, to inform it, to secure it. We had this head start, and it's allowed us, I think, to see some of the problems in advance and to position ourselves for what the AI revolution would require before the revolution itself was really conscious of what it would require, if that answers your question.
Yeah, no, that's great to hear, and yeah, appreciate the other thoughts there. Maybe on the guidance, I mean, the Q4 growth for the continued acceleration on the cloud side looks pretty impressive, and I guess just trying to understand a little bit better how you're thinking about, I guess, the ability to sustain that acceleration as we start thinking about 2027 and how you're thinking about just the forward demand environment and how that maybe supports growth further in the future.
Thanks for the question, Steve. I guess I'll say three things. First, the Q4 guidance for mid-20s constant currency cloud growth is a function of a series of good quarters of new business, but particularly Q3, which, as we said, was excellent. The easiest way to see that is the 120% net ARR expansion for existing customers, which reflects the exit run rate for ARR for that business, and that sets us up for the guidance in Q4. It's obviously a function of our team doing a great job creating and capturing demand all year round, but especially when it comes to Q3. When it comes to next year, I guess I'll say first two things. First, we have a big quarter ahead of us. We have to get Q4 done. Q4 represents the starting point for next year.
So, on some level, it is a little premature to talk about how we think about quantitatively the demand for next year. What I will tell you, though, if I take a step back, what has 2026 taught us? It taught us two things. One is that the demand is there, that the enterprises understand that they need help with deploying AI at scale, that they need capabilities that Appian provides, and we have both the technology and the credibility to provide it. And then secondly, that we can capture that demand. Our go-to-market execution has been strong. Our productivity has been improving. So the combination of demand and ability to capture it gives us confidence, not so much for any particular numbering in 2027, but that we can build durable growth here at Appian.
We all show up here for work knowing that we should be a much bigger company in a few years, and that is how we are investing. That is why we are growing our sales org, that is why we are investing in R&D and everywhere else, and it is really that durability of growth that we are preparing ourselves for.
Okay. Great to hear, and thanks for taking the questions.
Thank you. Our next question comes from Sanjit Singh with Morgan Stanley. You may proceed.
Yeah, thank you for taking the question, and congrats on the strong quarter. I had a question just on the self-managed performance, the other subscription revenue, really saw a pretty material uptick in growth there. I just wanted to get a sense, Serge, to what extent is that structural? The self-managed demand, is there particular industries or verticals or customer profile types that is driving the self-managed revenue?
Yeah, there is. That's a great question, Sanjit. There is. First of all, the public sector has always leaned a little more towards self-managed than other sectors, and Q3 is naturally a quarter where the spotlight falls on the public sector. Secondly, this is the year of sovereignty in European business. We hear it frequently when we talk around the continent. It's one of their top priorities is independence from American technology. They're looking for ways that they can have more control than dependency. One of the ways they achieve that, and one of the ways that we uniquely provide that, because many of our competitors can't do this, is to allow them to run our software on their own premises under their own control.
It's a nice selling point, and it's also more in vogue than ever before, and it has led to an uptick in interest in consuming our software that way.
Yeah, and Sanjit, let me just jump in with an accounting reminder. Due to ASC 606, we need to recognize roughly 70% of those bookings upfront. Obviously, that explains generally the variability of this line, but also explains partially why the number was so strong and ahead of our guidance. However, there's nothing else sort of funky, if you will, in there. There's no multi-year pull-forwards. This is a, all of our on-prem contracts are one year in nature. So, it's a good representation of the demand that we saw, and then, of course, the accounting needs to be taken into account.
Yeah, no, I appreciate that color, Serge. My follow-up question, Matt, you sort of detailed the Appian AI strategy, and certainly the message seems to be resonating in terms of embedding AI with deterministic processes. I just want to get a sense, and you sort of mentioned the sovereign AI theme. How are open-weight models being adopted into the platform? Are you guys pursuing a post-training strategy given all the processes that you guys have a lot of familiarity with? What could be the potential financial from a growth perspective, and even from a margin perspective, if open-weight models become a more prominent role in the platform?
Yeah. Hey, I expect that over time, there's going to be less dependency on certain models and more willingness to take into account many models. I don't want to predict which ones they may be or whether they're open-weight, but I do think that we're going to head toward a spectrum of different cost models, because we're going to be seeking a more efficient way to get work done. Sometimes you'll send a job to the latest frontier model, sometimes you'll send it to a less expensive model, sometimes that job will go to a digital worker, even a human being. So, the cost and the efficiency of job resolution is going to be a critical driving factor to the adoption of new models, and we're here to facilitate that.
We believe that an organization should have multiple cost points to which they can refer work, and we are built to facilitate it, to save them money, to get the job done with an equal amount of accuracy, but not to initiate unnecessary power. I'm not at all surprised to see this trend continue. It's an appeal for us, and we believe that customers do not want lock-in. So, we're designing in advance to give them the key to avoid it.
Appreciate the thoughts, Matt. Thanks.
Thank you. Our next question comes from Raimo Lenschow with Barclays. You may proceed.
Hey, thank you. Two questions. One, a more accounting question, so I apologize for that. Serge, cash flow, if I look at the quarter, everything looked pretty amazing. Can you talk to the cash flow performance, though? It seems like timing, but maybe you can give us more detail.
Yeah. I mentioned that it's timing, but let me give you a little bit more color. Generally speaking, collections are seasonal. Q4 and Q1 tend to be stronger than Q2 and Q3, so that's kind of the first thing to keep in mind. Secondly, as we do more large strategic deals, as we do more multimillion-dollar deals, and with them come multimillion-dollar collections, where those land, even on either side of quarter end, can impact timing. In this particular case, October was one of our strongest collections months ever, and so it's just a timing issue, and obviously, we continue to focus on it, but we're generally happy with the trend.
Yeah. Okay, perfect. Matt, the really great performance across the board, it looks like it's all coming together. Can you speak to the role of the improved go-to-market? You guys have been working on this for a couple of years now, so it's industry positioning, and you spoke to that in detail. But also, can you speak where we are on the go-to-market function and how much that is kind of benefiting the good trend here? Thank you.
I think we're really connecting with our customers. I love that we're doing it at a higher level than we ever did before. We're speaking to the top of the organization about strategic things. Take for an example, the EGP structure, the enterprise growth plan structure of deals, which we've been favoring more than we ever have before, where we pre-negotiate a price on an annual basis that will stair-step upwards into the future, and then allow the customer to have an essentially unlimited amount of usage under that figure. Well, EGP plans are facilitated by our modernization story, because when we talk about replacing old applications, the customer can understand why the value of our software will be exceptionally high.
While they may have only had a couple of projects in mind that were new to create, they've got all these old applications that need updating. Suddenly the addressable market within one customer is far larger than it would have appeared before, and justifies a larger EGP. I think that we're approaching the market correctly today to get the most impact and the highest level of access and strategic consideration for the value that we're providing. We're getting into the right kinds of dialogues about things like big dollar commitments and unlimited amounts of usage and modernization of vast amounts of old applications. We're going to market the right way for our product and for this moment.
I just add that the discipline of our go-to-market org has been impressive. Productivity is improving. They are on top of their forecast. Deals are not slipping. We like our win rates. It's been really great to see come together, and kudos to the team.
Yep, I agree. Okay, thank you.
Thank you. Our next question comes from Devin Au with KeyBanc Capital Markets. You may proceed.
Good morning. Thanks for taking my questions. I know that you've seen really strong momentum in public sector, and it is still really early following the midterm elections, but maybe given Appian's sizable exposure there, how do perhaps changes in political environment typically impact customer priorities or spending patterns? As we look into 2027, is there anything you are watching that could meaningfully change the opportunity set in that sector?
Yeah. First of all, I want to say that the election is expected to have no impact at all on the way we do business. We are focused on fundamentals, not politics, always have been. We focus on efficiency, and the government cares about efficiency more than it ever did. We focus on solving problems with technology, and the government is more willing to see technology as a way to solve a problem and a proactive instrument of improvement than it ever has been. We sell to impulses like that. Yeah, we are not affected by the politics.
Understood. Thank you for that. Serge, sorry if I missed it, but did you say that EBITDA margin expansion next year to be at a greater rate than the one-point expansion entering 2026? Maybe just perhaps speak to kind of your confidence in driving further expansion, and what areas you are looking to invest as we look into next year. Thank you.
I actually didn't say anything about EBITDA margin for next year, but let me kind of zoom out for a second. We are committed, and we continue to be committed to balancing our revenue growth and margin expansion. That's true for this year and for the future. Look, we believe that that's a good thing to do, not just for external reasons, but also as a driver of internal discipline and efficiency and return on investment. As you think about the magnitude, obviously premature to talk about it, but the magnitude this year, the surprise to the upside, is really a function of doing so much better on the revenue side than we expected, and sort of the timeline of how you can reinvest those investments. We're really happy with how we did this year.
We are committed to continued margin expansion into next year, but we'll give you the actuals when we get together in February.
Got it. Sorry to interrupt, Serge. Thanks for the clarification there.
Thank you. Our next question comes from Pat McIlwee with William Blair. You may proceed.
Hi, Matt, Serge. Congrats on another really strong quarter here. Thanks for taking the questions. First, you previously talked about the opportunity to monetize some AI consumption beyond the usage included in your subscription tiers. As we think about some really great results here, I just wanted to ask if that usage-based monetization has begun to contribute to cloud growth or NRR, or if most of the AI-related uplift is still coming from customers upgrading to higher tiers?
It's a great question, Pat. Let me just kind of step back for a second to make sure everybody's on the same page. You're right. We have a two-step process to monetize AI. First is, in order for a customer to even have access to our AI features in production, they have to be at a certain tier of subscription. By far the most popular is our Advanced tier, which is a 25%-35% premium to our Standard tier. We continue to see, and we've talked to you guys about this over the last year, more customers and greater percentage of their ARR moving to the Advanced tier. As you said, the Advanced tier and other tiers include what I would describe as a moderate amount of AI usage, enough for maybe one medium-sized use case.
If you have multiple use cases, which we're starting to see, or very large use cases, like one that Matt has talked about in his prepared remarks, then you actually have to come and buy more AI usage from us. We see more and more customers being in that situation every month, where they exceed the allotted amount, if you will, and then they either pay overages or they negotiate with us to buy more AI usage bundles, as we call them. Consumption grows over time, so it will be building. I wouldn't call it material right now, but it's definitely an added sort of vector of growth in the future.
Okay, great. Thanks, Serge. Last quarter, you mentioned that legacy modernization ops were increasing, but were not yet a material contributor to growth. As we think about this acceleration in the business, has anything changed over the past few months? At this point, are you seeing more customers move beyond initial modernization projects towards broader multi-app transformation programs?
Modernization is absolutely driving some of the largest transactions we are doing today in the form of EGPs. We discuss with customers the vast amounts of applications that they are modernizing or will soon wish to modernize, and it justifies the largest enterprise transactions. It is part of the conversation. We just had our conference, our Appian Government conference last week here in Washington. A lot of enthusiasm, great energy, but the thing that struck me most about it is, in every presentation I watched and every conversation I had, modernization was the ever-present topic. Everybody is thinking about this. They are concerned about old applications not just being expensive, but now liabilities. They are excited about the ability to reinvent the enterprise. This topic has come on quickly in the past year or two, and we knew that it would. We were ready for it. It is now ubiquitous in major conversations.
Okay. Thank you both. Great results here.
Thank you. Our next question comes from Lucky Schreiner with D.A. Davidson. You may proceed.
Great. Congrats on the quarter. Really good results here. I wanted to ask on the fourth quarter profitability outlook, the implied margin looks a touch lighter than we were anticipating. Does that specifically reflect the increased sales hiring that you discussed last quarter ahead of demand, and how has the ramp of new salespeople trended so far? Thanks.
Yeah, let me come at it from a couple of different ways, Lucky. First, I would say, yes, definitely part of the equation is what we announced last quarter, which is a head start, if you will, to hiring, to growing our sales org so that more of our people are in seats, enabled, and ready to go in early 2027. So that's definitely a major investment part as you think about the back half of the year and as you think about it for Q4. Just maybe a couple of other points as a reminder. We have sequentially higher expenses in Q4 than we do in Q3. Matt just mentioned Appian Government. That's one of our largest events. Appian Europe is in a few weeks here in London. So it's a very busy event quarter, large and small events, internal and external, and that obviously drives expenses.
Also, by Q4, remember, this is the quarter where bonuses and commissions tend to be larger than they are in other quarters, so you got to think about it on a sequential basis. Finally, because of the revenue outperformance and the accelerating cloud growth in Q4, we took the opportunity to make a few smaller tactical investments that will further help us into 2027, and those sort of add up to the EBITDA expectations that we have. Also, I would just kind of zoom out and remind again that we're delivering 300 basis points of margin expansion, 53% EBITDA growth for the year. So, we're very happy with our ability to both grow the business and expand the bottom line.
Great. Makes a lot of sense. Then just with the retrial against Pega coming up, I wanted to ask, can you provide any additional color or reminders around how your $500 million insurance policy works that you purchased in relation to the case?
Yeah. We can talk tremendously about the insurance publicly. We're excited about the case and our ability to go to trial again and argue for our situation. What I will just say, the insurance is there for what it is, insurance, and it will play in at the end of the legal process after all the appeals have been exhausted and so forth. So, we're glad that we have it, and it's sort of out there in the future.
If you want any information on it, the best way to go is to look at our webpage.
Appreciate it. Thanks for taking my question.
Thank you. Our next question comes from Derrick Wood with TD Cowen. You may proceed.
Great. Thank you. This is Cole Erskine on for Derrick. Matt, I just wanted to come back to some of the conversation around open-weight models and model choice. Could you just talk about how, on the Appian platform, you guys go about model routing? Is this something that is the customer's choice? Do you guys automatically route based on the task and the requirements needed for that workflow? And then how you think that over time this could impact usage?
Yeah. The key is that it's completely the customer's choice. We offer them scores of models from which to choose from, and they can route the work entirely at their discretion. I do think it's going to help us get more adoption because it can contain AI bills. By making the AI experience one that's less daunting, more flexible, less expensive, I do think it will encourage adoption. At the same time, for us to make it less expensive and much more reliable, safer, able to improve, able to be better informed, I think we've given people the fast lane to AI facility. I think all what we're trying to do here is just help the customer with the things they're naturally concerned about, the trepidations they have, the restrictions they have, the red lines they have.
We're just going to help them across all of that in the name of making AI a useful participant in the world's most important work.
I would just add, Cole, that we noticed that the customers are becoming more sophisticated in the choice that they have. Whereas, a few months ago, it would just be kind of click to default across the board, now there's much more conversations about the nature of the model, and the specific need that's be used. We welcome those conversations because, to Matt's point, it will help adoption, it will help accuracy, it will contain costs, and we stand to benefit along with the customer.
Super helpful. Thanks. Then, Serge, just one on the net expansion rate. Impressive step up. Going forward, is this kind of a new level that we can think about this trending at, through the end of the year and maybe in the next?
We don't guide to this metric, so I will refrain from speculating as to where it can go. What I will say is, ultimately, it's one of the best ways to look at our quarter and realize just how good a new business quarter we had. Yes, it's unusual, and certainly in my career, to see this kind of jump from a quarter to a quarter. That speaks to continued building of the demand and continued improvements in execution to capture that demand. Again, as I said, we're obviously very happy with Q3. It comes after a series of otherwise good quarters, but this one was particularly good. It tells us that we can continue building a durable business and gives us confidence to continue investing, and that's what we're going to do.
Great. Appreciate it.
Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.