Appian Corporation (APPN) Earnings Call Transcript & Summary

All right. Well, thanks, everybody, for being here today. Steven Enders, part of the software research team here at Citi. With us for a final session of the day, we have the team from Appian, Mark Matheos, CFO; and Srinivas Anantha, on the IR team. So I want to thank you both for being here.

Great. Thanks, Steve. Thanks for having us.

Yes. Maybe just to start off, maybe talk a little bit about just Appian, what it is that you do and a little bit of background for those who may be a little bit newer to the story.

Yes. Appian, I think is focused on the process automation market. By that, what we mean is we do everything that's related to building a process, running a process, diagnosing a process, changing a process. So everything related to a process. We largely or predominantly sell to large enterprises that have complex processes or workflow needs. And these companies are trying to make themselves more efficient, provide better customer service, make their processes more agile and make themselves better. The one thing that's different about Appian versus others is we provide all the capabilities related to process automation, and that includes capabilities related to workflow, RPA, process mining, AI, all in one single integrated platform. We think this is good for the customer as they're dealing with one code base. That makes -- help them deploy applications much more faster, quicker and [Audio Gap]

The demand environment. And what you're seeing out there on the macro. I guess has there been -- how would you kind of characterize the impact that you have seen so far from customers? And how are they thinking about their automation and AI initiatives at this point in time?

Sure. Yes. I can take that. Demand seems to be pretty resilient and strong. I would characterize it as robust. The product is certainly resonating with companies that want to improve their processes. There is a macro component to how quickly deals get done. And I think that's probably the biggest impact from this backdrop of macro uncertainty. And so a few quarters ago, we started to see some deals slipping. We weren't quite sure at the time if it was just kind of noise or what. But over time, we've kind of seen it persist. And I think it certainly is, as expected, companies that are trying to scrutinize their spend and get some more approvals in place before they spend any money. The good thing is we're able to close those deals. We're getting through those approvals. And things are going well. They're just taking a little bit longer. And that's something that we continue to see in a macro, but overall, the actual demand from the business and the product users is really robust and we're happy with it.

The only thing I would add there is this is something that we've contemplated in our guidance. So the positive thing as Matt -- Mark said, some deals are getting slipped, but none of those deals are getting canceled or we're losing to competitors. It's just taking a few weeks longer to close those deals.

Okay. Has there been, I guess, any like downsizing or a change in the scope of deals as they've come in either from the ones that maybe are slipping a little bit or from some of the others?

That's a phenomenon I think those that are in line to approve the deal would like to happen. I think the approvals would go a lot quicker if we started cutting price, and we're really resisting that urge. We really do have a compelling product proposition. We've got an ROI attached to every deal that is pretty compelling. And so we stick to our guns and we try to get what we're owed for us. And that's why we're choosing again, for -- again, this is the minority of deals. But for when there is such a phenomenon, we choose to let a slip happen and have it closed in a few weeks rather than start negotiating down a price and just trying to close it to close it.

Okay. That makes sense. I guess we'll talk about the topic of the moment of AI. It feels like every single person is talking about it at this point. So I guess to start, maybe talk about like what your strategy is today around AI? And what have been maybe the early -- early feedback points from customers when it comes to what you're doing in the market today for AI?

Do you want to take that one?

Yes. So we've been working with AI as part of kind of what Srin talked about -- when you look at our overall process approach, we look at all the different actors in the given process. And so we've got different systems we're integrating with. We've got people that are obviously playing a role in the process. Then we've got RPA, if there's an instance where you need a bot. And we have AI capabilities like e-mail summarization, intelligent document processing, things that historically, businesses have found very useful to do things faster and better. And so today, I think we've turned a new leaf, right? There's a lot more excitement that's warranted. There's more investment around, hey, you could do a lot with generative AI and LLMs. And so that AI component of our approach is getting a lot more focus. And we're looking at ways to make kind of a bigger boost out of our apps and the mission criticality of what we do through using some of those technologies. And our flavor of that when it comes to generative AI and LLM is really a private AI approach. Our CEO has spoken about this, he's written editorial in the Wall Street Journal. It's really much more inherently -- it makes more sense apparently, right, that a company who values its data, which, by the way, is pretty much every company, would want that data to be kind of contained in their own ecosystem, in their own domain, in their own servers, if you will. And so they want to own the LLM and they want to train it. And we're really positioning ourselves to help facilitate that. And kind of this public and open domain really isn't the place for large enterprises, governments, insurance companies, pharmaceuticals, what have you, to test out and design new stuff, right? They need control over that, and data is really the crown jewels of what they have.

I guess when it comes to your AI strategy and the product set, what is it that you kind of built out natively yourself? What are you partnering on for other things? And just how are you thinking about how the ecosystem develops and what your kind of place is in that?

The way I would approach that is two-part, right? I know Matt has talked about this publicly. I think AI is going to be a product very useful -- it's going to increase the overall market opportunity for us, but AI works well when you combine it with humans. That's our view. So from that perspective, when you look at it, private AI is one approach that we have taken, where we will build domain-specific LLM models, either on our own or we will partner with customers and -- few customers and build those models and give those models to our customers. They can train those models using their own data. That's one aspect. The second aspect of it is, at the same time, we'll also provide opening AI integrations to some of the public AI models up there. So if a customer chooses to leverage some of the capabilities of the private -- public AI, they can do that. But they will be training their own models on their own. All we are doing is facilitate customer to help base train those models. Again, this is within the context of that private AI. When we talk to our customers, clearly, there's a lot of excitement about the AI technology, how they can deploy within their IT stack. But it's still early at this point of time. We also plan to have a separate pricing or premium pricing for this AI functionality. That is something we will likely to come up later this year or early next year.

Okay. All right. That makes sense. Maybe switching gears a little bit. Just maybe back to the demand environment. I mean, it seems like public sector has been really resilient, has been really strong for you all. I guess, how is public sector doing maybe compared to the commercial side of the business? And I guess, what is it that is helping with customer budgets at this point and where you're seeing strength versus maybe relative weakness?

Sure. Yes, there's a confluence of things happening. Conveniently, the macro impact is a little bit further removed from the public sector. There's just a less of a focus on that. A little bit more resilient to macro headwinds and concerns. Sure the sales cycles are a little bit longer, but they end up kind of being less dynamic and subject to kind of what's on the news today and kind of the ups and downs of the economy. But the other thing is we have pretty robust solution in the government space, and it's the government acquisition management suite. And so this is kind of a prepackaged offering that is really an eye-opener for those involved in procurement in the federal government. And we think it has implications in the state and local government as well. So we're definitely investing and looking at that as the next stop, if you will. But there are success stories in 2023 in the GAM space in the public sector in the federal level that really have helped propel us. So those kind of combined things, the insulation from the economic headwinds and then our resident application solution that we've built in that space has really helped the public sector do quite well for us this year.

Okay. I do want to touch more on GAM and I guess the learnings there to maybe drive to other prepackaged solutions and those opportunities out there. So I guess what have you learned from GAM that maybe could be applied to a different use case? And how should other kind of prepackaged apps built on Appian will look like in the future?

Yes. So GAM has been, I think, our third or fourth solution that we've built. There's definitely a practice to it, right? We're finding the best components of what we know how to build that we built before. We're showcasing the power of Appian and the different components of our platform that we've spoken about. And we're putting it all in front of the customer and saying this is what really will solve your pain points in a fulsome manner from an end-to-end process automation perspective. And usually, we're holding hands with a partner on what -- when we're doing it. So we're talking to the customer in front of -- alongside a system integrator who kind of is ready to go and -- starting to build the solution, and there's a lot more of a speed component to the deal, right? You've got everything ready. You don't need to sit around and decide what the application might look like, and you're really ready to hit the development. And in this case, since the solution, the development is really plugging it in, configuring, it's not starting over and building something from scratch by any means. And so the actual deployment is faster, too. And these days, speed is a lot, right? It matters a lot to our customers. And I think focusing on getting an app up and running to go live is really a lesson that's learned. We know that customers care about that, and we know we have the capabilities to do that with our platform.

Okay. Is there a way to quantify how much quicker the -- I guess, the sales process is? How much quicker the actual implementation deployment is for something like GAM compared to when you're selling kind of the fuller Appian platform and kind of more tailoring the use case?

Yes. The impact is more pronounced on the sales cycle. I think that's the most -- I mean the reason I say that is because usually, our development is quick either way. And we have the Appian Guarantee, where we can build a really useful application in 8 weeks. So that's pretty good, right? The sales cycle itself is where we see a dramatic improvement. We've got longer sales cycles when we go for a platform sale, upwards 9 to 12 months. That's the thing that gets cut down dramatically. I don't think we have statistics on it that we've shared anyway.

The only thing I would add there is with Appian platform, right, customers can build anything they want, right? Everything will be going to a customer. Everything is pretty unique. They are trying to build something that fits with their unique workflow processes. Whereas on this particular solution, even though they're buying a solution, it comes with the platform. It's not like this is a stand-alone product, and they're not trying the platform. The only thing here is they know exactly what they're -- for example, you take Know Your Customer. Let's just say that's a situation, right? If I come to Citi and you say, "Hey, we have this requirement." So you buy the platform, you're building that. I know the use case that's using for it, but we think the same use case could be relevant to other financial services institutions. So if we build that use case and now when I go to that, I'm saying, "Guys, this is a use case. So they don't have to guess what they are building." They're readily seeing it. And that's where we think that sales cycle is much more faster and because the product is already built, deployment is quicker. Yes, there's going to be some element of customization would -- but it's a much smaller.

Okay. So I understand GAM, understand the continuity there. How do you think about like what the next use case is that could be prepackaged is? And I guess, internally, what is the process look like for that decision to get made and for engineering, sales, whoever is involved in that to make that decision?

Well, we think the pipeline is one thing. We do have a list of potential solutions that we are working through. However, we're not actually going down the list. At Appian, we really believe in doing something well and thoroughly. And I think the -- with the GAM solution right now is a really good opportunity for us. It's only the beginning. There could be a lot more there. And like I mentioned, the next chapter of that same space and that same solution is the state and local government space. So while it would be kind of interesting to switch to other solution, I think we're kind of focused on that pivot to the SLG space as the next step. And there are differences, right? You have a different ecosystem, a different procurement process. There are application changes that need to be made but we want to do it right, and we want to have the product to manage people. So that's -- with a company our size, we'd rather focus on those. And again, the market is huge, and we do see a lot of opportunity. So we're going to take it one step at a time with those 2 for the time being.

So in this base, I think we have a senior executive who runs our solutions and he is very important to Matt. So he is not only identifying what are the next solutions opportunities, but he also works closely with the product strategy and the engineering team to have them built out. Just I think 2 months ago, we launched this connected underwriting solution that's focused on the insurance industry. So there's going to be a few more solutions that's going but we have a senior executive who is exclusively focused on the solutions opportunity. So this is like a key priority for Matt and the company.

Okay. All right. That makes sense. I want to touch on something, Mark, that you said a minute ago. Just large opportunity for everything that you're doing. And when I think about your net retention rate, I think it's held in remarkably low in kind of the mid-teens area. So I guess -- to ask, how have you seen changes in your customer expansion behavior? And is there a way to think about the differences between pricing, more users, more use cases that is helping support that net retention level?

It's really the additional use cases and more applications, more projects. What we found is actually pretty encouraging. And it's that the prolific Appian user that's built 10 apps or 15 apps. They're a large enterprise or government agency. They're already paying us perhaps 7 figures ACV per year. Those -- that cohort of customers is actually expanding more impressively still than other cohorts. So the more you use it, the more you like it, the more you buy. I think it's a phenomenon that we're encouraging ourselves. And we don't really see a trailing off of expansion. It's really the more you buy, the more you want it. So it's probably helped offset some of the difficulty expansion -- expanding in other parts of the economy or in other parts of our cohorts. And we're happy with our resilient NRR. We'd like it to be higher. We are focused on getting NRR as high as possible. But we'll take steadiness in this macro.

So I guess with that point of view, the largest customers that you have continue to expand at an outsized rate, how much more opportunity is there for Appian to get deeper within those accounts and kind of expand the already large footprint that you have?

I think there's a lot of opportunity. It's not the case that we think we're tapped out by any means on any of these accounts. There are balances, you have to strike, right? We are a company that focuses on new logos and acquiring more customers. We use our partner ecosystem to help with that. So it's certainly easier to expand and sell more to an existing customer, especially in this macro environment. But we can't just do that and be good stewards of our proposition here. Our product and our market demands more. So we are focused on new logos as well. If you want to add there?

Yes. Maybe look at the penetration among our large customers, either by IT spend level or the number of applications that they're managing relative to what they have built with Appian, we still think we are only 5% or 10% penetrated. So it's still a big market opportunity. On top of that, we are constantly broadening our platform, adding new features and functionality. We've talked about data fabric in the past few quarters, since it's been launched, it's one of the most widely adopted. The uptake has been 90% plus among our new cloud customers. So we think as we roll out more features, this -- the penetration within the existing customers is still pretty big.

I do want to touch on data fabric since you bring it up here. I guess, first of all, what is it? And what is it that it helps your customers do? And how does that kind of differentiate the Appian platform?

It's a cloud native data and analytic engine. More importantly, what this enables customer is to keep data where it is and easily integrate into our Appian platform in a relatively easy fashion. When you look at other -- the way other companies use that is, for example, if you go to any other vendor, they make you write a custom API in order to migrate the data onto their platform. But as with this particular data fabric feature, you don't have to do any of that. In fact, it's a patented technology. And we think it not only helps customers manage their data easily in a highly secure fashion, but also it enables them to build new applications much more faster.

Okay. So is it something you're directly monetizing? Or is it more about the usage and being able to build faster, and so you monetize it kind of secondarily?

I think it's more about the usage, and I think the monetization is going to come later. Today, it's not like if they adopt data fabric, we're going to charge them a separate price. We want to drive adoption and usage, and we think of monetization aspect overall.

But we've been really successful at driving adoption. It's very highly adopted amongst the new customers, over 90% are using it.

So I think Matt has talked about it quite a bit of data fabric helping with the AI strategy and helping kind of differentiate it. So I guess, what is it about data fabric that enables that private AI strategy that you've talked about?

Well, to create -- part of this private AI strategy hinges on creating your own large language model, right? And to do that, you need to bring in data from all sources. And oftentimes, in a corporate or government, large enterprise environment, the data is everywhere. There's disparate systems. There's different entities that you've acquired over the years and it's a proverbial nightmare, right, to get all the data in there. And so with data fabric, you not only just get the data, you can do it securely, you can do it in a way that is controlled, trusted -- it's fast. It's performant. It's low code. So all of those things enable you to bring in all the data you need and even use intelligent document processing, which is AI-driven to get it off of let's say you have paper forms that you need to read and incorporate in your LLM. So it's kind of the first step, right, before you start training and getting the iterative machine learning and different things you need to do, you need the data, right? So that's where data fabric comes in.

Okay. Last question on AI and then we'll leave that to bed here. Monetization, just how are you thinking about it at this point? And kind of what are the implications for what charging for AI will do for Appian?

As I said, AI, we've already been investing in for quite some time. In fact, some of the solutions, products that we have already have some AI capabilities. Government acquisitions management solution is a prime example of that. So AI is already built in. So customers do pay for it. It's not like we have a separate SKU, but it comes as a part of the platform, and so they pay a premium pricing for it. With respect to GenAI and LLM models down the road, we probably will have some premium pricing for it. And to -- if customers were going to sign up for that AI functionality, there is going to be a separate premium pricing. What that is, we have to determine that. That is something we're probably going to communicate later this year or early next year.

Okay. All right. That makes sense. I guess what we'll stay on the product side, I think you have made quite a few acquisitions in the space to build out your automation platform, RPA, process mining. Where does the automation portfolio sit today? How is the adoption of kind of those other solutions that you've acquired into the Appian base going? And I guess, finally, how much of a differentiator is that compared to some of the automation peers that you're competing against?

I think our advantage is that we have the hard stuff down. The hard stuff is the actual workflow, application building. That's our DNA, where we're actually getting all these integrations built, all the different actors, like I said earlier, to play together to actually get stuff done in a company's environment. I think that's our forte. When we have RPA, it's -- our approach has been different than RPA-led companies, right? Those companies pretty much want to put it a bot everywhere they can. We're putting a bot where we absolutely have to. We'd rather build an integration. We'd rather use data fabric. But sometimes we need to put a bot. So we have that asset. The other asset that you mentioned, which is more and more important today is the process mining piece. I'd say it's important today because we're using it as a way to show the value of the things that we're building in Appian. And Process HQ is our kind of way of framing up that proposition. We're showing all the information that the customer would care about, about what their process was before Appian and kind of how Appian is helping streamline. And it's a continuous, real-time assessment of how long things were taking to go through the process, how much time or money you're saving or whatever unit you'd like to track. All that is enabled by process mining and getting the plethora of different processes on a piece of paper in front of the customer, or on a screen in front of the customer to actually show the power of what is being built and how it's running. And it plays into this maintenance piece, too, like you need to -- it's one thing to build an app and go live but you have to live and breathe it. You have to go through it every day to make sure it's still doing what you want it to do. And that part is facilitated by process mining very well. So those assets are kind of the building blocks. There are, at this point, not too many things that we can't do with Appian, but we'll certainly look out for -- on an opportunistic basis for anything else that might enhance our technology.

Okay. I remember when you made the process mining acquisition, there I think was a view that maybe it would kind of be like a tip of the spear to help kind of bring customers in. I think maybe it shifted towards then being more of like an expansion opportunity. So I guess, how are you viewing the monetization potential of process mining today? And just what's the view of how it changes maybe the go-to-market here?

I don't think our strategy in process mining has changed, right? We still said it's going to be part of the platform. So when the platform is fully integrated customers still get it. It makes sense. The existing customers who have been using Appian platform for a very long time, are likely going to the first adopters. So again, this is very consistent what we cited in the past. At the same time, we're also putting that capable team from our partners also so that they would have to talk to new customers out there. So it's not like one is more important than the others. Like it's equally important about everything, but it's kind of that initial adoption, we think we will see much quicker adoption by existing customers because they already know how to use their Appian platform, and they've been using it for a very long time. So from that perspective, I think we've seen some pretty healthy uptake. And we've talked about customer examples that used process mining. Some of these are large global multinational customers who have signed up for the process mining capability. So we feel very good about it. And in fact, we still think the process mining capability is very relevant, especially in this kind of macro environment where customers are looking for quick ROI. They are looking, trying to find process inefficiencies within their IT stack. So process mining is a perfect tool for that. The one advantage is, it's fully integrated to the platform. So a customer doesn't have to like write an open API and say, oh, I wanted to integrate with someone else. So just dealing with one software out here.

Okay. That makes sense. Maybe switching gears a little bit. I do want to ask about the 8-K that was filed this morning for the insurance on the lawsuit. I guess a couple of questions. I guess, one, can you just give some more details about what was actually announced today. And I guess secondarily, I guess, what kind of led to the decision to do this?

Sure. Do you want me to give a background on the lawsuit or...

Yes, sure. Let's start there first.

I think I'll do it briefly. There's a lot of materials out there, but there was a trade secrets misappropriation lawsuit. We won the lawsuit and the $2.08 something billion verdict against our top competitor Pegasystems. There's an appeal process underway. What we've done is succeeded in getting a judgment preservation insurance policy awarded to us or purchased by us. And that kind of creates a floor or a locked in value for us at $500 million. So no matter what happens in the appeal process, we feel very confident in that process. We feel confident in our overall victory. But now we don't have to worry about anything getting overturned because we have a floor of $500 million. So the mechanics are simple. If you -- if the judge for some reason, throws it all out and rules that there's a $0 award, we get $500 million from the insurance company. If the judge says the award is $200 million, we would collect $200 million from Pega and get $300 million from insurance company. So we get trued up to $500 million. If the award is ultimately, let's say, $700 million, insurance company wouldn't pay us a dime because that judgment was preserved at above $500 million, and we would have to collect from Pega. So it's really a win-win at this point for us. We have a new kind of floor like I said. And we'll fight to get it all, but we're -- we know the minimum is $500 million.

Okay. Maybe a follow-up on that. I'm sure that there's going to be appeals -- I know they were in appeals process now with that. I guess what's kind of the time line that you're kind of thinking about today? And I guess, I'm sure no matter what happens, there's another level of appeals here. So at what point does this ultimately kind of get finalized?

It's hard to put a time frame but give and take, it's probably going to take another 2, 3 years. The oral argument and the appeal process may be like [3 years with us ] and the judge needs to rule on it after the oral arguments are done depending on the outcome of that either party can chose to appeal that. So again, these are all in a moving process. But I think at this juncture, we think it could take another 2, 3 years.

Okay. That makes sense. That makes sense. Maybe if we go to the go-to-market a little bit, new CRO, Chris Jones, has been in place for, I think, a little over a year now. What are some of the highlights of what he's put into place to maybe change or evolve the go-to-market? And what are kind of the strategic priorities today when it comes to the go-to-market function?

Yes, CRO has been there for a little more than a year. We're extremely happy with what he's doing so far. What I think Chris Jones wanted to do was more fine-tune rather than reinvent the wheel. So it's a part of that, I think he has flattened the organization structure so that there is quick communication between different teams so that it enables them to make decisions much more quicker. But more importantly, I think what he is aiming to do is drive sales force productivity, number one and number two, accelerate new logos. So if you were to boil it down, those are the 2 most important things that he's looking to improve on.

Okay. I know that there's been a big shift for multiple years now to push services more and more to the partners. Where does this effort stand today? And maybe how is the channel and the partner opportunity different now relative to a few years ago?

It's pretty -- it's a longer game for us. We've been at this for a while, but it's been pretty successful. And so far as a good chunk of our new logos are in fact, influenced by our partner ecosystem. You can see the proof in the pudding with our services revenue being much more slow growth. In some cases, not even growing, which speaks to the fact that work is going to the partners, and that is because they brought us business in new logos. Dovetailing to the earlier question around Chris Jones, our CRO. He's a huge partner guy, and he definitely wants ways to get those partners more involved. There are strategic partners that are very involved at Appian that we want to even double down on with those relationships. The ones that we're doing a lot of business with -- ones we're looking at to really say, okay, let's up this even more. And so we can be the vendor of choice for those clients of theirs that are looking to deal with automation effort. And the AI chapter here is going to bring in a lot more conversations around that. So we want to be ready to have them.

Okay. And so I know you mentioned that there's a lot more partner influence in the market today. At what stage are we in terms of partners that are, I guess, pulling you into opportunities rather than maybe helping fulfill the services angle. Like is that becoming -- I guess, where does that stand in terms of the go-to-market?

No. I mean we don't -- our success and our efforts in the partner ecosystem and our partner initiative is exactly that, for them to bring us new business. And it's not necessarily just firms that can do Appian work so we don't have to do it. It's really we want their footprint out there, which is huge in the case of like KPMG is a strategic partner of ours. It doesn't really get much bigger than them and Accenture and then some of the other partners that we have. Those massive workforces that are out there in front of customers are ones we want to tap into. And right now, the majority of our new logos are influenced by those guys. So we want to keep that going and make it even more.

Okay. You made a comment at the recent Analyst Day about wanting to drive a significant increase in rep productivity and continue to increase their efficiencies. I guess -- what are the key levers that you can pull to help drive those improvements? And what does it take to make that significantly higher?

I think some of the initiatives are just basic blocking and, tackling right -- as a part of that, as I said, Chris Jones has already made changes where he is flattened in the organization structure. I think one of his conclusions when he was looking at -- there are multiple management layers that's delaying the process and in turn, that's taking longer to close deals. And that, in turn, impacts the sales force productivity. So that's part of number one, He's already done that. Number two, we talked about verticalizing sales force, right? He's already doing that, but also creating a separate sales force that are focused on the very large customers. We still think there's a ton of opportunity with these bigger customers that are spending -- we're trying to see -- we have customers that are spending $1 million ACV. How to move them to $10 million ACV? What are the things that we need to do there? So we are having some dedicated sales force that are exactly focused on those very large customers. The third one is also having a renewal team separately. In this way, your existing sales force is not worried about the renewal because this is basically, hey, renewal comes up, just making sure they signed the document, et cetera. So we have a separate renewal team. Even though team takes care of it, the sales force still gets the credit from accounts. So these are some of the things that are going to make to improve their sales force efficiency.

Okay. So if we tie that go-to-market and trend to drive efficiencies to the 30% cloud sub growth number that you've talked about. I guess, how does that kind of tie together -- and how do you -- how should we be thinking about the growth equation to get to that 30% level?

So we've talked about when you look at the growth, like 2/3 comes from existing and 1/3 comes from new logos. We think that particular math, that balance is going to remain relatively the same. But more importantly, within the existing customers as we see a pickup in sales force productivity, there's opportunity for that to be a bit more higher. And over time, again, our ambitions are at least 30%, but, certainly, it's much higher than that. So as we see sales force activity improvement, we should see some movement both within the existing customers as well as the new customers. But with respect to the mix of contribution within existing and new, it should be to [ roughly the same ].

I think we're running up against time here. So Mark, Sri, I want to thank you both for being here, and I want to thank everyone for -- in the room today for being here as well.

Thank you.

Thank you.