Okay. Hello everyone. Thank you so much for attending this session. We're getting to the end of the day, and we have a great session now with ZoomInfo. So we have ZoomInfo's CEO, Henry, and then we also have ZoomInfo's interim CFO, Graham. So Henry and Graham, thanks so much for joining us.
Of course.
Perfect. And just so everyone knows, I'm sure you're tired of hearing this all day. But to the extent you haven't heard it, if you wanna ask a question, there's a QR code in front of you, so you can just take a snap of that, and I'll try to get to any questions at the end. So with that, let's dive in. Maybe I think a good place to start would just be in terms of what you guys are seeing in the demand environment. I know last print, I think there was some expectation around investors where you guys have been starting to see some improvement in the enterprise space, in the mid-market.
When we look at the forward guidance and the reiteration, I think there were some questions on whether you saw anything weaken in the demand environment or anything got more challenging, or if it was just, you know, you seeing more, more stabilization. So could you maybe comment a little bit on that and, you know, if there are any areas that are maybe tracking a little bit slower relative to your expectations or if it's pretty much what you expected?
Sure. Yeah, I can start. You know, coming out of Q3, we saw the same positive trends that we saw coming out of Q2. Nothing really changed from a positive momentum perspective. When we guided in Q2 and reset expectations, we also sought to reset expectations for 2025. We thought that it was prudent when we came out with the Q3 results and the updated guidance to reemphasize those expectations. You know, the Q4 guidance was in line with that effort.
Perfect. And then, maybe expanding on that, Henry, I know you guys are getting a lot of questions just in terms of what you're seeing in sales hiring, right? And if you are seeing any green shoots to name, if that's across any of the, the verticals. So mind commenting on that. And then as a follow-up to that question, how quickly, I guess, if we do, let's say, see a recovery in the environment, how sensitive would your guys' model be to that?
Yeah, I think, you know, I don't think we are seeing improvement in the hiring environment. But we're not seeing degradation in the hiring environment either. We're seeing less layoffs in the go-to-market land. We think largely our customer base, particularly in technology and software, has gotten to a point where they understand what their operating model is now, where they spent the last two years cutting staff, cutting spend, trying to get to a lower rate of growth but at higher margins. And now they're at a place where things are stabilized, and they're looking for opportunities to improve growth going forward. So I wouldn't tell you that the environment's gotten better, but it certainly hasn't gotten worse. And I don't think I would've told you that in Q1.
Perfect. And then.
Oh, and the sensitivity of the model.
Oh, yeah. Sensitivity.
It's more symbolic when companies start hiring more salespeople. It's not like none of our customers are we fully end-to-end, not every account executive, every account manager, every SDR uses ZoomInfo in any of our clients, even our largest clients, or maybe in our small SMBs where there's one salesperson. But everywhere else, we're not end-to-end, seated. And so if you hire a new salesperson, that doesn't just mean that we're gonna get that seat. But I think if you start seeing real hiring in sales organizations, it's symbolic that companies are investing back in growth. And if companies are investing in growth, we're a very obvious partner for that.
Perfect. And I know you guys have talked a lot about this path to low 90% dollar-based net retention rate. So in terms of what needs to happen to get there, do you need to see a recovery to, to get there? Maybe you can talk about, you know, what the big drivers of that would be.
Sure. I don't think we need to see any meaningful change in the macro to get there.
Mm-hmm.
I think what we're seeing, what we anticipate drives our dollar-based net retention up, is a continuation of the trends we've seen with Copilot, where we're monetizing our customer base on migration from legacy SalesOS to our AI-driven Copilot platform. When a customer migrates from legacy SalesOS to Copilot, we're seeing strong double-digit uplift on that migration. We're also seeing higher utilization and engagement rates from customers who are using Copilot. And historically, higher engagement, higher utilization correlates directly with higher net retention rates. And so as we move more and more customers onto Copilot and they're utilizing and engaging with the platform more, we think that'll be a tailwind to our net retention numbers. In April, we made a change in the way that we brought on new business.
Mm-hmm.
Where our SMB customers who were flagged in our New Business Risk Model as riskier customers to pay us had to pay us upfront before they gained access to ZoomInfo. We think that's bringing in a much healthier customer base. We also resegmented new business so that we were putting more resources upmarket into mid-market and enterprise. And so the percentage of new business coming into the account base is more mid-market and enterprise today than it's ever been in our history. And so we feel good about the customer base improving, and that's gonna improve net retention rates. And then our DaaS, our Data as a Service and our OperationsOS platform grew 22% last quarter.
That's us going into largely our enterprise customer base and mastering and cleansing and enriching their data in a lot of instances to get that data ready for a training model or to put into an AI motion, but in a lot of other cases just so that they can do better targeting and segmentation and total addressable market calculations, and so those kind of pieces continuing to drive momentum in the business, we think gets us to low 90% dollar-based net retention in the foreseeable future.
Perfect. And maybe let's dive into three, those three pieces. So first, starting with Copilot, since this is an AI conference, obviously this is very topical. So I think one thing that's come to light, where I think there might have been a little bit of a misunderstanding, is that Copilot is really a platform shift, right? It sounds like from SalesOS. So in terms of what you guys are seeing today in adoption, when you went from, you know, $18 million in ACV two quarters to now greater than $60 million, how much of that is coming from new logos versus existing customers? And when you see those existing customers shift, can you comment a little bit on what that does to the average size of the deal, right? How are those conversations going?
Is there any concerns around this being like a, or I guess, is this any offset for bringing headcount or seats down? How maybe you could just give like a.
Sure.
Color on that piece.
Yep. So there's three ways that Copilot ACV comes into the business.
Mm-hmm.
One is on new business, where today over 90% of our new customers are coming on to Copilot. And then in the customer base, one way is through off-cycle upsell. So there's not a renewal event, but our account managers are generating demand for Copilot and then selling an off-cycle upsell to move the users at the company to the Copilot platform. And then third way is at renewal, migration at renewal. Migration at renewal can come when we're migrating off-cycle. We're seeing a strong double-digit upgrade, a strong double-digit growth on the per seat that we're migrating to Copilot. When we do it at renewal, where it's an upsell, we're also seeing that strong double-digit growth. Oftentimes, the customer might be in the middle of a downsell. Hey, you know, we have less salespeople. We need a downsell.
And we're able to use Copilot in those instances to get the contract back to flat or maybe with a little bit of growth. And we're also using it in instances where a customer might churn, and we're able to introduce Copilot and keep the contract. And so there are elements that reduce downsell. There are elements that reduce churn.
Mm-hmm.
And then there are elements that come in on new business, and then we're selling it with a double-digit per seat upsell when we're migrating our customers as well.
Perfect. And this might be a very tough question to answer because it's obviously Copilot has just been rolled out. But when you think about when those Copilot customers come up for renewal, maybe you can talk about what you're seeing in terms of usage of the platform, maybe breadth of users that it applies to relative to what you saw to SalesOS. So when those Copilot customers come up, is there any insights that you guys have gained that suggest, you know, you will see good, good expansion u pon those renewals?
Yeah. So first, we're seeing significantly better engagement and utilization rates on Copilot with the first cohort. This group won't come up for renewal until May of 2025, or the large percentage of them won't come up for renewal until May of 2025. But we're seeing higher utilization rates, higher engagement rates. Our customers are telling us that on average, they're generating 25% of their pipeline from Copilot-related recommendations. They're setting 58% more meetings, and they're saving eight hours a day using Copilot, so we feel really good about that. The engagement and utilization rates haven't shrunk down either, and so what you would usually see with a new product, like you get a new car, you use like everything in the first, like, couple of weeks, and then utilization dies down. Maybe you're not using like the, I don't know, the heated seats as often.
Mm-hmm.
With Copilot, we haven't seen that utilization shrink when customers come on, so we feel really good about that. And so we do anticipate better net retention rates when that, when that customer base comes up for renewal. I think that's.
Yep. That's good. And then, just building on that, so thinking about the enterprise ACV, which was up 1% quarter over quarter, so would that suggest that most of that uplift is coming from Copilot? So maybe you could delineate a little bit in terms of how are the enterprise conversations that are going that are adopting Copilot, right? And then what are the conversations for those that maybe aren't quite there yet? And how do you see that evolving over the next couple of quarters as Copilot, you know, gains steam? It's been in the market for a while. Like, how are you guys thinking of that migration in the, in the enterprise space?
Yeah. So enterprise growth is coming, you know, from two places: Copilot, Copilot migrations and our Data as a Service and OperationsOS products. The conversations in the enterprise, you know, what's interesting is when you think about AI, and AI agents that are out there today, what you're hearing most about is customer service agents, and agents that answer customer questions, support questions, and you could, you know, replace a support person with an AI agent that answers the same support questions oftentimes, you know, even better than a human can. The reason why that's been such a success is that it only requires first-party data to be successful.
If I wanna build an AI agent to answer customer service questions, all I really need to do is go take my knowledge base about the product, take some technical docs, product docs, and then the previous tickets that I've had that have answered similar questions, put it into the LLM, and the LLM can start now answering questions in the future that it already knows the answers to. It's all of my first-party data. When you think about go-to-market for AI, or AI for go-to-market, your first-party data is just not good enough to drive go-to-market AI. And not like it's close and it needs a little bit of massaging, but like it's very far away. And so when we're out there talking to chief revenue officers about what they're doing for how they're using AI in go-to-market, it's largely crickets. They're not using it yet.
And when they go to take the first step, the first step requires them to go to the data source that they anticipate leveraging within the AI, within the AI models and within their AI agents. And that data source is their CRM. And their CRM, I haven't met in 18 years of doing this, I haven't met a single customer who's told me, "You know what? I can absolutely rely on my CRM data for territory assignment or planning or anything." Everybody is very disappointed with the quality of the data in their CRM systems. And so for them to be able to leverage that data to build AI on top of, it requires them first to master that data, to cleanse it, to enrich it, to append it, to have their entire total addressable market available within that data.
But then, even if you were to master that data in your CRM system, there are very few CRM systems that have data on your entire total addressable market. They have it on your customers and a couple prospects that have been turned into opportunities, and if you really wanna prioritize accounts and prioritize who you should be engaging with, you need a full view of your total addressable market. That doesn't exist in your CRM, and then one of the other things that we've done with Copilot in ZoomInfo is we're ingesting a whole plethora of additional signals into the platform. Signals that would include, like, whether a company is researching your competitors, whether they're researching products and services like yours, whether somebody at the company was on a podcast and mentioned something that was relevant to your business.
If I just use one of those examples, let's use a podcast example. Imagine, you know, we're ingesting thousands of podcasts every month. A nd we're taking the transcription of that podcast, and we know it's the CIO, and he's talking about a cybersecurity initiative. That's really important to lots of cybersecurity companies. But it's not important at all to ZoomInfo. Like, a seller at ZoomInfo doesn't care that a CIO is talking about a cybersecurity initiative.
So we have a brain in the middle that can look at all of these inputs, like podcasts and press releases and job postings, and then say, "Oh, this is important to these three customers and not important to these other customers." The reason I use that example is that if you try to wrap your head around thousands of podcasts living natively inside of a CRM system, it's not, it doesn't exist like that. It's not designed for that. And so for go-to-market, the current legacy systems are not designed for AI to be built on top of, particularly for go-to-market use cases.
Mm-hmm.
I think that's why we're so bullish about Copilot and also why we're seeing growth on the data side, because the obvious first step is I gotta get my data house in order, and the next step is I've gotta build the agents and the AI on top of that, and we've spent the last almost two years thinking through what are the things that an account executive does, an account manager does, an SDR does day to day, and how do we start automating all of those tasks out one by one by one. We started with account prioritization and then buyer group identification and then content creation for that, and then pre-call meeting briefs and post-call follow-ups and account planning.
And so we're thinking through all of those different things that the go-to-market organization does, and then we're using that core data foundation to start automating out those tasks.
That was all great color and provides really good context. Maybe if we shift to the second piece that you mentioned, which I think was around the new business risk model that you guys put into place. So Graham, this is probably a good one for you, but we did get some questions post the quarter on whether the increase in disqualified pipeline was expected, right, and built into your guys' expectations. And I guess, one, any clarity that you can provide on that. And then two, just as we think about the level of disqualified pipeline going forward and what that trend could look like with write-offs and bad debt expense too, any insight you could provide?
Yeah. Absolutely. So just a little background. The new business risk model is something that we implemented in April of this year, and we started running through all of our new business pipeline through it. So what we're effectively doing is looking at the characteristics of any lead, and then assigning basically a grade and drawing a line and saying anything that's below a certain grade, we're gonna require cash upfront for before they actually get provisioned on the platform. So in Q2, think about it as like beta testing that model. And we, at that point, we were disqualifying about $1 million + per month of high-risk new business ACV that we would've historically sold and likely written off or had trouble collecting cash on. In Q2, we optimized it at about that $2 million level per month.
We knew this when we guided, like, this was something that we were doing in July and August ahead of the call in early August, so this was not kind of a new dynamic in our trajectory. I think we feel pretty comfortable at this, you know, $2 million, $2 million plus-ish level. I wouldn't expect it to change much, and keep in mind, like, this is not economically, this is a cash we probably weren't going to get. So if you think about the kind of the life cycle of a high-risk transaction that we don't collect on, we sell it, we get a DocuSign, we bill, we go out, and we try to collect the cash, we try to collect again, we follow up, potentially we send it to a collections agency.
It's about, you know, a six-month delay to the point of actually writing it off. So we should start to see the benefit of that in a higher quality of revenue, more efficient collections and sales processes in, you know, certainly in the first half of next year, but potentially at the, you know, as we exit Q4 of this year.
Perfect. And then in terms of what you guys are seeing across customer segments, 'cause I think you've seen very different trends in the enterprise versus mid-market versus SMB. So could you unpack that a little bit more? I guess when you think about NRRs, right, new logo adoption, what is, what historically has it looked like across those three different customer segments? And then as we get into next year, what are the puts and takes? It sounds like you have expectations for mid-market and enterprise to be on a path towards getting to mid to high single-digit growth. SMB might still be the drag. When do we start to lap that? What does all of that look like?
Yeah. Absolutely. I'll just start with the mix of the business right now. Enterprise is 41% of our total ACV. Mid-market is 28%, and SMB is 31%. If you think about the trajectories of all of those, enterprise is increasing as a percentage of the mix. Mid-market is, you know, stabilized at that 28%, I would say, and SMB has started to decrease. SMB is decreasing as a mix of our total ACV. When we look at the retention or kind of the recent history there, you know, historically, enterprise retention was well over 100%. We think about 2019, 2020, 2021, and actually dipped below 100% at its mid-year. It's gotten back up to 100% and has sequentially improved in the last two quarters. So we're really positive about the trends in enterprise.
We feel like we are, you know, correctly resourced to go continue to drive growth in the enterprise. Mid-market is where we saw, you know, that was the greatest driver of our revenue deceleration over the past two and a half years. Mid-market's a very software-heavy segment. Again, this is a segment where we had well over 100% retention. We saw a 30+ point swing the other way, and we saw mid-market stabilize in Q2 and actually improve sequentially in Q3. Getting mid-market back to and above 90% on that path will be, you know, a huge, potentially a tailwind to getting that 2/3 of the business mid-market and enterprise to growing mid-single digits or higher, as we get into 2025. Then, yeah, SMB has the retention there degraded this year, year-to-date, relative to 2023 year-to-date.
We've seen signs that degradation is slowing, and we're approaching stabilization. I would think about when we lapped the introduction of the new business risk model, which happened in Q2 of this year. So when we start to lap that in Q2 and the optimization of it in Q3, that, you know, that kind of period is when we would expect SMB to be certainly a smaller percentage of the business, but a healthier version of an SMB business.
Makes sense. And if we just bring all of that together and try to understand, you know, what that means for growth next year and the mechanics behind it. So is it such that some of the SMB pressures and the new logo pressures are still gonna weigh on the full year despite you starting to, like, lap that in Q2? Can you just, like, walk us through, you know, the scenarios and when we do start to reach that inflection point and potentially, you know, what that means for the full year?
Yeah. We, you know, I'd expect the SMB pressure from lower retention levels we've historically seen and kind of the intentional disqualification of some of that higher risk to weigh on or, you know, potentially impair growth when you look at a RPO or a bookings perspective.
Mm-hmm.
And we're certainly not taking credit for any stabilization or improvement there when we talk about expectations in 2025. By that same measure, the positive signs we're seeing in mid-market and enterprise, we are also not taking credit for those continuing as we start to think about 2025. We're just as, you know, bullish as we were coming out of Q2 about the positive momentum we have for two-thirds and growing part of our business. But we're gonna discount the positive trends and certainly not, you know, have any expectation of improvement on the negative ones.
Perfect. And now let's turn to the third opportunity, which is the DaaS opportunity in the operations. I know that's something that you guys have been talking a lot about, Henry. You were talking a little bit earlier about this need to cleanse your CRM system and that being a big play there. So can you just talk about what the momentum, you know, you're seeing on that, the pipeline, maybe the conversations that you're having with enterprise on this that's not necessarily something we see in numbers or appreciate today? Any update there I think would be helpful.
Sure. I think the DaaS and OperationsOS business is growing, primarily in the enterprise. It's an enterprise offering, and it does, you know, a couple of things. One, it'll cleanse, dedupe, enrich, all of the data in your CRM system, but also maybe in your data warehouse or in Google Cloud or in Databricks. Wherever that data exists, we can plug in and start cleansing and enriching and appending that data. Once we've done that, it also puts in guardrails to keep that data updated and consistent, and you know, what our customers are telling us today is really two things. One, they want to build their own prioritization systems.
They wanna get more specific around targeting and segmentation. They wanna be in a place where they can start building AI on top of their data. And so they're bringing DaaS and OperationsOS in as the first step in doing that. Again, we're super early for AI and go-to-market. When we have these conversations, it's very rare that we meet a company that's already implemented AI across their go-to-market or in any pockets of their go-to-market. And so we're still pretty early in that conversation. And DaaS tends to be the first step of the process. When we get to have that conversation, one really interesting thing that happens is a customer tells us, like, "Yeah, we wanna master our data so that in the future we can start building," and then they describe Copilot. And they say, "Look, we wanna prioritize our accounts.
We wanna use signal data to tell our sellers when they engage. We wanna write the content for them." And so then we're having the conversation of saying, like, "Listen, your data and your CRM should be cleansed and maintained and, and integrated. But if you're going down a journey to try to build your own AI solution on top of that, just take a look at Copilot. You know, you could spend the next five years with dozens of developers, and you're gonna have much less than what we have today. And so if there are little missing pieces, let's bring our Chief Product Officer in, have a conversation about what the roadmap looks like, and get you guys onboard Copilot 'cause it's gonna continue to be something that we invest in and that continues to mature." So it opens the door to Copilot conversations.
Mm-hmm.
As well, and that's been a pretty successful motion for us so far.
Yeah. Interesting. Would you say it's similar to Copilot in the sense that it's becoming a lead-gen opportunity for you guys? So are you finding more new customers that are interested in that, or is it still really a cross-sell, motion for you?
I mean, it's a little bit of both, but it's more of a cross-sell, upsell motion in the customer base than it is a net new logo motion today.
Perfect. And then, Graham, maybe turning to margins for a little bit. So you've been talking about free cash flow per share growing materially in 2025. But could you maybe offer a bit more color on, you know, what is meant by that? What are the drivers behind them as well too?
Yeah. Absolutely. So when we think about our adjusted levered free cash flow per share and growing that meaningfully in 2025, we view that we have a few options or a few levers to do that. The first one and the one that we're prioritizing and the one that we feel we are resourced for is to grow the top line, to grow revenue. You know, after that, we can start to talk about adjusting margins. If we are not growing revenue sequentially, you know, of a timeline that we expect or want to, then we would start to, you know, expand margins and take some costs out. Then the third lever is, you know, continuing to be opportunistic, buying back shares. When we think about the Adjusted Operating Income margins, we had a 37% AOI margin in Q3. We've guided to 35% in Q4.
Some of that is just timing. It's not as much trajectory. I wouldn't, you know, describe it as a declining margin business. I think that 36%-37% annualized AOI margin is the right baseline. And again, if we're not getting to the sequential revenue growth on our timeline, we would start to expand that closer to, think about like a 39% run rate on the path to 40%. There is some seasonality to margins as there is to revenue. So please, you know, keep in mind that Q1 of 2025, we'll have two fewer revenue days. We recognize revenue on a ratable basis by days, as well as we'll have some seasonality from benefits and retirement, 401(k) resets that will drive margin down in Q1, but then there will be offset in the back half of 2025.
Could you talk about the sources of the operating leverage and cost efficiencies that you see in the business? So let's say we run into a scenario where, you know, you're not necessarily maybe seeing a recovery as quick as you would have hoped in enterprise and mid-market. And I think you've mentioned that, you know, there's still opportunities for margins to come up. So what would be the drivers behind that?
Sure. I mean, some of them are variable just to growth. So you can think about cost of service and hosting costs, data acquisition, and AI consumption costs. And then, you know, some of it is what level of marketing spend we have out there. That's something that we can dial up and down pretty fast. And then it's just a question of how fast or slow we grow headcount.
Perfect. And then maybe in the last minute or so that we have, a question for both of you. So I'm sure that you've had a number of conversations post the print. And just in those conversations, I guess, what parts of the story do you feel like are misunderstood? And that's 'cause it clearly sounds like in your guys' tone and your messaging, you're very optimistic on the business. So would love, you know, your thoughts there and to leave everything with the audience.
Yeah. I think a couple of things. One, I think we spent kind of two years in this really difficult environment from a tech spending perspective with our software and technology customers. And we spent those two years rebuilding our technology infrastructure, building technology and software for the future, rebuilding our data infrastructure. Today, when we're out there and we see 22% growth in DaaS, where we see growth in the enterprise, where we see Copilot migrations increasing, month over month, quarter over quarter, we don't think that those things are flash in the pan moments for us. We've made concerted investments behind these. We think they have great product-market fit and that the improvements that we're seeing and the uptake on those products is gonna continue. I've been a personal buyer of the stock in a pretty meaningful way in the last two quarters.
I have a tremendous amount of conviction about the future of ZoomInfo and feel really good about the products we're putting out in the market, the customers' satisfaction that we're seeing, the NPS scores we're seeing, the utilization and engagement scores we're seeing, and we, you know, I really do feel like we've, we're really clear that we've stabilized, and now we have a real opportunity to continue to improve.
Perfect. Graham, anything to add?
No, that's it.
Okay. Perfect. Well, awesome. Well, thank you, everyone in the audience for attending, and let's give Henry and Graham a round of applause.