Thank you all for joining us at Elevate 2026, and thank you particularly for management putting on this investor Q&A session. I really appreciate it. I am Byron Stephen, Vice President of Investor Relations. To my left, I have Eliran Glazer, Chief Financial Officer of monday.com, Eran Zinman and Roy Mann, who are Co-Founders and Co-CEOs of monday.com, and Casey George, our CRO. This is an open format. I will remind everybody that we will be reporting our Q3 earnings here very shortly. We will try to keep this less focused on numbers and not talking about our upcoming Q3 results, but more on strategy and product as much as possible. Anybody who is listening to this on audio, we are live streaming this through our investor relations website.
If you have any questions, feel free to email us at ir@monday.com, and we will be happy to take those questions as well. We have two mic runners around here, and we will just open it wide for Q&A.
Great. Can you hear me okay?
Yeah, we can hear you, Steve.
Great. Steve Enders from Citi. Thanks for hosting this great event today. Maybe just want to ask on, a lot of focus on agents and trying to drive adoption. I guess, how are you thinking about the organic, getting people to adopt it that way versus needing a services element? And I guess, what are you doing with Frontier and the services model to enable that for your clients?
Yeah.
Thanks.
Thanks for the question. Look, I think I'll start first with generally speaking, feedback that we constantly hear from customers is that nobody's skeptical about AI. Everybody understands AI is a transformative technology. I think we're now in a phase of adoption where customers are saying, "We've been on the side of let's do token maxing, just buying AI for the sake of buying AI." And we're now transitioning into a phase in the industry where customers are saying, "We want to see return on investment. We want to see actual business impact, whether it's reducing cost or increasing the top-line revenues." And I think there's a big gap between the fact that technology exists to actually adopting it. We hear it across the board from customers who bought licenses to use all kind of AI software but saw marginal gain in terms of efficiency.
They understand, and the feedback we hear is that in order to drive a real change in their own company, in their own business, they need deep adoption and guidance more than anything. We heard that feedback. What we announced in our keynote earlier this morning is first that we're establishing a new organization called monday Frontier. It's basically a combination of managed services for deploy engineers. We're going to scale this massively going to 2027. We see a lot of demand from customers who basically say, "We want your experts to drive the change within monday, the deployment of monday." And what we also hear from them is that they want to do it on the platform because they have most of the data inside monday, the contacts. They have people already using the platform from across the company.
Something we emphasized during our keynote is that we want people and agents working together. Having people build agents in a siloed way doesn't help. In order to drive real change, you need people and agents, data and context all live in one place. So I think this is a big opportunity for us. This is the way companies are going to adopt AI going forward. And this is exactly what we're going to focus on as a company.
Hi, thanks so much for doing this, Matt Bullock, BofA. We hear a lot about context being critical from software vendors. It would be great if you could help us understand the context that lives inside the monday.com platform, how you guys differentiate versus a lot of peers in that context advantage, and then high level your strategy for API, MCP, and allowing third-party agents to tap into that context longer term.
Yeah. So I can take this. So yeah, it is critical. It's like a bunch of things. It's context, but also guardrails. Like access to data and governance. We've just been with the CAB. That's one of the number one concerns in enterprises. I have my data in here, but I really want to know who is accessing what, when, and to trust that. So that's another really critical element. And the context is work for us. You want the agent to do stuff, it knows what all the team did and your projects, your company, your role, what do you need to accomplish? It's already in there. So that's a critical part, and then it knows also to ask for other stuff that might be in other areas. And we see this as the magic part, right?
You ask, it knows already, and like Daniel shared on the stage, he said, "What do you need to do for it?" Says, "Nothing." And I think that's a killer part because it's not only that it's easy to use, it also allows us to create really good agents because we know before, and we can analyze all the data ahead of time, so the agent can also suggest stuff. It also knows it, and it's a better agent. It's not just like pulling the data. It's also having the right agent to do that stuff and look in the data in a certain way. But that's one part of it, and the other, I'll answer your question of external agents. We don't think we're going to be the only agent platform, and it's all-in-one.
There is obviously massive advantage for having a suite that covers everything like we are doing. But there's also a lot of power in connecting to anything else. So whether it's any other source of data that you can pull in and any other agent that you can pull in, and we're equally dedicated to opening up the platform, because again, like we said, the real value comes from having everything work together. When I write something to you and your agent jumps in the conversation, and whether it's on monday or not, and has that context or another context, you can imagine synergies happen, and it becomes like a multiplier. So we can also provide all of it if you're a smaller customer, probably. But for the larger enterprises, we're going to be a bigger let's say, piece of the puzzle of all their AI infra.
And maybe just to be more specific, today you can use monday through the API, through MCP, everything is open. You can use monday headless if you want, and also you can import any third-party agent, whether it is on Claude, ChatGPT, or even personal agents into the platform. It is going to be like any other user, and we do not distinguish between native-built agent within the platform to third-party agents.
Thank you. Arjun Bhatia with William Blair. Roy, Eran, your keynote, I think, highlighted change as a pretty important theme. It seems like there is a lot changing, obviously, in the environment, in the market, and then at the company also. Historically, we have always looked at monday through the lens of work management, CRM, dev, et cetera, service. Is that still the right way to think about the company, or does that need to evolve, and are you going to become sort of AI agent orchestration, and is there a new sort of framework or project product lens through which we should view the business going forward?
Yeah, it is a good question. I think it definitely evolved, I would say, specifically in the last year, in two ways. One, I think historically, monday was perceived as just a project management software. I would say more broadly, like work management, but very project-oriented sense. I think all the use cases you have seen today and from customers and also things Daniel highlighted in his own presentation, this evolved tremendously in the last few years. People today, what they do on monday is to manage critical business processes. Now, it is different kind of processes across the company that there is not like off-the-shelf software to do that. But because monday flexibility and nature of how you can use the product, companies from different industries manage critical business processes.
I think in a way, this is a big opportunity also in the age of AI, because as much as our goal as a company remains the same, which is to drive acceleration for businesses, improve their efficiency and revenue, I think the TAM has dramatically increased, not decreased. Because what now companies or our customers can do with the platform is much greater in terms of the business impact that they can generate on top of us. And just in terms of CRM and service, also with CRM and service, we think this way, so they have taken a more independent path of how they evolve, also put a lot of emphasis on agentic offering that we offer in each one of those verticals.
Yeah. From my perspective, you can look at it like work management has like it's always been a place to connect the organization collaboratively. You can think about it that it's expanded a lot with agents because what they do is very general with us, and they can connect, like customers can connect a lot of different aspects of the organization. You can argue that other agents do the same, and there, I think, the, let's say, historical quadrants is at question, right? Because the borders are kind of diffusing between areas, and that's also why it's important for us to first connect to other vendors. I think that's where the customers are, that they are asking themselves, "What do we want?" It's not as if AI was a technology that was born to solve the problem.
It's now a technology looking for things it can improve, and I think that's what customers are doing. With that, borders are kind of blurred in a way for the different boxes, and I think we and that's what I think Eran mentioned as an opportunity for us in a lot of ways because we're entrenched in the white space and connectivity of a lot of things. If we go to market well, we'll capture that area.
Hey, thank you, guys. DJ Hynes from Canaccord. There's tons of great AI ROI examples today in the presentation, whether it was growth, productivity, efficiency. To me, it seemed like a lot of the examples you cited skewed a little bit more SMB or departmental. A, is that the right read for where we are today? Then, B, what needs to be the unlock to take that into the enterprise, right? Is Frontier part of that?
Maybe I'll leave it there, and any observations you have of kind of where we are in that evolution.
Yeah. I can start. We actually tried to give a mixed use case study, so we had one from a 100-person company, one from a 10,000-person company. Actually, we are seeing way more adoption with our enterprise customers on AI agents. It is triple the adoption compared to SMBs. I think we have a larger portion of SMB in terms of percentage, of course, from the total customer count. But in terms of adoption, willingness to use, willing to spend, we see much more potential in the enterprise segment. Also our strategy is very much focused on continuing to go up market and focus on enterprise customers. I think in a way, and Roy just mentioned, we just met with a customer advisory board, speaking a lot with enterprise customers.
They are actually thinking about monday as a potential solution to how they can adopt AI, because their biggest fear is that if we give AI all the freedom in the world without guardrails, it can hurt their business, and there is a lot of concern around it. Because we built our agents and our AI offering on the monday platform, which is enterprise safe and secure, they have full control. They know exactly what data is in it.
I feel they feel much more confident, and I think monday Frontier will accelerate that, because rather than saying, "We trust the brand, we trust the company, we trust the product, people love the product, I need your help with those experts to find those use cases that are going to make an impact on my business." I think this is our biggest opportunity, and this is the biggest segment we are going to go after in the next year.
The only thing I would add there is we have to be successful at both. The product organization is hyper-focused on building what I would call accelerators. As an SMB customer, if you can give them 80%, they can get all the way home with the rest of the agent. That is one of the things we are hearing loud and clear, and as Eran just mentioned, as we move up market, it is very clear and our customers are screaming it, "Please help me. I don't know where to start. I don't know how to do this. I need someone to ask the right questions." That little bit has surprised us. I would say that the enterprise has been such a huge growth opportunity for us, for agents. But we got to do both.
Yeah. Hi, thank you guys so much for the time. Taylor McGinnis with UBS. Two-part question on Frontier, because that seems to be the big announcement coming out of today. The first is, this seems like a bit of a change in strategy for you guys, because if you think about monday historically, it has been very easy to get up and running. So one, could you give us a sense of how big of a lift that might be for you guys, just in terms of costs and bringing all of that hiring in? Then secondly, to the extent that you guys have been trialing Frontier with some customers, I would love to just hear early feedback on what you guys are seeing with ROI and use cases being unlocked and payback periods. Just anything that you guys could give us a sense for what this opportunity could become.
Yeah. I will start.
Go ahead.
You guys can continue. Just one point that is important for me to highlight, it is still the core of monday to make things easy and simple. I think you saw that in the presentation. It is critical for us, the Frontier part, is to help them think about the business, the value, and not about using it. It is about because then we will build something, and they will own it. It is about holding their hand and helping them and be consultative in that respect, so yeah.
Yeah. I think what Roy is saying is directly correct. I think the problem is not the user interface or the easiness of use. I think this is something we know very well how to do. I have a lot of confidence on that front. I think the biggest problem is to identify which areas in your business can be improved with AI, and then how do you drive the change? Because it is usually, from our experience, it is not one thing you need to change, and there are multiple people involved, multiple risks involved, and it is very complex. Look, we have been very focused, working closely with our customers in the last six months. Like I said, they love the brand. They love the product. They have good usage and already a lot of data in it. It is a big advantage.
I think one of the biggest issues right now with companies trying to implement AI is they go to market. It is a very noisy market. Everybody is saying the same message, and people do not know. They feel fatigued about all the messaging they hear around them, and eventually, they want to go with brands they trust. So I think it really plays to our benefit. We have seen a lot of success so far. This gives us the confidence to scale monday Frontier going forward, to do more implementation like this. So we are going to hire for that this year, next year. We are going to put a lot of focus on that and offer it to majority of existing and new deals. Yeah.
Great. I have one question online that has been submitted. It is going to be for Eliran. Without discussing near-term guidance, if growth remains below historical levels, how would you adjust spending to protect free cash flow?
We continue to operate, as we said in Q2 when we disclosed the numbers, we continue to operate in a way that is efficient, meaning we are looking at growth, but on the same time, we are looking at our bottom line, continue to be efficient on the way we manage post-RIF, and we did the RIF, and we announced it. And we said that we had the growth savings of potentially $100 million that we are going to invest the majority of it, but we are still looking at our cost components such as AI cost, such as investment in headcount. So we keep the balance of investing, but also looking at the efficiency metrics that we always looked at.
Hey guys, Allan Verkhovski with BTIG. Congrats on a great event. The keynote energy was phenomenal. I want to ask a follow-up question to one of the things you were talking about earlier, which is around your API limits. I think I heard that customers are now getting notified when they exceed their daily API limit, and you'll be monetizing API add-ons next year. Can you talk about the activity and factors you saw that led you to begin enforcing these limits? How meaningful can this be? Then lastly, can you talk about how the user mix has trended between monday.com users that are exclusively using your MCP versus broader AI capabilities inside of monday? Thank you.
Yeah. First of all, we always had limits to our API per plan, so that haven't changed. Basically, when we think about pricing strategy, and this is something we announced a few quarters ago, we want to lean more of a hybrid model of seats and consumption. We unified our credit system. Whether you use monday vibe or Agents or AI columns within the board, or AI workflows, it's going to take the credit from one pool of credit, and we're going to allow a customer not just pay for seats, but have consumption model in addition to that. Over time, I think we'll introduce more and more of those capabilities into the product as we see that customers are leveraging the platform in different ways, and want to allow them to scale regardless of their headcount.
Again, over time, I think it will shift, but I think we're very focused on a hybrid model right now. We kind of might see more features like that, coming into play.
Great. Hi, Elizabeth Porter from Morgan Stanley. Thank you so much for today. I wanted to follow up on the idea of the different quadrants and how AI's blurring some of the lines between these. My question is, does the shared AI layer change some of the cross-sell opportunity across work management, CRM, service, and dev? What are you seeing from customers as it relates to AI driving a willingness to adopt more products versus just monetizing more of those existing deployments?
We continue with a multi-product strategy, and all of our AI capabilities and our assets span across all of our products. That's definitely an integrated point that our customers really like. So they can leverage what they build in work management and also leverage that in CRM and service, et cetera. That obviously helps us as we cross-sell. The second part of your question was, remind me.
[inaudible]
Yeah. Obviously, if they can use some of these assets across multiple products, obviously their utilization goes up. So it is a self-fulfilling prophecy in a lot of ways, and we are starting to see that. We did accelerate some of our capabilities in CRM. Some of the agents that we built that were specifically for CRM came online sooner than our AI Workspace. So those things have helped us in the reverse with our AI Workspace platform. So it's again, a self-fulfilling prophecy.
If I'll use Daniel's metaphor from his session about batteries being included. I think work management broadly is a more diverse product. Every customer is using it in a different way. I think the benefit of both CRM and service specifically is that we can build very specific agents and AI use cases that are built specifically for CRM and service, so it's a little bit easier to adopt for our customers. And there's a lot of potential with those specific agents as well. So it's a little bit of a different use case. I feel monday Frontier will be, let's say, more needed for the work management part, where CRM and service, again, this is just our intuition, is going to be more easy to adopt using the built-in agents that we offer.
Hi, Mark Schappel with Loop Capital. Regarding the rebranding strategy, I was wondering if you could just talk a little bit about what specific commercial problem the rebrand is intended to solve here. Would it be limited enterprise awareness, the perception that you're just a work management tool? Maybe just talk a little bit more about the rebrand strategy.
Yeah. First of all, we feel we've been with the monday.com logo for a long time. This is after we branded from our previous name, which was dapulse. And we're very excited for this change. It's not a major change as it was back then, but I think we feel that the company strategy have evolved since then, and the offering we can offer to our customers much greater than what it was. We've seen that from the use cases. We've seen that from the feedback we get from customers. I think in a way, it will open up people to listen more to our story and kind of rethink what monday.com can be for them. I think now with everything going on in the market, there's a lot of noise.
I think it's a method for us to stand up a little bit above the noise, tell the monday story again. I think for a lot of customers that heard about monday in the past, it's still perceived maybe as a limited work management or project management. I think in a way, the rebrand will help us reposition monday as the AI workspace for teams, open up new messaging and new ways to communicate to customers. They can do much more with the platform today. Also talk about the fact we're going to be there to help them solve, and maybe find opportunities to leverage AI more. So I think it's more of a facelift, a way to tell a story again, open up to new customers, and continue going up market, the way we've done in the past few years.
I have one more question online. I think this one's probably more for Casey. What most often limits expansion within enterprise customers, and what are you changing to address it?
I would say, a lot of our opportunities as we move upmarket, as you can imagine, these enterprise customers have multiple platforms, right? There's constantly rationalization going on with their vendors, and this is where we shine a lot of times. We typically win in these instances. What typically happens, and it slows us down sometimes, is they've got to organize whatever they have today and figure out how they're going to move it to us. To the extent we can help them accelerate that with our services to migrate over, that's great. As you can imagine, they build a lot of workflows and a lot of different platforms, and it's tightly integrated into what is a pretty complex infrastructure. That's the one thing that slows us down sometimes.
I don't think that's unique to us, but we do see that a lot, and we see it a lot because a lot of our expansion opportunity upmarket is taking out the competitor.
Thank you, guys. This is Ivan Radojicic from Wolfe Research. Maybe one question on the voice AI agent, Harmony, that you launched, which is obviously pretty exciting, and it's one of the highest growing markets out there, but also one of the most competitive ones with a lot of AI startups competing in there. Maybe what are you seeing there with early adoption and early feedback? Then two, what do you think will be your right to win and what differentiates you there versus other products in the market? Is it specific use cases or anything around the models? What is sort of the right to win there?
Yeah. It started with us making a decision that voice is a strategic path for us or something, a capability we'd like the platform to have. So we acquired OneAI, the startup, and that's the seed that grew Harmony. Essentially, the product has a lot of unique capabilities. We build our own models, really small, lightweight models that identify a lot of stuff really needed to have high-quality conversations and high-quality voice conversations. Just give an example. When someone answers, identify it's a voicemail and not start talking into the voicemail as if it was a real person. So we're really good at that and a lot of other stuff.
We do have some advantages right now that we are winning over competitors in quality, but I think it is more strategically something we need to own and to incorporate in a lot of other solutions, especially with the Frontier that we see a lot of the use cases. If you pull voice in, it kind of solves for, opens up the door to a lot of other use cases and connectivity throughout the organization.
I will just add to that two things. One, it is a separate website, separate go-to-market, Harmony.ai. So we separated that from the platform. It is a completely separate product. One of our assets as a company, we have over 200,000 customers, so again, the go-to-market right now is very fuzzy. A lot of our customers express interest in voice agents, so I think it is an opportunity to sell that to our customer base. Also what we have seen from customers is a lot of the times that a voice call from their own customers become a process later on within the company. I feel there is a big chance of one plus one equals three here. Integration with monday, each call making it a workflow. So we also see some integration there. They have their own go-to-market, and they are approaching our existing customers.
They have a booth here at Elevate, and we see a lot of interest from our existing customer base.
I will add one thing here as well, and I think those of you who know this industry and market with contact centers, I think everyone would agree it is probably the most ripe industry for disruption. This platform is not an SMB platform necessarily. It scales all the way up to the largest customers. So we can support and go all the way up to the top of the market, and as I just said, in a market that is ripe for disruption, and this stack that they have built and this platform is second to none. So it is a huge opportunity. We are super excited about it, and I am even more excited about the fact that we can continue to support our upmarket motion.
Yeah. Having said that, it is a startup. So we're just starting, and we'll see how it goes. Yeah.
Thanks for doing this, guys. Mashu Nishi from JPMorgan. Casey, you touched on this earlier, but with the significant portion of your ARR still from the sub-$50,000 customer cohort, how much are you going to continue or start to offer Frontier to your lower-scale customers? On that point, we've been hearing from channel partners that maybe the enterprise customers are receiving a little bit more attention in terms of standing up monday's AI products. Do you see that as true, and how much are you going to invest into your base customer cohort?
The timing of your question is very good. We just had a couple strategic partner meetings, and what I would tell you is our FDE motion is, for the most part, in place to support our mid-market, up-market customers, right? Enterprise. Obviously, we could not scale that for our 50,000 SMB customers. We are leaning on and building a program and a model where our partner ecosystem supports those customers with pre-built assets, obviously services. This is what they do, and it's what they do very well, and they're very interested in this space and building out a set of resources that can go deliver for those customers. So we're doing both, but through a different channel, and our FDE motion really there to support our up-market customers.
Yeah. Again, going up market is one of our most strategic things we're focused on in the company. When we went public in 2021, I don't remember the exact number, but I think we have a single-digit percentage of accounts that are $50,000 or above. We've been gradually replacing our customer base, now over 40%, and we'll continue to grow that segment going forward. It's a strategic direction for us as a company. I mentioned this earlier to a previous question. We're seeing much more adoption of AI agents within our enterprise segment, three times compared to SMB, relatively speaking. So, we're very focused on that. Frontier is going to focus on that mid-market and above. We see great adoption and willingness to adopt more AI, and the ability to pay for that. So definitely, more kind of focus on where we're going and where as a company.
I have another question online for Eliran. How do you determine how much cash to retain versus return to shareholders?
As a reminder, we did the buyback that we completed in Q3. It was an $870 million program that we announced. When we did the analysis, what would be the amount that we take is appropriate for us in terms of capital allocation? We said, okay, what it would be the requirements of cash in order to run the business, just an example, for six months without interruption. What would be the amount of cash we need for potentially M&A? With regards to M&A, we said that we are going to look at tens of millions and not hundreds of millions. This is what the strategy that we defined. The third thing was looking at shareholder return by way of shares buybacks that we did. This is the way we also think going forward, in terms of the strategy and our execution plan.
We are not thinking currently about any additional buyback. This is something that, if in the future we want to think about it, then we have the flexibility. The main purpose or uses of our cash is running the business and keeping some cash for potential M&A.
Great.
Thank you. Prashik from 8th Wonder Fund. My question is, for enterprise moment, how much of enterprise growth you guys are thinking from coming from having more agents and seat usage, versus more product penetration? A second part of that is kind of a dumb question, but I like asking a dumb question. How do you define a true enterprise? For an example, if I have 200, 300 seat customer, but I have a lot of your products, so I could easily be spending $100,000 with you. So what is true enterprise customer for you guys?
Well, we identify enterprise anything over 1,000 seats. Is that right, Paige? Yeah, 1,000.
1,500.
1,500.
1,500 now.
Okay. Thank you. The definition has changed, but 1,500 is the amount that we identify for an enterprise customer. Do not get that confused with enterprise tier. That is our highest tier. But if you see any stats where we break it out by enterprise, mid-market, SMB, enterprise is over 1,500 employees.
There is one caveat there. If we do see a customer that has the potential based on how they are using the platform and they have growth potential, we do treat them as we would a customer that has 1,500+ seats. We have the ability to do that, and we do that on a regular basis.
I would add one more thing that changed now because of the new offering. Let's say if we had a 1,000 people account and we had wall-to-wall deployment, like 1,000 people on the enterprise, we felt like we kept the ability to expand. Now, with the hybrid model, the fact we have consumption, we see this as a potential to expand. So it's no longer like an account that we kind of reach its limit. So we try to upsell it like agents and vibe. So I think it's opened up a new vector for expansion that was previously limited by the amount of people within the enterprise, within the company. Now with the consumption, the hybrid model, it opens up new opportunity for expansion. So you might see like a mid-market account paying equivalent annual contract that an enterprise could only pay previously.
Howard, do you have a question?
Thanks. Howard Ma with Guggenheim Securities. For what it's worth, I don't think dapulse is such a bad name because you could have done a play on words with the heartbeat of the
It was the dapulse, so it's not. Anyway.
Anyways, I digress. My question is, have you guys done an analysis on the incremental usage that your platform scalability investments unlock, and how much of that usage you believe will take the form of agents? Eran, you mentioned pricing earlier, pricing packaging. Could you see that evolving into more agentic tiers?
Yeah. So maybe I'll start with the pricing part. Look, I think everybody's trying to figure what's the best way to price AI offering. I think we see this already. I think it's a larger problem than just how monday.com is pricing its product because you've seen also the Frontier labs, when you price per token and you manage to reduce the token cost, what happens to the revenue and how this affects the run rate of those businesses. First of all, with the Frontier, we try to think about more of a solution that we can tailor to our customers as opposed to just price it on tokens and usage. So we're thinking about also new potential pricing strategies. What I'm trying to say is, it's a moving target.
I think we'll evolve a lot over the next year or two and learn what customers are looking for in terms of predictability, attaching price to results, and we'll adjust it along the way. I think what's important to emphasize, which I said earlier, is that it's no longer just a seats model, it's a hybrid model, and we'll continue to expand it as we move forward. Just in the first part of your question, sorry, I forgot the first part.
It was measuring the incremental usage, that scalability. Some of your customers say they use a lot more.
Yeah. We heard that a lot today. Yeah.
It's like a lot of them see that as opening the door for leaning 100% on us. Some of them have maybe a few vendors, and they want to say, "We want to go all in on you," and that's an unlock for a lot of those stuff. We see it also as this critical path to agent usage and AI usage. Why? Because there will be so much more information in the system because of agents. We need the scale, we need the speed, we need all of that. For us, it's like the same investment, both in enterprise and in the future of AI on the platform.
Great. As a reminder, if you're listening online, please feel free to submit any questions to ir@monday.com, and I'll be happy to ask those questions. I do have one for Casey. What are the biggest opportunities to improve rep productivity, and what needs to change to realize them?
Yeah, so very proud of the direction of our rep productivity. We've done a lot with our RevOps organization to build in processes that streamline a lot of the things they do, right? Obviously, our reps want to spend more time doing what they love, which is selling. If we can remove as much of that as possible, it's a big help for us. One thing that we did, and you heard this in the keynote, you've heard this a couple of times, and I think this is the best thing we could have done as a company, definitely since my tenure started. We had Agentic Week, right? We unlocked the power of the platform. We handed it to every individual in the entire company, and we said, "Hey, go build. We're going to take the week off. We're going to go build assets.
You come up with whatever you want. We gave them no guardrails whatsoever. And you would be amazed at the creativity and the things they thought of. And of course, a lot of the sales reps came up with processes and built AI capabilities that removed a lot of the administrative things they do, some of the things that they do not want to do again, they want to go sell. And so we created a ton of productivity or capacity in our system by leveraging our own platform. The other thing I would tell you, and I can say this, I have worked at and been a leader at a number of technology companies, some of the best in the world. I have never been at a company where everyone was a super user of what you were selling, right? So we all love the platform.
We have used it before AI, and now with AI, it just exploded. So our productivity on the back of us all using our existing platform has just exploded. So we continue to keep a close eye on that. We no longer have to grow our sales headcount in a linear fashion to support our growth. We are growing our headcount, but we do not have to do it in the same linear fashion as we have done in the past because our productivity has increased so much.
Yeah. It is also we did something really interesting. We connected the platform to BigBrain, to our data source. And so it is like everyone out of the bat have access to their data and all their sales data and everything. So they pulled up their own agents, reports, whatever they wanted, and it is also like the previous question of context. That has been a massive unlock of context for the stuff we build internally, that we connected our own data lake into the platform. And it is a bespoke connection, but obviously any one of our customers, we are going to help them do that as well.
Hey. Ryan MacWilliams, Wells Fargo. A lot of talk about the fusion of AI through the economy, and I am happy to say that my uncle, who is a plumber in Philly, called me today and said, "We got to talk about this AI thing." And I was like, "Okay, we have broken containment." And he wanted to talk about voice, which was interesting. But I guess I have a two-part question for Eliran here. Just a lot has changed about how we think about AI, but how people are adopting it. So I know you are not giving financial targets today, but has AI usage or investment maybe changing the ways you think about the financial components of monday's business could be going forward?
And then for Eliran, for those customers that have been on the front foot of adopting AI throughout their organization, whether through monday or larger AI labs or different AI features or implementations, how have you seen those more AI-forward customers change their monday implementation?
I will start. So with regards to internally, how we use AI. So one and a half years ago, or maybe 14 months, we hired five people in the finance team. None of them were the financial person. They were project managers, they were developers, and what we did is we established a team that everything that they do is just they have to look at all the processes in finance now, also outside of finance, and automate the processes.
There is a belief for some reason that you will get immediate ROI. I spoke with one of the investors here earlier, and I said that I'm attending all kinds of webinars. And people are saying we're using AI, and then there is a question, how many people did you let go because of AI? And no one did. How many people you didn't hire because of AI? It start to do something. So I think, for us, there is a lot of work that we are doing, but currently it's not reducing work. It's actually replacing. I think there is a transition period of about, I would say, 18 to 24 months in which we are going to complete.
And people are having now more time to do things that are more strategic, or they can be more knowledgeable about other things that working with other parts of the organization, replacing repetitive actions like, I will give examples, customers billings matching with invoices. And all kind of manual things with regards to budgeting, with regards to free cash flow modeling, with regards to the modeling of the business itself. We now ran, I will give an example, I'm proud of it. We worked on this for the past six months. We took an Excel spreadsheet, and we basically that was everything, the Bible of monday, and we now run it in different AI tools in order to, like a statistical model, a business model. It takes a lot of time. So I would say by way of efficiency, people are doing much more with AI tools.
There is a lot of work still involved. Over time, I think we're going to see the productivity gains coming very soon, but already we're changing the landscape of how we operate in monday.
That's the finance team, right?
The finance team.
I can share stuff we're doing in the R&D team. It's like about 30% of the code is now being written automatically by AI. I don't like all 100% of the code is being written by AI, but it's developers. Like 30% is just like automatic, like bugs coming in and cycle through, like without anyone touching it. It helps there because we have the infrastructure to validate quality. We had it all set up before. But in other areas, like Eliran said, we're building it and also for our customers.
One anecdote, Ran told me about a year and a half ago. He said, "One day we'll sit in an earning call and you and seven agents.
I wanted to do it before.
I said, "I will never, you know, never going to do it." No, it's the agents with
We need to understand the legal aspects of that.
I think it was illegal.
Yeah.
Legal hurdles. Not the IR hurdles.
Yes. Just on your second part of the question about, are we seeing enterprise adopt AI, the ones that actually kind of more forward-leaning. Look, definitely some customers are more forward-leaning and more experimental with AI. Look, for SMBs, it's easier. But also they're less willing to pay, I would say. Because the impact is not as great as enterprise companies. With enterprise companies, we see some, but as I've said, we see a lot of them need guidance and help with adoption. This is why we put so much focus on monday Frontier, and we tend to double down on that. I will say we did see some customers build incredible things. The way we think about this is also how we can increase usage. If you attended Daniel's session, one feature we introduced recently was Agent Directory.
Because a lot of our customers says, "We built amazing agents. We want to share it with the rest of the company." Some customers build an agent that anybody in the company can ask it questions about the business. They connect it to the data source. Some build an agent that can process receipts, and a lot of other things. So we're trying to think about what kind of features can we build for enterprise customers to drive virality and to increase agent adoption. The same way we've done for other features we've developed. This is one example, and we plan to launch more of these. So definitely thinking about how the product can encourage more people to create agents, to use agents other people are using. If you think about it, this is a big problem right now in companies.
People, individuals might be very proficient, they might use Claude or ChatGPT, any other vendor. When it comes to actually sharing what they've been building with other people, making it official, who's taking responsibility for that? For us, every agent we build, we have a human owner that's responsible. AI is a great technology, but it will never take responsibility for mistakes. You can trust me, that will never going to happen. We need to be accountable. Our customers are accountable for the results, and they want to track what the agents are doing. They want to have a log of what's going on. The way we think about our product is exactly that. How we can give them the confidence of what they're doing, how we can drive virality, and make this amazing technology available to everybody in the organization.
I was wondering if you could spend a minute talking about any trends that are different or at high level, if it needs to be on the SEM side. I think maybe more importantly, looking forward, what SEM was 10 years ago and what you built with BigBrain should probably be pretty useful for like agent-led growth. Agents who are deciding whether they're going to use a product management tool to handle themselves and choosing monday.
Alongside of that is obviously the GEO stuff, but I'm just curious whether there's any effort inside the company or thought or you're seeing any early results from what I'm sure is work to do to surface in those places the same way you've surfaced really efficiently in SEM.
Yeah, definitely. First of all, we have deep partnerships with Anthropic and ChatGPT. We pair the one of top vendors that they recommend in terms of integrations. We see our MCP usage and discovery exponentially growing month-over-month. Definitely, it's a new way to discover the product. We see very good conversion from that. At the same time, last we disclosed, I think over a year ago, we've seen the pressure on the top of funnel. I think the industry is going through a shift, where previously it was about displaying ads for people doing search or surfing online. We're shifting now to more decision-making and agent's recommendation. I think over time it will shift more towards that. We're very focused on that. We have teams working on that, on those partnerships and discovery. We'll see. It didn't offset the pressure we felt previously.
But I think things will eventually come into equilibrium of discovery, and of course, we'll do whatever in our power in our BigBrain engine to seize every opportunity that presents itself.
We are seeing internal AI opportunities like you mentioned, like agents optimize stuff automatically and also launch campaigns multilingual right off the bat, stuff we couldn't do before, like create 100 ads and not one in almost no time. So those stuff we're rolling it out now and also closing the loop. So I think we're going to see some gains from it. But again, like Eran said, nothing changed from what we reported before on the pressure in that area.
Thank you. Ivan Feinseth, Tigress Financial Partners. Going back to Harmony, how do you see the product developing to gain the ability to increasingly interact, and where do you see deployments, let's say, like in apps eventually? Even as early as next year, for example, phones are going to go from a regular operating system and using apps to a voice-activated interface that then the operating system drives the agents that interact and drive the transactions, for example. I think that's going to be a huge opportunity for voice interfaces, as well as the growing demand for voice interfaces in all levels of customer service.
So the question was. Sorry, Eran. Are you good?
Yeah. I can start if you want. Yeah. Harmony is a very interesting opportunity. I will just say that Harmony's product can do both things. It can do outbound calling, so for example, calling sales prospects and customers, and also a very popular use case is inbound calling. So customers calling into the business. It can take over the business phone number and route calls and basically work against the script of what you want it to do. I think over time, what we are going to see is an explosion of agents doing calls, agents calling one another and scheduling appointments and acquiring businesses. I am not sure what is the end game of that.
I just feel customers' expectations in terms of the level of service they are going to expect, how fast they are going to get a response is going to dramatically change, and businesses that will not adopt such a tool will be left behind. So I feel we are going to see an explosion in that field. Like we have said, this is strategic for us. We acquired OneAI, we built it, packaged it into Harmony. We plan to offer that to our customers. We feel it is a big opportunity. We focus more on the mid-market enterprise segment, and not the SMB segment on that product. Again, I feel it has a great potential.
We have time for one more quick question.
Awesome. Ian McHugh with Jefferies. I guess I am wondering just on the shift from the seat base to the seat base plus usage model, just broadly what inning we are in there. Any color on number of customers that are using usage, if they are topping up their tokens, and then if there has been any pushback by customers on the new model. Thanks.
Yeah. Overall, the tool was very well received within our customer base. We are introducing improvements and adjustments as we move forward. A lot of the features we've built in the last few months is to provide more visibility into the way they're consuming their credits. That was a big request from our customers. In the admin section, actually, there's a whole booth here just for the admin section. It's pretty wild. People are really excited about that. We have a full report of who consumed what per month over time. We're also trying to experiment with credits, so give personal users credits and then overall company pool credits. We'll adjust it over time. Overall, it was very well received. I think it's becoming more of an industry standard going forward. Over time, we'll learn how to properly improve the way we package it to customers.
But definitely customers' reception for that and willingness to pay have changed in the past year.
Yeah. But we still get the same feedback everyone gets, that they want predictability, okay? Which is hard to give them because it depends on usage, but that's like a shift we see for everyone, okay? Moving from a super predictable model to a consumption-based one. But within that boundaries, where they understand it and we're solving for a lot of those issues, like Eran said.
Great. With that concludes our time. I appreciate management for making time for our investors, and I want to thank the investors, particularly for some of you that have traveled to be here. We really appreciate it. For those of you in the room, we do have some swag in the back if you're interested, and some additional snacks as well. Thank you once again, and hopefully we'll get to see you at the rest of the Elevate sessions. Thank you.
Thank you.