Docebo Inc. (DCBO) Earnings Call Transcript & Summary

[Audio Gap] Docebo; and Giuseppe Tomasello, Head of AI. I have some disclosures to get started. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. Sukaran, I wanted to start high level to help level set the opportunity, some of the key products and use cases for Docebo's Learning suite. If you could start there, and then this -- yes, that would be great.

Yes, I'll give a quick overview of what we are at the business and what we do and kind of what we've done in the last year. So when you think about Learning and Learning Management Software, most companies or most folks think about it from a perspective of employees, what's different about Docebo is that we not only serve that audience at 35% of our business at a minimum, but we serve an audience that supports all facets of learning at every organization, whether it's your customers, your partners, your members, your franchisees, and that's what we call an external audience, anyone but your employee. And then you have the internal audience, which is typically compliance, onboarding, sales enablement, customer support, so on and so forth. The difference -- the reason why we've done reasonably well in the past few years is that when you think about the landscape from a total addressable market perspective, the employee -- the internal audience, which is employee training for sales enablement, onboarding, compliance, so on and so forth, is typically thought about from a perspective of number of enterprise companies multiplied by number of employees. If you take that in terms of what we do 65% of the times, at a minimum with our customers is to support them on external training, external learning. One example, which is one of our top 10 customers is AWS. All -- close to 6 million annual customers of AWS go through our platform, give or take from a learning perspective for all things AWS certification, and that audience is so much larger compared to the AWS employee size. That gives you some sense of the addressable market and predominantly the external audience, whether it's customer education, partner education, our franchisee and so on and so forth, is dramatically greenfield. Most of the times we're generally competing from an in-house bill solution versus Docebo. And we started this story in early, let's call it 2015-2016 era, with our first marquee customer, Thomson Reuters, which still is my top 10 customer, effectively uses our platform for all of their finance and legal intellectual property, that they effectively use our platform to use that content, that intellectual property that sits in their ecosystem, convert it as a product, skew using our LMS, and sell it to their customers as the end user. So a lot of our customers utilize our platform to be a distribution mechanism of that beautiful intellectual property which is our LMS. That's what it does. But at a high level, we have 3,700 customers, enterprise customers, just under $200 million in ARR, 81% gross margin. We just reported 28% subscription revenue and growth a week ago, 17% free cash flow margin. So give or take, on a free cash flow basis, a Rule of 45 on an EBITDA basis a rule of 41%.

Great overview. Thank you.

A lot of focus from you on the external training opportunity. You also have a full platform to address the more traditional internal learning, LMS market. And over the years, you've gone from kind of core LMS to a broader Learning suite. There's 2 questions there. Are there other parts of learning, specifically talent development that is still ahead from a product rollout perspective? And then more broadly, are you looking to stick within talent development? Or are there other areas within HCM that makes sense for you to get into whether it's acquisition, core HR or the broader HCM suite?

Yes. I think generally, what we would say is that the HCM space is very important for us. It's -- if you think about -- I'm not seeing legacy players, but folks that are in the HCM space or even a company like Cornerstone, which is now private, which used to do this a lot, and we compete with. They built a platform which is designed around the employee experience, right? Effectively a Workday or some of the other platforms, they are not only just -- they're doing a multitude of things from HRIS, HCM perspective, and LMS is a component of that ecosystem; one of the challenges with that model from a learning perspective is that it limits your product and road map because you're designing a bigger platform from an employee experience perspective. So when you think about learning in customer education, there's so many capabilities required, even just simple things as e-commerce capabilities that you have to embed so that the customer can charge the end user. Those are not things that are thought through when you design an LMS from an employee experience perspective. And that's where I think to kind of answer your question in a straight way, we are going to not just be in LMS, but our vision is to be the learning company, and we're not necessarily saying we're going down the space of HCM, but there's some functionalities that are important that we capture, one I'm getting early and I'm sure Giuseppe will speak to. There's a legacy problem where a lot of systemic shifts are happening in multiple industries where workforces have to move from X to Y, whether it's EV transition, whether it's onshoring, so on and so forth, we work with some major customers on the manufacturing side. And that question has been on the HCM side, which is a product functionality that I'm sure Giuseppe you can step in after this, is that the question of skills ontology, right? So, if I've got a move, we've got great customers like Intel and others, I got to move onshore a new chip facility in North America. I also have to move a number of my employees from X to Y function, the ability to be able to map those skills of those employees and then to build a career path that they can move to a new function of an organization. This is one problem that we're going to take head on and release some feature functionality this year. I'm sure Giuseppe can jump in here. But -- that's how we think about it. We're not necessarily going to the talent space. We're focused on being the learning company end-to-end, whatever the use case for the customer may be. I don't know, Giuseppe, if you want to comment on the ontology side.

Yes, absolutely. So every customer that we onboard in Docebo, and more and more, actually, they're going to have their own specific skill oncology. So Skills is becoming one of the core foundation for all the HR world inside the corporation. But what we're seeing right now is that we're building an AI side, the capabilities to map the specific skill of a company -- the specific ontology to our Docebo core ontology. And this is actually very important because we are -- by doing that we are capable of generating content that is specific for those specific skills of each company. So what generative AI is allowing us to do right now is not only to tag existing courses with those ontologies, which is like, let's say Layer 1 of AI capabilities, but what we're doing now is that AIs are enabling us to create content that fits specific skill gaps that are identified inside the organization. And the big part of our work right now, our big focus is on the content generation capabilities that is going to allow us from a knowledge management standpoint, we're able to ingest all different kind of documents and even like transcripts of calls and any kind of unstructured data. And then our engine is capable of getting this knowledge and create the specific content that is mapping with a specific skill gap identified. So this bridge exists between skills and content generation is going to be one of the main advantages that we're going to deploy in Docebo as we move forward.

Perfect. And we'll dig more into Skills and AI in a little bit. A couple of other things I want to lay out upfront, like one piece of your story has been move-up market kind of away from SMB based over the years, and we've already mentioned AWS and Thomson Reuters and Intel. And there's a lot of other customers that we know about as well. So could you talk a little bit about the move-up market, kind of where we are and some of what those largest customers are using you for?

Yes. No, it's a great question. So I'll just give some history very briefly and some specific decisions we've made. So we went public on the public in 2019 and then subsequently in the NASDAQ in 2020, and we called out in our perspectives that we will continue to move-up market, there's a number of reasons for it. If you think about smaller, first time SMB type of customers, they are generally macrosensitive, less strategic, do not have the capability from understanding whether LMS is particularly important to their organization because they're smaller, nimble organizations. And that's the space where it's not optimal from a growth and net retention perspective. So, we made a deliberate decision even at that time, looking at the unit economics, that the benefit of serving mid- to large enterprise like we talked about, is that our ability to penetrate, give or take what I spoke about earlier, we serve customer, partners, HR, compliance, sales enablements, revenue -- customer enablement. That effectively means that in mid- to large organizations we can target, give or take 6 to 7 buyer personas. And each of those buyers use my platform as independently as in it's 1 Docebo, but each of those users or departments is in charge of their own destiny. An example just to iterate that better. And the reason why we did that was if you take a customer like Bridgestone, they use us for customer education, partner education, compliance and onboarding for 1 Docebo, 4 different instances. And each of those owners of those instances have nothing to do with each other, right? The Head of Customer Education at Bridgestone doesn't necessarily talk to HR to run customer education. And the reality is the SMB customer was one that we -- is not in a position to be able to do so. And then from a growth-net retention and an LTV to CAC perspective is just suboptimal. I'm just being frank here. And so we made that concert as a deliberate move up market, where mostly we work with companies from a price point perspective that have mid-market will be anywhere above 2000 employees and above, and then enterprise customers will be 5000 to 10,000 and strategic 10,000 and above employee size, just roughly, just giving you some kind of segmentation information. What that effectively does is 2 or 3 things. One, we can support them from external use case, which is we spoke about customer education and most -- and a lot of the times, the customers, these are revenue generating departments. It's much more integral to their business strategy from a customer, whether it's a product adoption, customer education or partner enablement perspective. And then it gives me the ability to target multiple departments, so it's much more efficient from a CAC perspective, because the second best CAC in my world, first is churn downgrade saving, but the second best is expansion into that base and it's much more efficient. And so we have been very deliberate in being a company that supports mid-market to large enterprises because not only helps us to win business on the onset, that is at a higher ACV relative to SMB, but it also gives us the ability to expand. And I'll just shut up after this point. Almost 60% of what we did in the past quarter came from large enterprise customers, the ones with 5000 employees and 10,000 employees above segment. And so as we move forward, I think it's reasonable to expect that we'll continue to work our way into the mid- to large enterprise customers because they're just from every unit economics, KPIs makes sense for us and that's what the product is built for.

So does it -- is it a goal of yours to become the like wall-to-wall LMS? Or is that not even important because you are getting into these big organizations and can go department-to-department use case to use case?

That is generally what has been the case in terms of the last 4 years. What is -- in terms of -- firstly, I'll say Alessio will -- my CEO will say it's different, is that our goal is to be the learning company, not just an LMS, meaning that we can do more from a software and capabilities perspective, whether it's content authoring, whether it's performance management around learning and other things. But if you would take an example, which we spoke about, AWS, the ability for us to go from an AWS to Amazon Logistics, to AWS Engineering and multiple departments, we've historically done that quite well in terms of land and expand. And the reason we did that initially was, if I start with a learning management -- a use case of a customer, that is, customer education. I'm way more integral to that organization and I'm much more -- necessity. And that leads me to play the consolidating the tech stack game better over time and eliminating the folks that do learning on the internal use case, which is employee training. That has worked quite well for us. But what's changed since 2022 is there's a new villain in town, which is people like us, the CFOs, who are trying to consolidate the tech stack and want to get -- want to make sure that their tech stack is also they benefit from having less software in their orgs from a risk perspective, but of course from a pricing and consolidating the tech stack perspective, we've been able to increase our ACV. You have seen this quarter too, because what's happening, the CIOs and the CFOs are now more thoughtful about why do I need this LMS that does employee education, but I have this LMS that does customer education. And that's where, we've been a bit more targeted in consolidating that on the onset and lifting our ACV in the last 2 years, that's just -- the market has changed that way.

Got it. And one on macro. When you think about the last couple of years, definitely some headwinds there that maybe have met some reprioritization around Skills and AI and Learning. So how would you assess the current macro backdrop and maybe there's 2 different answers between SMB and Enterprise?

Yes. I will start with saying, listen, in our world, part of it is -- we deprioritize to an extent, but also part of it is that in the world of learning, SMB customer is the one that's more macro-sensitive, budget-sensitive. And we haven't -- I'm not saying that's been an area that we continue to participate. But I would say that's where the weakness is. Where we saw strength in the last 2023 fiscal year, as you would have seen, was mid-market and large enterprises. There are significant themes, and learning is integral because one you have from an employee base perspective, that productivity gains are only going to come from products like ours, from enhancing what employees are doing, or our customers and partners are doing. So if you take content as an example, Docebo Shape is a Content Automation module which creates content. Effectively, what you're doing in that case is instead of paying X-million dollars for employees and outside providers, you're automating creation of content at 1/5, 1/6 the cost instead of paying a consulting agency. And so you're effectively reducing costs or you're being able to manage higher productivity from an employee base perspective, and that's been a bigger problem to solve. And then there's themes that are playing out across multiple spectrums. Customer education is huge because that drives higher product adoption and higher usability and that customers are looking at that and saying, "Well, why am I not driving? Why do not -- I don't have", like if you looked at customer education 5 years ago on LinkedIn as a role, Head of Customer Education, you would find that this theme has played dramatically because now customer academies drive product adoption. And so those type of themes and onshoring and multiple industries are giving us some benefit at the mid- to large enterprise to penetrate better. I will say that as you move forward, one thing I will highlight this macro question, this is my personal belief that the rates that are being cut and that macro improving won't necessarily change the buyer behavior in the next 2 to 3 years. If you look at it from my seat, you have to demonstrate value, ROI, how quickly you can actually help the customer. Because if you don't do that, they're not just going to loosen their wallets because the rates are cut. The buyer behavior has changed dramatically, and the last 2 years has clearly shown that ROI driven fast enablement at the customer level. If you can't demonstrate that, you're just going to get kicked out by the CFO or the CIO's office.

Got it. Let's dig into AI a little bit more. So maybe Giuseppe, starting with you. Some of this you touched on, but hoping to really lay out clearly the differentiation for Docebo. Like, why are you well positioned? What's proprietary? And where does your advantage come from?

Absolutely. So first of all, I would like to say that AI has been of the -- a big focus of Docebo for the past 5, 6 years. And so I would say that the advantages that we have right now are in 2 different domains. One, I will say on the data perspective, and the other one is on the product workflow integration perspective. So on the data side, we have been organizing our learning management system, having a very solid integrated content lifecycle workflow, where actually from the content creation with tools like Shape to content consumptions, and then actually the business impact of the content has been all integrated with different product suites. We have Docebo Shape, Docebo Learning Impact, Docebo Learning Analytics, and this data consolidation perspective give us a very -- richness of data in terms of all the content and how the content is consumed from learners. That now this is actually a big advantage for us, because it really set the foundation for a solid data base, to then build on top artificial -- generative AI system. On the product workflow perspective, we have already tools launched in the market. And I want to say that Docebo Shape has been in the market already for a few years, so actually a few years earlier than all this generative AI hype entering the market. And we already been learning a lot from our customers using this kind of tool. What we are doing right now, since we already own the workflow for content creation, we are adding a lot of other new capabilities to our Docebo Shape product. So I believe that this is also giving us a competitive advantage because we already have a product in the market, we already have customers using the product, and now we are leveraging our data and also and also know-how that we acquired through also M&A in order to create really like the next generation for -- content creation for learning.

And the one -- only one thing I'm sure we'll get into specific products. I will say that I'm sure this is a theme you're hearing is ultimately a Learning Management Software company like us has -- the biggest benefit we have is that, I won't -- I mean, he'll slap me afterwards for saying this. But the intranet, the knowledge of my customers goes through the LMS, which means we have the luxury of the data of my customer in our ecosystem, that is intellectual knowledge of the customer that sits in our platform. We have over 40 million registered users, 3700 customers. That's where you get the power. AI functionality is great, but if you don't control the data, that's where generally companies will struggle, in our perspective.

100%, Sukaran. And actually that is one of the obstacles for newcomers in the market is the fact that you need to get this backbone of data knowledge inside the platform before you start having all these content generation systems that you can fully leverage. Because the way that generative AI works is that of course you have large language model capabilities, but those large language models need to be grounded to the company knowledge. And the company knowledge live within these learning objects already in the learning management.

Very clear. Maybe it would be helpful to go back to what exactly is Docebo Shape? And then what changes with version 2 from a feature perspective?

Absolutely. So Docebo Shape, as I mentioned, is a module of Docebo. It is already in the market for a few years already. And by the way, has a great attachment rate. And the reason why Docebo Shape has been successful so far, is because it's generating micro learning pills from, for example, let's say you have a documentation you can upload to Shape this set of documents, like very long 50 pages documents. And that Shape is capable of reading through all this documentation and getting the knowledge points that are relevant for the learners and organize that into a micro learning pill, like these are learning objects that as, for example, voiceovers or even like videos. And it's great because once you create this content with Shape, you can automatically translate it to 50 languages and that will be the voiceovers and the images will be adapted depending on the different culture for each language. And this is great for multinational companies to create content, that then is distributed in different geographies. So this has been Shape so far. But now, of course with generative AI, we are bringing those capabilities to a next level. First of all, we are working on the knowledge management standpoint. We are able to ingest, for example, all the learning objects that are already in the learning and LMS, right? So also thanks to generative AI, we're able to understand what's inside those learning objects and create a knowledge management which is rich with all this information. But as well, what we are capable to do is that we are able to generate the next generation content that is hyper-personalized. So we are, first of all, as we mentioned before, we are working with skills and skills ontologies that is the backbone for personalization. And a new Docebo Shape will be able to understand the specific skill gaps that you'll need to create as an organization, and therefore like create content with that specific skills in mind. Second thing we're doing is that, we are creating a next generation of altering. So it's not going to be just -- we are extending the capabilities of Shape, because we are anticipating what are the needs of the market. And now, we're seeing that the kind of content that Shape will be able to generate are more similar to a traditional altering solution, so there will be different kind of richness in the content that Shape will be able to output. Shape is going to be able to have 2 different access points. One we call the top-down access point, which is subject matter experts or instructional designers in the company can tell what kind of content you would like to create. You have a copilot interface, where you can have a conversation with an AI system, and then the AI will be able to generate automatically all the content, which is by the way, pedagogically sound, because we are embedding pedagogical principle inside our content generation capabilities. The second part is what we call a bottom-up altering solution, which is basically you have kind of like a traditional altering, and then you have a generative AI that is embedded each step of creation of each asset that goes in the content. So the AI in that case, is kind of like an assistant that it can help you to craft the output, specifically to the need that the instructional listeners have. This is just a picture of the new capabilities of Shape, and very important is that Shape is going to be able to produce very engaging learning activities, and yes.

Very great. Yes. Awesome. So then from a monetization perspective, like you still have original Docebo Shape, what's going to happen to that? What's the timing on this next version? How will that be monetized?

Yes. Let's just put it that the incremental features and capabilities, I'm not going every feature but that Giuseppe just spoke about are going to be separately packaged, as we call the Creator pack. Just simplifying some of the kind of the go-to-market motion, and that is the add-on that the customer will acquire from us. And so, it'll be separately monetized. We can probably touch on if you want, Josh. There's a relation, but there's another separate monetization on virtual roleplay, which plays nicely with Shape, because once you -- what Giuseppe said is important, because once you generate content and it's hyper-personalized to the end user, my learning path will be different to his and even simple things as Q&A quizzes, so on and so forth. But then at the end of it, the question is, how do you practice it? And that's where our technology, which is the second product, which we will separately monetize, will be virtual role play. And I'm happy to go into it, but yes.

Sure, why not? Go ahead.

Yes. So the virtual role play is a new, let's say, like, way of learning. Like, usually the problem that we have with learning inside our learning management system, let's be frank, is the engagement. So we call like click-and-sleep, because people click buttons next, next -- next slide or watch a video passively. And that's actually a big problem, because if the learner is not engaged, you're not generating the value you want with the software. So we are solving that problem of engagement by creating new and fresh ways of consuming content. And this is a virtual role play, which is actually like a simulation-based learning, where -- let's say like you are a salesperson and you want to do sales enablement, right? So you want to learn how to sell Docebo style, in case you are a salesperson inside Docebo. What we're doing with the virtual roleplay is that we are able to read all the transcripts of calls, are also successful calls that a person in Docebo will do, and then we are able to import this kind of style inside the system. So now the learner is able to simulate what can be, for example a call with a client, and the client will mimic the same kind of language and style of a real client of Docebo because it has ingested the data and it can do a simulation of what could be a real call. After the simulation is completed, our AI system is able to generate an assessment report that is based as well on this kind of data that's ingested. And this actually is game changing, because now the learner can get feedback directly from the interaction and very importantly that there are other software in the market that have this kind of simulation base, but those software have kind of like a scripted simulations. So the simulation will always be the same. In this case, since we have this other part which is Docebo Shape that is reading all this information, every simulation is actually new, and the learner can really enjoy every time like a new conversation also, but following the guidelines that we set up.

So I mean, just to give some perspective, when we built the initial prototype, I tried to -- effectively I tried to sell to this virtual avatar. It's fascinating how challenging objections I got. I was trying to practice and trying to sell to an enterprise company myself and the avatar -- it's an avatar on the other side, real time speaking to you, and he -- it was a he, I guess it's virtual but it's he. And the objections I got, I would never have got myself talking to a regular customer because he picked up on the -- I don't know if it's he or she, doesn't matter, avatar -- but he picked up on certain capabilities I spoke about in the call and are like, well the other competitor also has it, why are you so better? And this is an extreme level of challenging in term of a cold call and you get a full report card at the end of the virtual call and now your manager or whoever it is in the sales organization understands what your capabilities are, how you should have started the cold call, what features you told the customer about, what did you miss, how did you close, what products and features you didn't tell them about? And I mean that's the power in our role. We'll start like I think, Giuseppe spoke about sales enablement, but for the 2 big use -- 3 big use cases for us will be sales enablement followed by customer support. And then third just onboarding and just helping employees within the organization move different roles and practice how they can enable themselves.

Great. Really interesting use case. I want to spend a couple of minutes on competition, partnerships before moving to financials.

Yes. Yes.

You are the CFO. Competition, I guess, one observation if an analyst and investors coming to do some work on this space, you'll see LMS, LXP HCM providers, content providers, like specialized -- there's a lot of vendors, it seems crowded. So who do you actually compete against? And where do you win, where might a company select another vendor?

Yes. It's a good question. So I would say in our world, mid- to large enterprise specifically I'll speak to you are generally competing with the legacy employee experience platforms, such as a Workday, Oracle, SAP -- listen, I'm not going to criticize, but we'll say they are building an LMS that is a subcomponent of a big HCM/HRIS suite. And so generally that's kind of what we see in the mid- to large market. There will be individual niche players here and there that may do customer education only, but that's kind of the biggest game. And then I will speak to one competitor that perhaps we continue to work positively winning some of that business, perhaps is the word I'll use is Cornerstone. They used to be a company that was public previously. They have acquired. It's a bit of a roll-up situation there. They've acquired 3 legacy platforms that basically compete with each other. So certainly the questions that customers are asking these days is, and again, that's only serving employee experience. Internal use case, not external, where 65% of my business at a minimum is external or hybrid. But in the Employee Learning space, what has happened is it's either legacy platforms that are not focused on an LMS because it's a small component, or the ones that are focused have acquired assets. The problem in that scenario is that which of those 3 assets survives on a 3- to 5-year basis, where, as you probably heard here, this is one place in terms of one vertical Learning Management Software, where AI is real and there's real benefits between content automation skills, ontology. So you have to be on top of your game and progressing innovation, and we certainly are taking a good share from them. We just landed one of the top U.S. banks in Q4, globally, serving all of their 200,000 plus employees, and that's a full replacement of the legacy player, which is I just spoke to. So that's really where we see the competitive landscape. I forgot the other question. Sorry.

No, that was all I asked. And then the second thing I wanted to ask about was partnerships. I think, in the interest of time, we'll just kind of focus on OEM and the OEM strategy. And I wanted to -- I mean, last earnings call, you directly addressed one of your largest OEM partners, Ceridian did an acquisition of a similar smaller, but somewhat similar from high-level capabilities. I think that you addressed that well, not sure if there's anything else that you wanted to talk about in relation to Ceridian, but did want to ask like how you see your OEM strategy evolving?

I will call the OEM strategy as OEMs, SI Partners and other partners we work with. Because if you think about the vision of the company is to be the learning company, part of that will be -- what you want to make sure is that we are able to give the customer end-to-end support, whether it's content, managed services, so on and so forth. And that's where SIs and other partners come into play. I'll start with OEM first, quickly. OEM is a business where effectively what we do is because when you heard me talk about the HR buyer, if I'm in a situation where an LMS is part of an overall HCM suite offering, where the buyer is buying the full HCM suite, I'm not necessarily in a position to win that business because I'm just an LMS in a big HCM suite buy. In those scenarios, we partner with a number of OEMs out there, whether it's Dayforce, MHR in the U.K., Darwinbox in India, EY globally is our partner there. There they effectively white label our product. As part of their offering, they do this implementation, they do the servicing, they excel it and we have a nice rev share that we collect. And that gives me the ability to actually sell to a buyer that I wouldn't be able to participate in the first place. The second thing I will say that's important, I want to hit. I know, 2 minutes left. Is that on the SI world it's really important as to be built out partnership with big 4 and other large system integrators that we'll announce in due course who they are. But as you think about those system integrators, one specifically and the one segment, one part of the business we didn't talk about, I just want to hit on quickly, is that as we look, we basically reinvested almost 2.5% to 3% of free cash flow or EBITDA last year in FedRAMP certification. We have a big 4 SI and some of the partners that are going to be integral. Today, almost 5% of the business comes from Federal and State. We are not doubling down, but we're going to triple down on that business. We see significant opportunities in the federal space, huge market. And the players that play in that market are 20-year old. And we are -- we spoke about in the earnings call, we are in the last -- we have 2 federal agencies that are going -- potentially willing to sponsor us to be FedRAMP certified, if we get through that within H2 of this year. We will have the ability to sell into the federal customer, which we couldn't until we get FedRAMP certified. But the SI in this space is the one that's giving us that opportunity and bringing those 2 federal agencies that are going to sponsor us. The reason I talk about that, I'm going shut up now, is that the SIs are equally important as you move mid-market to large enterprise is because they bring a significant amount of qualified opportunities that if they don't believe they can win, they won't bring it to us. And from a CAC and the size of the opportunity, that is one reason why we've been able to show success in 2023.

Perfect. Glad you hit on federal. With the last minute, I want to touch on margins.

Yes.

You guided for 15% EBITDA margins this year. You've seen like over 20% improvement over the last several years. What's really driving this? What should we expect on a go forward basis?

Yes. I'll keep this simple. In the 81% gross margin, business should be printing reasonably good free cash from the EBITDA margin. My life is simple. We're at 81% gross margin. G&A is at, let's call it 17% today. A company that gets to $250 million and $300 million, we are $200 million today. As you think about the next $50 million to $100 million that are going to come from ARR. G&A stays flat. You're just going to -- there's a gift that keeps giving, which is another 7% to 8% coming in operating leverage from G&A. If I just delivered that, you're getting close to 20% EBITDA margin. And we've demonstrated in the past year going from, I think 2%, 3% to 13% EBITDA margin. So people should believe me when I say that we will exit this year full year 15% EBITDA, which means that my exiting Q4 quarter should be higher than that to be able to deliver a full year EBITDA. But we will continue to invest in R&D and innovation, as you saw here. We will leave that at 18% of total revenue. We have a benefit because the majority of our organization, from an R&D perspective, is in Italy. So it will cost arbitrage, because of the salary comparisons relative to North America. Sales and marketing will remain our bigger investment around 32%, 33%. But it's pretty straightforward. Just some operating leverage in sales and marketing majority coming from G&A, you get to 18% to 20% in relatively short order.

So easy.

It should be, it should be.

Awesome.

Gross margin [indiscernible]

Enjoyed the conversation. We're out of time. Thank you, Sukaran and Giuseppe.

Thanks, Josh.

Thank you.