All right, awesome. Well, thanks everybody for being here today. Welcome back from lunch, day one of the Citi's Global TMT Conference. I'm Steven Enders, part of the software research team here at Citi. And with us for this session, we have Serge from Appian. Serge, thank you so much for being here.
Steve, thanks for having us.
Maybe for those who might be a little bit newer to Appian, can you just give a little background on the company and what is it that Appian does?
Excellent. There's no better place to start. So Appian is a platform that allows customers to automate and orchestrate their most important enterprise processes. So that's a mouthful. So maybe the best way to explain that is through a few examples. So for example, one of our largest customers is a global pharmaceutical company that applies us as an enterprise standard workflow tool to manage processes such as clinical trials, logistics, deployed to 50,000 of their employees. Or a branch of the U.S. military is in the process of consolidating a number of their legacy systems that do finance, procurement, supply chain management onto an Appian platform. It will eventually be deployed to hundreds and thousands of users and will save the government tens of millions of dollars.
Or one of the largest asset managers in the world, I've actually seen them here at the conference earlier today, uses Appian to onboard and manage their customer relationships. And they replaced a number of manual processes when they deployed our platform. Or just to give you a little bit more color or context, a medical device manufacturer uses Appian to automate the lifecycle of a device from order all the way to installation. We replaced an internal homegrown tool that wasn't scaling. And then as a final example, and to add a little bit of AI to the equation, a mortgage lender is using Appian to audit tens of thousands of applications annually. And they use our agents to cross-check information across multiple different documents with 98% accuracy and then loop in the human as needed to facilitate the process, save time and money.
So if you kind of take a step back from all of that, what you hear is automation. What you hear is orchestrating end-to-end processes, replacing manual effort, replacing homegrown solutions, point solutions, or legacy applications in the spirit of providing better agility, performance, flexibility. That's ultimately the value that we bring to our customers.
Yeah, perfect. No, I think that's a great place to start here. You recently came over to Appian from MongoDB. I guess, what led you to Appian? Why did you decide to move to, yeah, to move over to Appian?
Yeah, it actually begins right where we ended the last question. It begins with the product. The product is very good. You see it in our retention rates. You hear it when you talk to our customers. Our customers see us as a mission-critical platform that they concentrate their spending on, and they like partnering with us. They like our professional services team. Again, the product is very, very good, so if you're going to have a foundational asset on which to build a company, the quality of product is as good as it gets, so I was very happy to see that and find that out in my research, and that was the first thing that attracted me.
The second thing I would say is that our AI value proposition really resonates in the marketplace right now, and I think is what the enterprises need, which is a way to deploy AI that generates tangible value, but very importantly, subject to their own accuracy, security, and auditability requirements, and there's a lot of disconnect in terms of the promise of AI and what it can actually deliver and what the customers are comfortable using AI for these days, and Appian fills that gap today in a way that we believe will resonate even more in the market going forward. Three is the ability to improve execution across the board, but particularly with our move-up market on the sales side. We've seen evidence of continued improvements under the new sales leadership, and we think in the early innings of what can be done there.
So as you think about ways to generate value and sustain sustainable, efficient growth, sales execution is a critical component. And then the final thing is the culture. It's a company that wants to win. It's a company that is intense and wants to be successful in this large and growing market. And I'm very excited to be a part of it.
Yeah, no, that's great to hear. I do want to dig in a little bit more on the go-to-market. But before jumping into that, I think you've been on the job now for three or four months, I want to say.
Three and a half.
Okay, right in the middle there. Where do you see kind of the most opportunity to help kind of operationalize the business and see areas where you can find the most efficiencies and just help Appian become a better overall company?
Yeah, so let's talk about opportunities for growth and then maybe opportunities for efficiencies, and there's overlap, but there's also some differences. On the growth side, we're still in the early innings of our go-to-market, move-up market, and sort of continued improvement, efficiency, and effectiveness there. So I think in terms of hiring the team, we're upgrading our talent. We're making progress, improving all our processes anywhere from enablement of salespeople to forecasting to actual negotiation and closing of the deal. There's opportunity for growth and improvement across the board, and the way that you will see it manifest itself is ultimately in sales productivity, efficiency of sales and marketing spend, and we've made progress on that front, but we think that there's more to go. The second part when it comes to growth is going out there and telling our AI message better.
As I said, we feel like there's a great desire by enterprises to hear how others are using AI in actual production use cases, and we have plenty of examples that we can bring to the market and give customers an idea of what's possible with AI today, which will generate more business for us, so that's the next opportunity, and then on the growth side, we're increasingly feeling like the changes that are happening in the federal government are actually an opportunity for Appian. The drive to efficiency with DOGE has been doing. Yes, it's introduced incremental uncertainty into the market over the last few months here, and obviously, we're in the midst of the biggest federal quarter, so we'll have to see how that plays out.
But some of the changes that were made in terms of how the government wants to do business and what that means for our ability to work with the government is actually increasingly seeing like a net positive for us in the long term. And then on the efficiency side is with increased sales productivity comes improved sales and marketing leverage. So as we continue, we've seen improvements in productivity over the past 12-18 months. And we can continue seeing that will drive the efficiency of our marketing spend. It will increase the payback on our sales and marketing dollar, allowing us to invest more while expanding margins.
Second is tasting our own medicine, applying AI across various processes, whether that is customer-facing functions, whether that is, frankly, generating our own code, where we're seeing some early success, as well as all the back office function across the company. We can use Appian technology to produce incremental efficiencies. And then finally is just being smart where we hire globally and continuing to get benefits from a global and distributed workforce to drive better operating leverage. So all those add up to growth opportunities on the revenue side and opportunities to expand margins and balance revenue growth and margin expansion as we go forward.
I guess looking specifically at the efficiencies that you have seen in the go-to-market, I guess what was it that you've changed to be able to drive that? And where do you kind of see incremental opportunities to, I guess, kind of further enhance that approach and drive further efficiencies and leverage from here?
Yeah, I would say that there's been two specific levers. So if you go back 12-18 months when we first started talking about a move-up market, we had a clear effect of focus. So we've, frankly, shrunk our sales force. We've eliminated the least productive part of it. And we've improved sort of mathematically, we've improved productivity because we've decided to focus where we're seeing the most success. And so that was a difficult decision to do for any company. It was difficult for Appian to do at the time, but it was the right decision to do, not just for the purposes of improving our financial performance, but also for the purposes of driving focus inside the company and what we do day in, day out. So that's behind us.
And that was sort of a one-time change that was needed and improved efficiency, but also improved our sort of scope and focus internally. And then what we've done since then is really just build foundations for better scaling of our sales orgs in the future in terms of some of the processes that we talked about in the past. And whether that is hiring new sales leadership, whether that is actually replacing feet on the street as well, whether it is forecasting, qualifying deals, negotiating. And what you see happening is you're seeing that we're getting bigger deals, even bigger lands with new logos. And what we've seen, what you generally see with enterprise sales forces is the momentum begets momentum.
When you see success under your own success or your colleagues' success, when you see pipeline growing, it further increases sort of the confidence that we have to go fight for the value that we believe we deserve and that we get out there in the market.
That makes sense. Maybe we can start to pivot a little bit into the AI discussion. I know that you have started to drive a little bit of AI revenue, but I still think we hear concerns in the market from investors around how does Appian monetize it? Is there risk to the revenue stream from AI, whether it's a seed-based replacement, or maybe there's different ways to build applications? Just how do you think about the, I guess, risk versus opportunity for Appian? How do you kind of capture the next use case within a customer? And I guess as you think about that customer, a lot of ways to build an application now, where does it make sense for Appian? Where is there maybe more of a code-based solution? Just how do you think about that?
Yeah, so maybe we'll talk about AI at sort of a high level, and then we can drill into three different buckets of opportunity for us because we think this heavily skews in the favor of upside for us as opposed to risks, and I do think there's quite a bit of misconception about this in the market, honestly, not just about us, but software more generally, but so as you think about some of the use cases that I've talked about in the beginning and the processes that we help automate and orchestrate, AI supercharges the value that the customer gets from that process when applied properly.
And so whether that is further replacing manual steps in the process, whether that is automating processes that weren't automatable before, whether that is adding an incremental step that accelerates or simplifies the process, AI has many ways of adding value inside the process, but critically inside the process. And because that's where we play, we think that the opportunity to us increases in three different ways. The first one is it increases the number of processes that customers are interested in automating, how they're willing to automate, and the value that they believe they can get from that process. And because we ultimately, no matter what the mechanism it is, we ultimately try to price to value, that means that the value that we ought to be able to extract from those applications is growing for us as well. So we share proportionally in the customer's gain.
That's in terms of the types of use cases that are going to be built, some of them that we're already seeing being built in the ROI that the customers are getting on it. It's true for new and existing customers as well, which is very exciting for us. Once customers are ready to deploy AI in production, they find Appian quickly because of our approach and philosophy to how to deploy AI in a process in a way that is, frankly, sympathetic with the constraints that customers want to put on the AI technology. So that's the first bucket. The second bucket is AI, and we talked about this a little bit on the last call, is AI offers the promise of accelerating modernization of legacy tech stacks.
Because with the combination of services and AI, it becomes easier to extract application logic from existing, in some cases, very old applications and deploy it onto some new stack, or in our case, Appian. And we have an incremental benefit, incremental advantage in that process as well because what customers want to do is not just take old apps and build them into some new set of code or a new stack, but they also want to reimagine their processes. And that's where our process expertise comes into play. And how do you think about if you remove the constraints from your legacy architecture, how would you actually build this process from scratch? And the art of the possible there is significant. So that's further incremental use cases.
And then finally, as building applications become easier, more applications will be built, and our platform will benefit from that as well, which I want to touch base to this idea that now there's many ways to build an application. And some of the fun buzzwords out there, like vibe coding and so forth, I just want to be careful because although those tools improve productivity of whoever is working in it, it's difficult to see it actually replacing a platform onto which to build an application because that platform requires an ability to communicate with the business user, an ability to visualize the process, an ability to then introduce security, auditability, have the real performance and scalability.
So, although bits and pieces do become better over time, and I'm sure that over time becomes the AI benefit will occur in multiple different ways, the complexity of the system and the interconnectedness of the data and the controls and the guardrails means that the platform is as needed as it ever has been.
So you're saying the platform approach, the auditability, security, governance, all that matters more, and that's not going to be replaced by.
I would say it's not replaceable. Matters more is semantics, but it's not particularly helpful. This is where our sort of engine versus the car analogy is helpful. So you've heard us talk about this in the last couple of calls. So AI is a new engine to put into an automation/orchestration process, but it still needs the process itself, and that's the engine versus the car analogy. And so all those pieces are exceptionally important. And if you talk to enterprise customers, they will all tell you the same thing to actually get the value at an acceptable risk from an AI strategy, and we don't see that changing.
Okay. All right. That makes sense. You have been working on your AI products at the solution set. And I think coming out of your conference, it seemed like the use cases were pretty compelling, a lot of new innovation there. How are customers at this point, how are they thinking about where does it make sense to build, where does it make sense to buy, and how does the platform decision-making process maybe change from AI?
Yeah. So what we overwhelmingly hear from customers is they don't have the expertise to build. And proof of concept is one thing, but putting something in production and putting your job on the line is a different thing. And so in the build versus buy conversation, certainly with all the customers that we have exposure to, and we have a pretty solid exposure sort of across the Global 2000, the answer seems to be buy. So that's point number one. Point number two is when it comes to buy and buy what, customers are looking for use cases where they can get tangible value at an acceptable sort of trade-off between risks and performance. And what that means is they still want systems that are 98%-99% accurate, where there will be a human in the loop if needed, where the process is auditable.
Because ultimately, when you take a step back, what you're trying to do with AI is you're taking a non-deterministic or a probabilistic technology and trying to shove it into a deterministic process to derive value without causing things to go haywire. And so what we see is partnering with us to engage our professional services because they need help implementing these solutions. And most of what we're seeing right now are documents-based use cases. So whether it's document summarization, extraction, comparing data, and those can be significant value in the context of the types of workflows that historically were deployed in Appian. So customers are excited to go down that path, but many of them are not ready, and those who are ready need significant help.
That makes sense. I mean, if I think about some of the use cases that maybe you're going after, I think there have maybe been some traditional either OCR technology or some of those kind of more document-based solutions. It seems like that's kind of where that's centered at. Just how do you think about, I guess, net new TAM opportunity that you're kind of going after and it kind of opens up the aperture for dollars versus maybe there is some kind of displacement kind of going on in that market?
I think it's both. I think it's both, and it's significant. So obviously, it's net new to us. So from some perspective, it's semantics. But ultimately, we have great opportunity to gain share with our existing customers. A number of our customers, certainly the government is this way, but they're not the only ones, are looking to consolidate on a fewer platforms, get out of as many legacy technologies as they can, and modernize on a platform like Appian. So that's opportunity for us to gain share in existing customers. And obviously, if you know us, chances are you'll like us. And if you like us, chances are you'll give us more business. And that's the vector in terms of acquiring incremental use cases inside of existing customers. But what we're also seeing is that our penetration is low when it comes to new logos, even in the global 2000s.
Some of the, frankly, one of the things that I've been positively surprised is the type of new logos that we're winning since I've joined Appian and the initial size of those deals. AI is actually a big part of their role because if you have an AI use case that you want to bring forth in production, you are quite likely to come and find us, which is why 50% of our new logos over the last couple of quarters actually came at the tier that includes AI features.
Sure. No, that's great to hear. We're about halfway here. If there's any questions in the room, I want to make sure that we can get to those and address those. But I guess continuing on kind of the AI discussion, I guess, how does the Appian AI strategy evolve from here? Where do you kind of see the most opportunities to build that out further within the product set?
So I will call out two particular vectors of further innovation that are exciting. First is our AI Agent Studio, which is in beta and will be GA later this year, early next year. And that's a set of technologies that will allow customers to build their own AI agents in a way that is congruent with our overall platform and approach to process. And some of the early beta use cases are exciting and show an incremental variety of use cases that we can go after. And so obviously, as we mature that platform and that product, the number of use cases that people will feel comfortable deploying AI for will only grow over time. So think of that as ways to win incremental applications either from new or existing customers to go onto the platform.
The second one is our Composer product, which is also in beta, which will allow for a more seamless communication between the business user of an application and the actual person building it, whether that person is Appian Professional Services, customer resource, or a partner. That conversation is ultimately where the value is happening. It's facilitating, making it easier to get the requirements, making it easier for the business user to explain what they want and the art of the possible to the person building the application. Using AI to make that process more seamless will only further reduce barriers to adoption to Appian everywhere. I see us investing in both of those over time, and that sort of expands both the number of use cases and how quickly they can get on the platform if you think about those two as the axes.
And then the final thing, which we're in the early days, but it's what is the product that we can build behind the concept of modernization? This will always be a solution that involves both services and software, but is there software that we can build in addition to the Composer to extract business logic faster from applications to facilitate our services and partner services to accelerate the process of modernization with our various customers?
That makes sense. I know you've given some disclosures around, I think it's the advanced tier in terms of AI contribution. How do you think about AI disclosure? How quickly maybe some of these capabilities are going to begin to impact the model? What does that look like on a go-forward basis?
Yeah. So maybe we talk about the enabling factors, and then we'll talk about the metrics first. So what we hear from the field is that the principal gating factor in terms of further AI adoption is actually customers being willing to bring use cases forward because of where they actually are in terms of their own comfort with their technology. And I think that's a bit of a misperception versus when you hear at conferences like these. So customers have deployed the productivity tools, so whether that's Copilot or Gemini or so forth. So that gives an up, that's a way to turn around and show their board, "Yeah, here, we are using AI. We are doing broad-based productivity improvements through these tools." They run a bunch of POCs, most of which have failed.
And then there's some subset of them that have succeeded, and now it's time to move to production. But the gating factor is those that are ready to be production moved. And that's maybe surprising given that we've all been talking about AI and nothing but AI for the last two years, but that's the reality of a cycle of enterprise adoption where we are with it. And so once the customer is ready, then they will show to us either as a new or an existing customer, and they will want to get our advanced tier features. And so what we've seen, and I mentioned already here today, is that over the last couple of quarters, 50% of our new logos is coming from customers who want the advanced tier, which likely means they have a production-ready AI use case that they're ready to roll out.
It's not the only feature, but it's the most prominent feature in the advanced tier. So we're happy to see that people are basically walking off the street and pick us as their AI solution. And then the same sort of gating factor exists with our existing customer base, which is, "Do you have a use case?" And when there is a use case, getting the upgrade is not difficult because the ROI is there to justify it in terms of savings that they generate in the process. The question really just becomes, "Are you ready? Are you as a customer ready to roll out?" And so they happen. And so the metrics are the output of customers' readiness to engage with AI as opposed to our ability to drive and incentivize through price or other means the adoption.
Because ultimately, once you have a compelling AI use case and a way to safely and securely deploy AI inside of a new or existing process, our 25% uplift is not an issue. The ROI is there. The issue is just like, "Are you ready to actually go through that exercise?"
Sure. I guess when you think about either the customer base or when you're finding new opportunities, just maybe where are we in that journey in terms of customers actually ready to deploy AI applications? Are we still kind of in the early stages of that?
Maybe not first inning, but second or third.
Okay. What does it take for that to begin to inflect for us to kind of turn the corner and kind of see a little bit more mass-market adoption?
I think it's the linear continuation of all the things that have been happening for the last couple of years. So the technology will continue getting better. By technology, I mean AI itself, but also the guardrails that people like us put around it. The implementation experience is growing. This is particularly where we think our professional services business gives us an edge because we have a very strong professional services organization that customers can pay premium for when they need to implement the most important and the trickiest of sort of Appian applications, which usually means AI. So we're building an edge on that side as well. Third is customer internal acceptance and tolerance for the risk. And ultimately, there's a little bit of FOMO needs to be happening. Customers need to see what other customers are doing. That's what we're trying to do.
We're trying to put together a marketing effort that demonstrates to customers what others have been able to do by naming customers, by putting people on billboards, by showing, "This is what I've used Appian AI, and this is the actual tangible value that I generated," which will give people sort of incremental ambition to go after those use cases themselves. And AI is growing in adoption. We talked about it being a contributor to our strength in the first half of the year. And we expect that to continue, whether there's an inflection point or continued steady build-up. We're okay with either outcome as long as we sort of keep getting more than our fair share, which we believe we are right now.
Sure. Last question on the AI side, and then we'll shift gears. But I think one of the things that is more product-related, I think one of the things that we've heard historically about Appian is you're selling a platform, right? And trying to find the right use case or really kind of narrowing down the message to, "We can help you solve for X," I think has maybe been a question from customers or partners that we've heard. So does this help simplify the messaging, or how do you think about what the real kind of killer use case application is that Appian can go after and target here?
I understand sort of the yin and the yang, if you will, right? Because if you have a solution that is sort of a hammer for a nail, that's perhaps an easier sort of incremental conversation or the first person that you find or the first deal that you qualify versus a platform which has a number of use cases, but then where do you begin? I would argue that that hasn't been a principal impediment to our growth, and the principal impediment to our growth has been more around the execution side, and if anything, the promise of AI and the variety of sort of agentic use cases that is available today, but it will become more available over time, only speaks to the value of the platform. Because you don't want to create a siloed AI estate the way that you created a traditional siloed IT estate.
What we're hearing more from customers is a desire to actually pick a platform, to pick a company to partner with, one that can bring breadth of functionality as well as professional services expertise. Those things, again, favor us and favor the breadth of what we have to offer versus, "No, no, no, here's my hammer. Let's go find all the nails to nail down.
Okay. That makes sense. I'm going to pause there and see if there's any questions in the audience. I'm going to go more into the financial side of the house. Okay. We have a mic come in just real quick.
In terms of the importance of metrics, how important is NRR to you? And then do you see a scenario where we see some acceleration over the next year or so?
NRR is an important metric because it demonstrates your ability to grow inside of existing customers. I have two caveats with our own NRR metric. The first one is that we have chosen to do it on a pretty lagged basis in terms of, and it's blended over multiple time periods in a desire to not, what's the word I'm looking for? Not overextrapolate any single move in any single quarter. We're sort of giving you a number that is more of a historical average of performance as opposed to the latest and greatest point in the quarter. And that causes some challenges because it's hard to reconcile at times with our revenue growth rate. But nonetheless, NRR and growth in existing customers as a concept is very important.
It's not a metric that we drive the business towards, but it is a metric that we care about as an output metric, and obviously, it's a metric that we're going to keep reporting. We don't guide towards it. We used to talk about this 110%-120% range, which I decided to stop that practice mostly because we don't actually drive the business. So why would we communicate a range for something that we don't actually incentivize anybody to hit? But we expect to continue to see strong expansion from existing customers, but we also have been happy with how the new logos have been performing over the last couple of quarters.
Okay. That's great. I want to ask about the DOGE side, the federal government, just kind of the deal environment. But I guess, what are you seeing out there? What are the kind of conversations like within the federal side of the house right now?
Yeah. So why don't we divide that into the right now and maybe the future?
Sure.
Right now or this year, we've seen disruption at the beginning of the year when the new administration came and DOGE was implemented or established. There was some period of uncertainty that was measured in weeks or a couple of months maybe because it wasn't quite clear what was going on, who still had a job, who's reporting to whom, that whole thing. That's largely subsided. And it feels like we're operating in the business's usual environment right now when it comes to the federal space. We've been happy with our execution in the first half of the year. We talked about our federal business actually growing faster than our overall business, which you wouldn't guess reading the headlines. And I think it speaks to the depth of the relationships we have and the quality of sales execution on that side of the house.
We are in the biggest quarter. Our quarters, like most people's quarters, are back-end loaded. So there's a tremendous amount of back-and-forth in terms of renewing and upgrading existing deals, new business that is out there in the federal government for us to win. Just given everything that's happened, although it does feel very much like business as usual right now, we'll have to see how the quarter calls out before we can actually say it was business as usual or no, something changed at the very end and that final approval that we thought we had, we didn't get or anything like that. But we have just short of a month to go. But right now, it feels like a normal federal quarter, hopefully closes out that way.
What I will say is if you take a step back, the drive for government efficiency has introduced two major positives for us that we believe will be a tailwind to our business regardless of how this particular quarter or year plays out in the federal. The first one is the desire to automate and consolidate legacy technologies. So there's a rallying cry in different pockets of the federal government, "Automate or die." And it is about processes that have existed for years and decades running on very old legacy systems, sometimes custom-made that the government wants to get out of the business of maintaining. And they're looking for platforms onto which they can consolidate that. And we have a track record of doing that.
I was talking about the military contract that we're consolidating a number of legacy platforms, where there's plenty of other opportunities in the fullness of time out there for us to compete, and we like our chances of winning. So if the government is really going to go through a monumental digital transformation effort, we like ourselves as being in the winner's column of that. And then the second thing is government has made it clear that it wants to deal directly with the software vendors as opposed to through intermediaries. And the reason is simple. The incentives are aligned. Our incentive is to actually get software deployed so that it can generate software revenue for us and savings for the customer, in this case, government, and professional services as a means to the end.
So to the extent that we get to talk directly to the government, we have two benefits. One is we can better explain the benefits of our product than anybody else. So it gives us a chance to sell directly as opposed to the intermediary. And then secondly is that we get to participate in the execution and the implementation, which will be incrementally benefit to how the product is implemented and therefore the value that it creates. So it really does feel like beginnings of a virtuous cycle, but nonetheless, the current spending environment is still to be determined. And obviously, we'll know more here in the next 30 days.
Okay. I guess when you thought about the guidance philosophy for 3Q, I guess in particular, did you take a bit more of a conservative approach given this uncertainty? And I guess secondly, I think we've heard from some of our colleagues who knew you from Mongo who maybe saw you as a pretty conservative guider in general. But I guess, is that how you think about guidance here at Appian, or is there something maybe different about the Mongo experience that is separate?
Yeah. So first, in terms of how we make the guidance sausage at Appian, in a short to medium term, given our deal cycles, it's pretty straightforward. You have a pipeline. You have stages of that pipeline. You have conversion rates. You have historical conversion rates in the sales forecast. You triangulate those, and you create the forecast. Nothing's really changed in that process, and nothing really needs to change based on the win rates and the conversions that we've seen in the first half of the year recruiting in the federal business. So to the extent that there's incremental conservatism, we didn't feel the need to introduce it into the guidance model. As for my MongoDB experience, the only thing I would say is conservatism always looks easy to call in the 20/20 and in the hindsight.
But I don't think that we are introducing an incremental level of conservatism here, either because of my newness or because of my nature when it comes to the Appian guidance.
Okay. That makes sense. We're in the final minute here or so, but last question for you. Just I think we've always got the question around balancing growth and profitability for Appian. And just how do you think about that moving forward? How do you think about top line versus margin?
I think it's exceptionally important to deliver both. I think we've done a remarkable, and by the way, this all predates me, so I don't get to take any credit for it, but we've done a remarkable job turning our focus to profitability over the last two years. So we're guiding to 7%-8% EBITDA margin for this year, which is roughly 2,000 basis points improvement over just two years. So that speaks to sort of when Appian decides to focus on something, we get it done, and the company should be proud of that. As we think about going forward, though, we do want to continue showing margin improvements because a business with great unit economics like ours should be able to do that. But we don't want to shortchange the growth because we do see opportunity to grow our sales org because of better execution.
We do see incremental AI adoption, so a need or an opportunity to invest on the product side as well as benefit from the go-to-market. We talked about DOGE being a long-term tailwind and sort of the opportunities that may come from that, so we want to find the balance, but it is important to continue showing measured margin expansion in addition to revenue growth.
Okay. That's great to hear. I think we're at time here. So Serge, I want to thank you so much for being here. I want to thank everybody in the room for attending today as well. So thank you so much.
Thank you.