Hey, thanks, everyone. Good afternoon. Welcome to day one of the Wells Fargo TMT Summit. I'm Michael Turrin, software analyst. Very pleased to have a couple of gentlemen from Five9. We have Barry Zwarenstein, CFO of the company, Dan Burkland, President. Post the rock and roll, and Barry would like to kick us off with a little bit of a safe harbor statement-
Thank you.
We'll get started.
Thank you, Michael. So before we start, I remind you that, we'll make forward-looking statements today, that, regarding future events, trends, expectations, projections that may affect our industry and our company, product developments, AI and automation, potential growth drivers. They're predictions and should not be unduly relied upon by investors, and may differ materially, and we undertake no obligation to update them. Please refer to our 10-Qs and 10-Ks with the SEC for further explanation of what, we call these things to change. Thank you.
Great. With that, let's get into it. Dan, maybe we'll, we'll start with you on just-
Sure
... the market backdrop. We've spent time over the years talking about this. It's been steady. There have been some kind of steeper slopes over the past couple of years, but maybe you can just level set with the puts and takes of what you're hearing from customers, the macro conversation, where we are, and kinda the large contact center transformation.
Sure.
Yeah.
Thanks, Michael. Just to level set for those who may not have the history of the space, we're a contact center as a service provider, delivering customer experience for brands of all sizes around the world. It's a massive TAM. We're dealing with an approach that right now has only penetrated, meaning, moved to the cloud, only 20% of the market has made that transition. Yet it's been validated that eventually, the whole market needs to make that move. What we're seeing is a couple trends occurring. One is that the legacy providers that have delivered on-premise solutions for decades have either discontinued those platforms or decided that they're not gonna invest any more development on those platforms. So it's kinda pushing folks to the cloud, like many software industries.
We're also seeing a pull, with AI and automation, being at the forefront, and having enterprises realize that to take advantage of the compelling ROI that AI and automation delivers, that they need to make that transition to the cloud. So to your point, Michael, what we've seen is, an inflection of interest and demand, on the front end of the funnel, to make that move. We've seen an increase, 66% year-over-year in RFPs. We've seen a doubling of our pipeline at the high end of the enterprise, what we call strategic accounts or the whales or mega deals that we talk about. That's that side is extremely healthy and only getting better as time moves, especially as the end of life of the other products.
And then, the install base, some of the puts and takes. The install base has faced some headwinds when you look at certain verticals like consumer discretionary, that's slowed over the last several quarters. The growth rates from that install base have not grown at the pace that they once did, and they grew dramatically during COVID, and then, following that, they did so as well. But, we've seen some of that macro impact the install base, growth. But, that's kinda the, the nature of our business, and, again, the, all the discussion is around AI and automation and how, how that can take advantage, or how enterprises can take advantage of that and, deliver a, a better customer experience to their consumers.
Yeah. That's a great overview. I'm an analyst, and you put a number out there, so I have to ask the question on the RFP volume.
Yeah.
So 66% increase, is that at all a function of different types of deals, or is that just more activity because of the market backdrop, more intense focus on AI? What's, what's driving the RFP up there?
Yeah. That, that RFP increase, it was 66% year-over-year and 21% sequential from Q2 to Q3.
Wow!
One of the reasons was... Several reasons. One is, and that's our enterprise business, that typically put out the, the RFPs. The industry had to evolve and mature, and we had to get to a point where we were able to deliver scale, reliability, secure, emulate, and deliver all the functionality they've had on their legacy platforms. And even for a while after that occurred, large enterprises said, "Well, I'm not gonna disrupt my business for two years just to get like for like in the cloud." And then came AI and automation, and that was really the pull factor, and they're realizing in order to take advantage of that, they need to make the transition.
There's quite a window of time that it takes to go through a sales process and then an implementation process to where they reap the benefits of those solutions. So lots of companies are embarking on that journey now. So the inflection really was the acknowledgment from some of the largest legacy providers to say-
Mm-hmm
... "These are end of life. You have to move to the cloud," whether it's with them or whether it's to something new.
Mm.
A lot of them go out to bid because of that.
Yeah, that makes sense.
Uh, yeah.
So if we compare and contrast that with what happened during COVID, which was, I can't come into the cubicle-
Yeah
to do my job, therefore, I need to enable a remote agent footprint that I hadn't-
Yeah
-thought about. That was... Can you compare and contrast-
Yeah
what you're seeing today versus
I-
What you saw there?
Really appreciate, Michael, you bringing that subject up, because people made the assumption that companies ran out and bought cloud contact center-
Yeah
... to accommodate those agents.
Yeah.
There wasn't time to do that.... Right? The shutdown happened so quickly that they had to make do with what they had. Now, with a solution like Five9, it was very easy for them to send agents home and just have them log in from there. 'Cause, you know, you need a PC with a headset and an internet connection, and you can have calls delivered to you. So it was very simple for customers on cloud platforms. The legacy premises-based solutions took time, but there's still ways that they were able to deliver that.
Mm-hmm.
The big increase we saw, and why growth rates went into the 40% and even 50% year-over-year, was because many of the brands that we serve, you know, if they're retail or consumer products, shut their doors and went fully e-commerce, fully online, and the traffic that they saw coming into their contact centers exploded.
Mm-hmm.
We saw a big increase in volumes. You can do two things: You can sit with the agent counts and the seat counts you have and have queue times just build up to hours, or you can hire more people, buy more licenses, and accommodate that growth. That's what we saw during that time-
Yep
... was just an increase in volumes that went into contact centers.
Yeah.
As things opened back up, they went back to the stores.
Yeah
... so to speak, as one example. And that growth slowed. Didn't go down, but it slowed.
Yep.
Yep.
Very helpful. So I'm gonna let the two of you choose who goes first with this one. But the AI conversation has been a mixed bag in this market, which has been a bit surprising to me because you've been selling AI for longer than many of the other vendors who are telling us about AI enablement today, right? This was not ChatGPT enabled.
Right.
You've had automation technologies and, other things that you've been building towards and bringing to market for longer.
Mm-hmm.
So what is your sense of the mischaracterization, if you will, and just as far as-
Yeah.
You, you gentlemen, like, I can ask about the financials and how this changes some of the implications there. We can talk about the monetization difference, if there's any, between virtual and in-person, but I'd just love to kind of hear your high-level response to questions around the implications of AI.
Yeah
... for the CCaaS market in Five9.
Yeah. I'll start it, and I'll let Barry continue. But if you think we've been out touting that this industry is gonna go through rapid change, we've been doing that for roughly five years. We went out and actually, in anticipation of that, made a couple strategic acquisitions. Inference, arguably the leading IVA company in the world, we acquired them several years ago. We acquired Whendu, which is a workflow automation solution. So we hit the ground running and really pioneered the AI and automation story that was gonna compellingly change how brands delivered customer experience. And so we think we're in the pole position there. We also took a very unique approach to how we deliver it.
If you look across most of the landscape of our competitors, we're all checking the same boxes and have the same set of applications. We have eight modules now in our AI and automation portfolio. We all check those boxes. The question is, how do you do it? And we took an approach to say: We're gonna leverage certain hyperscalers underlying technologies. We're not gonna hire hundreds of data scientists and try to go out and build speech engines or natural language processing engines. We're gonna leverage those-
Mm-hmm
... and provide those to apply that technology to the customer experience. And so what we see is more and more companies are. We're infusing AI throughout the journey of the customer, whether it's before the call gets routed, and we're providing more intelligence to the routing, identifying their intent, authenticating the caller, finding the right resource for them. And then once the transaction starts occurring with the interaction with the agent, we can listen to the call and, simultaneously transcribe it, use NLU to then fetch data, deliver it back to the agent to make them more efficient and effective in their conversation. And then even post-call, deliver a text message thanking them for the order, "It's gonna ship on Tuesday. Here's a survey." So we can apply automation throughout the life of that journey.
The key there is, there's a misnomer that happened for several quarters was, oh, the... You know, people get on the media and say: AI is gonna replace agents, and all these, all these seats are gonna go away, and the contact centers are gonna become fully automated. If you speak to the customers and the brands, they're, they're looking to deflect certain very rudimentary, highly repetitive, transactional questions. You know, "Where's my package?" You know, "Give us a tracking number." "Okay, well, I can do that on the website, I can do that on my smartphone, or I can do that now on using my, my phone and calling." So it's a, another convenience. There's other, you know, "Where's this? What's my balance?" Very simple questions, many of which can be done via other self-service devices as well.
So what we look for is companies want to do more with less people, and we're – that's been the mantra for 30 years that I've been in the industry, is always trying to help them do more with less people. And so if they can deflect even a small percentage, 2%, 3%, 4% of their calls to be fully automated, hallelujah. Most of the automation that we're delivering them is to help the agents be more effective in their jobs and be able to be more... deliver a better customer experience. So when you look at the kind of misnomer or what's happened with the AI, we're a huge beneficiary, 'cause if...
When we provide transactions straight to an agent, and the agent's using our software to interact with the customer, we get $200 per seat per month.
Mm-hmm.
When you add automation to that equation, whether it's during the call, pre-call, or post-call, you're adding to that 200.... If it gets all the way to a fully automated call, it's 400. So we're, we're beneficiaries. As the more AI and automation you can insert into a organization and how they deliver customer experience, the more wallet share we, we receive.
Yeah, that's a good... So the monetization-
Mm-hmm.
-changes, but in a favorable way. If I think about the economic profile of the customer, Barry, is that changing at all?
Yeah, well, let me just emphasize the point over here, is that there's a tremendous benefit to our customers who are adopting AI and automation solution, as Dan described, the various forms before, during, and after the call, and managing the contact center. Because they're paying 10x what we are charging them. That's, we charge for an ordinary live seat about $200 per seat per month, but double that for a port. And we both, it's a win-win situation. We are making a pretty respectable margin in the 80s%, typically, and our customers are saving a tremendous sum of money. And part of the key and understanding is going back to what Dan was saying about, we're not competing with the hyperscalers in developing NLP or LLM or whatever.
In fact, just the opposite. We're taking advantage of that because our technology people wisely perceived that that was gonna become going lower and lower as the as these technologies developed. So, as a part of our overall revenue, it's still in the single digits, but it's the fastest growing part of our revenue and profitable. And just one last comment. It's our roadmap on AI and automation. I feel a little bit embarrassed in talking about bookings with Dan here on the stage, but that's really one of the key things in terms of they want to be part of a company that is very committed to this, that was in the vanguard of it to some extent, and which is fully committed.
Yeah. And the 10x comment that Barry was referring to is the 10x from the technology spend at the roughly $200 per seat per month, versus the human being 10x, that spend of $2,000 per month for the human. So whenever we can offload the workload from a human agent to an automated agent, the customer is saving. The company, the brand that we're selling to, the ROI is extremely compelling, as you can tell from that math, just that alone.
Does it change the competitive conversation with the other CCaaS providers? Is this something where you feel like you have a competitive advantage or a lead relative to some of the other players? Because you talked earlier about kind of checking the boxes.
Yeah. We're all gonna check the boxes.
Yeah.
But it's the next questions that, you know, when they double-click on that and find out, well, what's your approach? Not only what's your roadmap and are you doing what is needed, but how are you doing it and how are you gonna evolve? We've always taken an approach that we're engine-agnostic, as we say, and that means, you know, when ChatGPT comes along—came along-
Mm-hmm.
and when, you know. When it comes along and you, "Oh, great, we can summarize calls." We can now take, you know, interactions, use ChatGPT to summarize it. We can take the summary of that call, insert it into the CRM so that it's available for the next agent, the next day to see what happened in the call. "Oh, I see you called and spoke to Barry yesterday, and here's what happened." That's a nice benefit. It also keeps the agent from having, in some cases, spend a minute or two typing in their notes from the call. We can now shrink that down to three seconds. You take that average call length from six minutes down to five, you've just saved their labor force, almost 20%.
You're basically able to get those types of returns very quickly and easily. Now, everybody's claiming that, "Hey, we do summarization." The big question is: How are we gonna evolve that story? We're not gonna try to build and have proprietary solutions that can very quickly get leapfrogged. The technology is changing so quickly. By leaving ourselves engine-agnostic, as companies come out with better and faster and more effective engines, we can leverage them and insert them into our solution. That's where large enterprises are saying, "Aha, I wanna, I'm protected in the future and not having to go down a single path that I may regret later." They wanna be agnostic. They wanna have an open system that can leverage new technologies as they come about.
'Cause in five years from now, something's gonna come along that none of us have thought of, and we wanna make sure that we give them a platform choice that allows them to leverage that and bring it into the fold.
Great. Barry, these things all sound very optimistic, so give us reasons for prudence.
Okay. I'll be happy to do that.
With macro or guidance and just the sort of puts and takes of what you're seeing from a demand perspective. We've seen certain verticals-
Yeah
So maybe you can just level set where, where we are currently.
Absolutely. In order to properly understand it, I need to go back and remind you of what Dan said. When you think about Five9 in terms of revenue, think about two buckets that in a typical year, each would contribute on a full year basis, about half of the annual year-over-year revenue growth. First bucket is the net new business, the second part of it is the install base. In terms of the net new, that business is firing on all cylinders, mission-critical systems that have to be replaced and take time to replace, and there's a motivation to do it, and it's been demonstrated it can be done successfully. I'll be happy to elaborate further separately on that.
Then we come to the installed base side of the business, which has, as Dan referred to, had some quite meaningful headwinds from the macro. In particular, we track 17 verticals in total, but the third biggest of those is our consumer vertical, and especially around consumer discretionary spending. The agents are there for a purpose. The purpose they're there for is to handle transactions that are coming into the contact center. And if there are less transactions, or even negative transactions, year upon year, there's gonna be—they're not gonna have agents sitting there idle. So when we came into 2023, we were looking at data that was quite optimistic. One way we can demonstrate that to you objectively, that you can see for yourself, is looking at debit and credit card spending.
In January and February, they were 12% and 8% respectively. And then, just, you know, you can debate how the strength and how healthy the consumer is or isn't, and what part of the consumer, airline, non-airline, whatever. Those numbers dropped down to 4%, ±1%, for every month thereafter, until October... till September, when it collapsed down to 1%, which is negative in real terms, which is what matters to us. So we've had some headwind over there that we, we recognize. And, typically, we would be putting through, you know, our full, beat and raise. In the first quarter, we reduced it to 50%. We saw what was happening. In the second quarter, we reduced it to only 17% and then 0% this quarter. Why?
Well, you've got to face reality. The reality is that auto loan delinquencies are at 20-year highs. Credit card delinquencies are at 20-year highs. J.P. Morgan's look at the... what the consumer is doing or not doing is showing that 42% of them are gonna spend less this year than they did last year. So we looked at all that, and we said, "We're just gonna take a very prudent stance," especially considering that in the third quarter, as I just referred to, things got progressively... For the whole quarter, it was okay, but it was not linear during the quarter.
Any update on the month that you've seen since you just sort of a real-time update on if those trends are continuing at a similar slope to how you would characterize them with the?
Very fair question, and we see nothing there that is effervescent and, ebullient or anything like that. In fact, if you look at, well, two things. If you look at the Bureau of Labor Statistics, we did not know this when we gave the guidance. They gave a, their interpretation of October hiring for the retail sector. Obviously, you hire in October for the November and December holidays. This year, they said it was weaker than even the anemic October of 2022. And then, Deloitte has also come out with another study. So we feel very comfortable with what, how we approach the quarter.
Does that alter the go-to-market focus in Q4 for you, if certain verticals are facing more headwinds than others? Does that at all change the conversation, or is it just the aggregate you might capture from that environment because the agent footprint may not blossom the way that it would in a stronger... Is it-
Right. Said another way, it impacts the installed base half of that equation only.
Yeah.
The net new side and the new logo side, it doesn't get affected because those are long sales processes that are a year or more, in most cases, for large enterprises. And they're on, you know, multi-year digital transformation strategies and projects anyway. Yeah, there's been over time, a year ago, we said there's been some elongated sales cycles. That was primarily due to kind of the uncertainty of recession or not, and a lot of scrutiny put on budgets. But I think in the last, you know, nine months or so, we've seen a pretty much of a return to normal as far as a little bit longer sales cycle still. They're doing more scrutiny, but there's... The volume has actually picked up, to Barry's point earlier.
The net new side of the business is extremely healthy for those two reasons. They got to get off those old legacy platforms. Economy or not, that has to happen. And, secondly, the ROI can help them, you know, from a financial perspective, once they do get to the cloud.
When we think about, there's gonna be a year, Barry, where we have this conversation, and it's not macro-driven. Whether next year or the year after remains to be seen. But when we think about a more normalized environment and the sort of factors that would drive a improvement and growth for your business, what are the sort of key indicators, key factors that we should be focused on?
You're not gonna like this answer. It's gonna be macro.
It's gonna be macro.
That's the only thing that's beyond our control.
Yeah.
Let me emphasize this point. Our logo retention has been excellent, on enterprise, in the mid-90s%. We spring-loaded to take advantage of the time when the U.S. economy, as it inevitably will, rebounds. And there'll be more transactions, more agents, more seats, and-
Mm
all the other good stuff on top of that as well, in terms of the net new, and AI, and automation. That's really the biggest single thing.
A couple of the growth drivers that you presented in kind of framing out the longer term targets for the business are beyond the move-up market, which we've touched on, are international, and then kind of the pull through you might see from the expansion rate as well. But on the international side, is it a similar conversation?
Yes. If you look at the growth rates of... So our international business has been going, you know, through the roof. There are more agents outside of North America-
Yeah
—than there are inside. But, this last quarter, it grew about 28 or 29%... but that's down from what it was earlier on. And it's no newsflash, the European economy is not doing particularly well either—
Yeah
Germany especially. So, it's, that's really the headwind over there as well. But we feel very good because we've made tremendous investments and continue to make tremendous investments that are clearly paying dividends.
Yeah.
We still have India ahead of us, South Africa, which is where the Europeans have a lot of contact center support, and then Asia beyond that.
Maybe you can touch on, I think many are familiar, but for those who aren't, your approach to guidance, how much of this is contemplated in the initial outlook for next year in the 16%?
Very fair question. So, almost as a throwaway, Michael mentioned 16%. Five9 is known for 16%. It does it every year at the beginning of the year, and then, as a starting point, and then-
For one. Yeah.
Except for one... Yes. Do you want to do this?
No.
So, the way we've said it, rightly or wrongly, is as follows: We gave guidance for 2023 of $909 million. Now, remembering always, install base and different set of drivers for them versus the net new versus the new logos. In the install base, every single quarter for four or five, six quarters, up through Q2 of 2023, we said the smart money is betting that the dollar- based retention rate will continue to decline, which it did. However, when it come to the third quarter earnings call, we said we looked at the data, and we believe it's either gonna be flat or down very slightly and reach an inflection in 2024. Now, without endorsing a number for 2024, just for ease of arithmetic, let's assume that it's 110.
That gives you $91 million of the $145 million needed to make the 16% that we mentioned as a starting point, leaving another $54 million for the other side of the business, which is the new logos. There's two buckets that will drive that $54 million: one big, one small. The big one is the backlog that we have from all the deals that Dan and his team have been bringing in that remain to be implemented in the course of the rest of 2023 and into 2024. That's the big bucket. The smaller bucket is the bread-and-butter million-dollar-plus deals that Dan and his team bring in routinely every single quarter, and the ones that they bring in through April or May of next year. Why April or May?
Because those are the ones that will contribute to 2024 revenue, because it takes, on average, for the smaller... these million-dollar plus, but not the mega deals, about seven months to ramp.
So I've asked Barry this question a lot, but I'm gonna ask you the question, Dan, because, Barry, I-
Uh-oh.
We've talked about the expansion rates of enterprise, and those were emphasized as an important point to get to the longer term targets. And when you look at the profile of an enterprise customer and what drives the cross-sell, up-sell motion, it... Can you just kind of level set what makes enterprise expansion rate more robust-
Yes
within contact centers than some of the other segments of the business?
Excellent question, Michael. And if you, if you take that install base side of the business, and then you look at enterprise, what we define as enterprise, and then you take a subset of that at the higher end, 50% of our recurring revenue as a company comes from companies that are providing us with more than $1 million of annual recurring revenue, that ARR. So those customers tend to be the, the largest, they're the larger enterprises. They're buying the full portfolio more. We, we like to say more of what's on the truck they're buying. So their, their ARPU is higher. Counterintuitively, you think, "Oh, a larger company, they're gonna get volume discount." No, they're larger company, they're gonna buy everything we offer in some way, shape, or form.
You end up bringing AI and automation to them, and they're gonna adopt and pick that up. That's why our dollar-based retention rate is north of 110. The overall is 110, but the dollar-based retention, as you move up in the size of the enterprise, is larger and larger, the bigger the company. So we end up getting more expansion and native expansion of them buying more ARPU by putting in more software modules, even when the seat expansion or their organic seat expansion may not be there.
Yeah.
So they may be hitting the headwinds, but they still are tending to buy more software. And so as we have a larger portion of our sales occurring with larger enterprises versus smaller enterprises, it's natural that we're gonna get higher ARPUs, and we're gonna get higher DBRR.
Yeah, that's really good.
I see you didn't like my answers before.
No, I just I don't want to badger you with the same question over and over again.
Okay.
You've heard it on many transcripts already.
Yeah. Okay.
I want to step back and ask a cultural question, because a lot of software companies have gone through change over the past couple of years, fairly significant change, whether it's management or just needing to go through headcount reductions or other things, and your company has been much more stable, despite there having been, you know, puts and takes of the macro and other factors at play. So what do you think it is about Five9 that's enabled that stability? You've still been adding sales capacity, and I think have just taken a more patient approach in general to the market than others.
I'll start, and then please add. But, we're very careful, cautious, prudent, whatever term you want to use about hiring.
Yep.
We're gonna hire when we see the demand and where we see the demand. So when you look at the go-to-market teams, as an example, channels has really picked up, and we've been getting a ton of business from the channels. And they're preferring to do business with Five9, more so than with others in the industry. And so we're getting large demand there. We're getting demand, as we mentioned, the strategic accounts pipeline growing 2x, double year-over-year. And the RFP is coming in stronger. So we're hiring to fit those.
Yeah.
We don't hire in anticipation and hope that the market's gonna be there.
Right.
We wait. Some could argue that, "Hey, we don't-" We like to use the term, we don't get out over our skis from hiring perspective and then have to pull back. We're waiting until we see the demand, and then we're hiring to fulfill it.
Yeah. You opened the door with the culture, and I want to take advantage of that. And so Dan's right, and you mentioned it, we've not had a layoff at Five9. We're just not a faddish company. You know, when everybody's saying revenue is... Why did you go with the revenue faster? No, we want balanced growth, revenue and profitability. And to do that, you need to be mindful of how fast you hire people. But, you know, having these benign conditions in terms of mission-critical systems that have to be replaced from premise to cloud is necessary, but it's not sufficient.
What you also need is 2,700 people who are experts and passionate about the company as a reflection, frankly, of, of Mike Burkland, our CEO, who has instilled this truly remarkable special culture, and it's natural. There's nothing artificial there. People come up to you at a Christmas party and say, "Just what an incredible company. We're so lucky to be there." You can crack a joke anytime you want, saying, "Oh, you're a refugee from that company, that company," and they just burst out laughing. They've never been happier in their lives, and that's how you do it.
Yep.
That's great. We're, we're down to the last couple of minutes, so I'll, I'll turn it over to you, gentlemen, for, for closing thoughts. Key takeaways, areas that you're prioritizing as you're thinking about planning for 2024, and outside of macro, just what the swing factors investors should consider-
Yeah
... around Five9.
I'll touch on one thing Barry just concluded with, which about the culture and the people. The leadership teams on the go-to-market side, you know, even in facing some of these macro headwinds, they wanna be at Five9, and they wanna stay at Five9, and they see the opportunity that lies ahead. So the message I would say is, we're at the early stages of a trend that's gonna occur and a transformation that's gonna occur. We talk about 20% penetration of companies that have moved from the premise to the cloud. Two transitions are happening. One is it's far less than that as you go up market, and you get into the large enterprises. They're just looking at this for the first time.
Now, we've validated and proven that we can serve the largest and most complex contact centers in the world. That's great. We feel it puts us in a pole position. We also feel that that transition will occur. We also look at the AI and automation trend, where just customers are just dipping their toe in the water and starting to experiment with different use cases and how they can really get a compelling ROI from AI and automation. So we believe that this market has a decade or more to really prosper and see some accelerated growth from it. And I think what will happen is, as we come out of this kind of macro headwind, that's when we'll see a change.
However, there's also, you know, a long cycle, a sales cycle, and an implementation cycle to get to the real revenue. You know, that's the challenging part, but it's also the part that gives us visibility into the backlog and into kind of the next few quarters with comfort that we can forecast our business probably more accurately than some.
The only thing I would add to that is a request, Dan, that you demonstrate the magnitude of the remaining market by talking about your 294 versus the 16.
The what?
The 294. The total. So, sorry.
Threw me off.
So, Dan talked about we're in the early innings. Just to give you some idea, we're one of the leaders, not the leader, in terms of size, in terms of the number of seats deployed in the cloud. We give that number each fourth quarter, and last fourth quarter was 294,000. To put that in perspective, there's 16 million agents out there. 20% say we're already in the cloud. 80%, what? That's 12-13 million or so. That's how much left to go to the cloud, and they have to go to the cloud. There's no on-prem vendor left in the industry any longer.
That's a good note to close on this anyway.
All right. Thank you.
Thank you both for the time.