Confluent, Inc. (CFLT) Earnings Call Transcript & Summary

Okay. I think we're live. Good morning, everybody. I'm Brad Zelnick with the Deutsche Bank software team, and welcome back here for day 2 of our 2022 tech conference in Sony Las Vegas, Nevada. Joined for this session by Steffan Tomlinson, CFO of Confluent. Steffan, thank you very much for joining us.

Thank you for having me.

Great to see you. The format of this session is going to be a fireside chat. I've got a bunch of questions for Steffan that we're going to go through. And if time permits, at the end, maybe we'll take some from the room as well.

But maybe just to start off, I want to talk big-picture Steffan, and get into the long-term opportunity for Confluent, but before we do, for better or worse macro remains top of mind for investors. And on your recent earnings call, you talked about greater scrutiny on new deal approvals in June and July. How has this trend progressed into August? What are customers telling you? And are there any updates on the deals that pushed out of last quarter?

Yes. So first off, we think the demand environment is very strong just in general for the new category that we're creating data in motion. We had seen some pockets where deals were taking longer to convert, to a purchase order and we incorporated that into our guide for Q3 and for the balance of the year. And so we ended up taking roughly a $4 million to $6 million haircut for the full year. We haven't really seen any material change to the macro from our point of view. And the deals that had pushed many of them closed as expected. And we're looking at seeing the balance of them close as well. So overall, things are good.

Good. Well, things I think in life or -- I don't believe anything stays constant. They're usually -- they're getting better or getting worse, but that sounds -- it doesn't sound like things are getting worse from your perspective.

Correct.

Fair enough. Thank you for that update, now that we got that question out of the way, maybe we can get into the more interesting stuff. So adoption of event streaming architectures remains relatively new, I think. I mean they were prior sort of iterations of what we think of it as event streaming architecture today. But could you remind us about some of the advantages over traditional message-oriented systems? And what the primary use cases you're seeing are -- where you're seeing strong product market fit?

Well, as you look at the technology landscape and how it's evolved over time, there really hasn't been data streaming and event streaming in the past as it is today. In the past, you had vendors doing ETL, point-to-point batch solution-related implementations of technology. In this world, what you really need is real-time processing of not only data, but then the use cases, what we call as connect-and-build you need to be able to connect all of the data stores, applications, et cetera, and then build real-time applications on top of that. And that is where there's a fundamental departure between Confluent and any of the legacy players that are out there. When I think of core use cases, it really depends on the vertical we're talking about. In the healthcare space, we have connected health care records, we have patient monitoring. You need real time for that. In the retail space, think of real-time inventory management, think of personalized promotions, both the back-end operations and the front-end customer experience needs to be real-time. In the financial vertical space, fraud detection, capital management, there's a whole host of use cases. And then there are use cases that cut across every industry. Think about Customer 360, you need to have real-time engagement with customers with -- in whatever industry that you're in. And so Customer 360 is definitely a use case that spans across every industry. And I can tell you that the technological position that Confluent has, given the fact that we're both on-prem and in the cloud. We call it cloud complete and everywhere. We need to be wherever our customers' data and applications reside because in this world of hybrid cloud, multi-cloud environments, [ we need ] a solution in a platform. And as you know, Brad, platforms win over time. We're a company that's 8 years old. We're turning is this month.

Happy Birthday.

Yes. Thank you. And we have a lot of runway to go in front of us to really capture the market, and do it in an efficient manner.

Great. That makes a lot of sense. And I think in today's hyper-connected world. Real-time means something different than what other similar sounding solutions of 10 or 20 years ago, we're doing real-time. Back then, it could have been minutes to hours versus millisecond-type time frames today, and that's really important. Steffan, maybe just in terms of the size of the market opportunity, how much of the $50 billion or so TAM that you've identified? Does the current solution set fully address? And as you think about your growth levers, how do you think about investing in things that expand TAM versus capabilities that enable you to drive penetration of the existing opportunity faster and more effectively.

We've constructed our TAM, both on a bottoms-up and a top-down basis, and we converge on that $50 billion number by really taking components of 4 distinct markets. App infrastructure and middleware, data integration and quality tools, database and analytics and BI. And we take fractions of those markets to construct the $50 billion TAM. Our current product set today addresses that $50 billion TAM. What we outlined for investors at the time of our IPO was that over a multiple year period, that $50 billion TAM should increase to roughly $90 billion as we continue to introduce new feature functionality. The innovation engine at Confluent continues to come along, and cloud is definitely where we're investing the most. And what we see is customers who are coming in for specifically the features around ease of use, security, data governance and building real-time applications on top of that. So the short answer to your question is our current product set addresses the $50 billion TAM today. And over time, as we introduce more feature functionality, we're expanding within those 4 markets. We're not really looking at adjacent markets to expand TAM because there's so much wood to chop in our current TAM.

Makes total sense. And just as we think about the adoption pattern inside of accounts, what are the biggest obstacles that you see within an organization to undertaking real-time data initiatives whether it's product fit, technical know-how, maybe unclear ROI? And how is Confluent addressing this?

Well, it's been an evolving landscape, on that question in particular. The lineage of Confluent really starts with the formation and foundation of Kafka. And for those of you in the audience who don't understand this, I think it's instructive to just give a little bit of a history lesson. So Jay, our Co-Founder, and Neha and Jun, the 2 other cofounders were engineers at LinkedIn, and they were trying to run LinkedIn in real time. And there's no commercial software out there, any of the legacy vendors today. None -- there are no solutions out there that actually hit the mark. So they created Kafka, while at LinkedIn, and they had permission to release it to the open-source community, and it became one of the most successful open source projects of all time. Hundreds of thousands of companies use open-source Kafka. Now this is where our details matter. It's not just about a single use case of Kafka. In order to get a company on what we call the customer growth go-to-market journey for us, it does typically start with the developer and the developer who is -- who may have -- historically have downloaded open-source Kafka, but now today, they are usually starting with Confluent Cloud as a free way to just try Confluent, and all of the great ease of use, future functionality that comes along with it. And so to answer your question are like, what are some of the barriers for really widespread adoption? It's when customers start to use the open-source version of Kafka, and they use it for a project. But when they go from projects to multiple projects to platform, unless they have hired a massive engineering team and some companies have done that, right? And some companies, large digital native companies have done that. But the vast majority of companies out there have not been able to hire all of the technical talent to have widespread adoption of Kafka throughout the organization. So a few years ago, Confluent made it super easy for companies to get over that hurdle of moving from having to hire loads of people and have lots of infrastructure in play by us introducing Confluent Cloud, customers can easily and seamlessly, and in a frictionless way, start with a managed services offering from Confluent. And when you look at any financial kind of hurdles that are out there, the ROI and the TCO of the Confluent offering brushes open-source Kafka. So if you look at ROI, the payback is within 6 months. And if you look at the feature and functionality of our campaign that we just released our marketing campaign, where we were saying, "Hey, we're 10x better than Kafka", think of better elasticity, better storage, better performance. There are so many things that relative to the open source alternative we skew a lot better on. And in this environment, when budgets are tightening and people are looking at the bottom line, ROI and TCO matter as much today as they ever have. And the Confluent Cloud solution in our overall Confluent Cloud story resonates very well with customers.

That makes a lot of sense. And we're hearing that across the board, our conversations with industry constituents, security infrastructure, all the way up the stack of just additional scrutiny and ROI really being very closely looked at. Maybe just as we think about, again, the customer's journey that you talked about, the importance of getting involved early in that customer's data and motion journey has been stressed on the past couple of earnings calls. What are you doing to identify and get in front of developers upstream that are looking to use Kafka for the very first time. So they don't necessarily have to go through the trial and error of eventually making their way to Confluent and Confluent Cloud.

Well, we recently did a number of things for what we call the pay-as-you-go service, which is the service that we've had for a long time. But developers used to have to start swiping a credit card to get going. Now we've removed what we call the paywall barrier. And we've seen a very large increase in just number of sign-ups of developers coming to our platform to start with.

And that was just in the last quarter that you've done this, right?

That's right. It was actually at the very end of March, but last quarter was the full quarter that we saw this. So we saw a 55% quarter-over-quarter increase in sign-ups, which means the developer community is coming to Confluent as the starting point for their data and motion journey. And so we removed the barrier around paywall, so they get to start for free. It's a free trial. And then down the road as they are understanding the product, they're getting value, there's a point in time when they have to swipe the credit card in order to convert from free to paid. And we believe that over time, that's a winning strategy because the customers that matriculate through that process will be more engaged. They will convert to ultimately bidded contracted customers because the journey is start with pay as you go, which they're paying rap rates for. And then as they get to a larger implementation, they will come to us or we will come to them depending on where they are in the customer segmentation and they will sign up for committed contracts. And once that committed contract is in place, they get additional financial benefits, that sort of thing. But it's also a commitment to really running a data and motion infrastructure that's a foundational element of -- that Confluent is the foundational element for.

Makes a lot of sense. Maybe if we could just double click on removing the paywall, and I'm sure you guys have given us a lot of thought. But just to appreciate, is that initial free use time-based, feature-limited, and is there any potential for trade down effect where somebody who is paying you in the past is now getting away with your free offering? Or am I not even -- you're thinking about it correctly. And that's okay.

Yes. Those are good questions. Yes. There are a number of features to our pay-as-you-go journey for customers. But it typically is time-based. And so when people get to a certain point in time, they're going to have to make a decision. The trade-down effect -- the potential trade-down effects that we've gotten questions on. And the trade-down effect is customers who may have been running Confluent they may trade down to open source Kafka, that really is not a prevailing dynamic anymore. Back a couple of years ago when the Confluent Cloud offering was less mature. We saw limited instances of that. Today, we're not seeing that really at all. And if you look at the underlying fundamentals of the cloud business, we have a greater than 150% net retention rate on Confluent Cloud. It's the fastest-growing part of our business, both in terms of just raw dollars in then on customer count. And then if you look at another subsegment of our business, which is our hybrid customers, which are customers that are running Confluent Platform and Confluent Cloud that has the highest net retention rate. And the way that infrastructure plays work like Confluent, once you're in, it's very hard to take out. And as we've added more future functionality as we've wrapped more services around the offering, the stickiness of the platform and the value add that we're bringing to customers is as high as it's ever been.

Everything you've just said, we actually hear in industry conversations that we've had where we've seen this pattern. We're -- and many times, customers starts with open source Kafka realizing they need help and moving to, in some cases, MSK or another public cloud vendor offering then ultimately finding their way to Confluent as better and a more complete governance solution. But beyond this journey, I think I alluded to this earlier, the same question of trial and error, like how do you go upstream, even as you talked about removing the paywall, is one of the things which sounds like it makes a lot of sense, and it's been very successful with the customer count that you then saw a jump up in last quarter. Are there other things that you can do in marketing campaigns? What else might you be able to do to kind of get them sooner on the right path, so to speak?

It's a really good question. So as we think about the go-to-market organization that we have, we have great leaders and Erica Schultz and Stephanie Buscemi, heading both field operations and marketing, respectively. The focus on use case is very important to the fundamental trajectory of a customer getting up that customer growth and go-to-market journey. And the reason why I say focus on use case depending on what industry you're in, you really need to understand what is the art of the possible. And for a new category-creating company like Confluent, customers in their own mind, they're trying to solve a specific problem. But then when they solve that immediate problem, the light bulb goes off and they are like, okay, well, we could -- in the retail space, they can go from inventory management on a real-time basis is very important. Well, then after they get that done, it's all about personalized promotions than there are curbside pickup logistics. There's a whole host of applications and use cases that can get built on our platform. So the focus on use cases from a marketing standpoint per industry is going to be very important for us. The partner ecosystem that we're still in the nascent stages of building that is going to be a very important aspect of our growth. And the reason why I say that is partners come in many different flavors. The 3 main cloud service providers that are out there. They're great partners of ours. There's also a level of coopetition with them depending on which 1 we're talking about. But that is a route to market. We are on all 3 cloud service players, marketplaces. Each of the sales reps for those cloud service providers retire quota when they sell Confluent through the marketplace. And when you think about from a customer standpoint, multi-cloud is very important. And our ability to provide multi-cloud solution to our customers gives them a lot of leverage as they're building out their own infrastructure. So the fact that we're multi-cloud, we have this customer growth go-to-market journey, we have our partner ecosystem that I would still say is in its nascent stage, and that will be a growth vector for us going forward, are all things that will be catalysts for us to increase our performance and market share over time.

Yes. The value prop as you describe it, and as I understand it, it's very clear, it's very distinct. I guess it's relative to the 100,000 or whatever number you quoted earlier of open-source Kafka adopters, getting them as quickly as you can to realize that distinction and to make their way over to Confluent. But that's all the great stuff that you're doing. And that's why we see strong metrics, strong growth. And to my next question of just net retention, like you've talked about greater than 130% even better, as you said, for Confluent Cloud, and what you also see for hybrid customers. But that clearly shows that once customers start using the product, their usage expands pretty quickly. And you talked about some of that with the different use cases. But up to the point where you now have over 100 customers with ARR greater than $1 million, can you just walk us through the typical expansion pattern within an account and then touch on what penetration looks like with some of these largest customers versus like how you might think out the wide space of, all right, if I can get all these other accounts to look like are most mature, what is that opportunity? How do you think about that?

Well, it's evolved over time. We are agnostic in terms of how big of a deal that we land originally. We just want to get our foot in the door because once we get our foot in the door, we show we have the versatility of our platform that can really show the expansion over time. So in some cases, some of our largest customers that are out there started with pay as you go. And they didn't even talk to a salesperson. They started using the product. And then when they got to a certain size and scale, which is usually more than 1 use case or a project. It's multiple projects, multiple use cases. That's when they effectively come to Confluent our sales organization, and then we engage. And then we show them what the art of the possible is. So some of our largest customers started with a very small footprint that grew over time. There are other customers, let's say, in the digital native space where their first or second purchase order was very large. And they started and they are like, we believe, in data and motion as a starting point for our business. And I would say that dynamic I think we'll be growing over time. I think we will start landing larger digital native accounts over time. But this pay-as-you-go journey that really starts with this frictionless way of getting people to use the product and then they mature along the curve. And it is about the multiple use cases. And then when you think about being a central nervous system, which is another metaphor that we like to use. You can't be a central nervous system to half the body. You need to be wherever the customer's data is. So as an example, we have large Confluent Platform only customers that have grown with us for the past 5.5 years. And they haven't even started using cloud yet. But they have on their road map to use cloud and move workloads to cloud. And as that happens, that's going to be a natural expansion opportunity for us because they're going to want to embrace Confluent Cloud as part of the overall central nervous system because they already have Confluent Platform in place. So this concept of being cloud-based complete and everywhere is a fundamental aspect of the journey that customers are on. And then it goes -- again, goes back to the use cases we talked about.

Cool. Makes perfect sense. You've talked about investment in dollars and engineering hours put into Confluent Cloud, and you're now seeing some of the returns in your results with record net adds, for example, in Q2. Can you talk about what makes Confluent Cloud a fundamentally different product than your self-managed offering? I mean you touched on this a little bit, but even just to put a wrap a bow on it, like how do you get that on-prem customer or self-hosted customer to really embrace, and how much of it is hybrid, how much of it is even scrapping what they might have and bringing it eventually over into Confluent Cloud?

Yes. So most of the dynamic at play is these are additional workloads that are being put in the cloud. We've seen very little cannibalization of companies using -- that had started to use Confluent platform and then when they move to Confluent Cloud deprecating Confluent platform. It's actually been additive. But the main difference between Confluent Platform and Confluent Cloud is, Confluent Cloud is a fully managed hosted offering that's cloud-based. Confluent Platform, and then if you were to look at like Kafka, as an example, if there is open-source Kafka user that's out there, dedicated teams of individuals are running that software on-prem. So you have lots of engineers you have lots of infrastructure, you have other costs that go along with that. And that's why we shine when we look at the TCO comparison between a Confluent Cloud offering. And specifically, when you look at open source Kafka, large implementations of open-source Kafka, the ROI and TCO is a no-brainer because what fully managed hosted offering doesn't win against open-source, do-it-yourself environment.

And I guess at some scale, right?

Yes. Yes, exactly. It is...

You have to be a hardcore technology customer yourself perhaps a hyperscaler to roll your own, but [ greater ] than 98% of the world is coming to you.

Correct. Yes. That's right.

All right. That makes sense. Maybe just on another topic. Your guidance for cloud revenue net adds suggests a pretty big deceleration from platform revenue growth in 3Q, actually I think that makes sense, especially looking at it relative to 2Q. Can you help us square this with the fact that it sounds like platform still accounts for nearly half of new ACV, and your statements that the growth in cloud is coming much more from expansion than any kind of cannibalization.

Yes. Yes. So I just want to rephrase the question.

Go ahead, please. Help me to help me.

So I think what you're getting at is the implied guide for Confluent Platform is down sequentially?

Yes. Thank you.

I think that's what you're saying. Yes. And why is that the case? And is there a cannibalization? The reality is we've had 3 quarters in a row of new cloud, ACV being the majority of new ACV. And so when you see that type of change happening where the market is speaking and we run our business on an ACV basis, right? And so ACV is a key leading indicator of what the kind of the future holds. And from a mix standpoint, we're still going to be looking at Confluent Platform being up like meaningfully year-over-year. But the mix shift is such that, that Confluent Cloud will be a bigger portion of our revenues over time. It will be growing in share. And the rev rec for Confluent platform is such that when you get -- when we get an order in, we take roughly 20% of that upfront, and then the rest ratably over the life of the contract. So when you see lower new ACV coming from Confluent Platform, that's going to be a direct impact on a near-term basis, which is why you're going to see Confluent Platform, not grow as quickly as Confluent Cloud. It's still going to grow, but it's just not going to grow as quickly.

That makes perfect sense, and thanks for the help because as you can see since the last time we met, I need glasses now, and I misread my own question, I think. Maybe if we can look at the business maybe from an RPO perspective if that's cool. The past quarter, we saw current RPO bookings grow much faster overall bookings, I think, for the first time. Any changes to note in terms of customer contract duration in the current environment or other factors that might have played into this? And any changes around incentives for duration?

Really no changes for incentives to duration. Our duration is customer led. And we see a healthy mix of both multiyear deals and annual deals. The quarter-over-quarter comparison of RPO is a little -- it takes a little explaining. In our fiscal Q1, we closed the largest deal in the company's history, which was a deal with New Relic. It was a 3-year, 8-figure deal, which had very little current RPO. And so most of that went to a long-term RPO. And so that actually created a little bit of just quarter-over-quarter comparability dynamic. Now the duration on a year-over-year basis is largely unchanged. I mean there's going to be puts and takes to it, but the duration is largely unchanged.

Got it. That makes sense. And maybe just to -- you are the CFO, so I got to ask you about margins, right?

Yes.

Operating margins took a nice step in the right direction this past quarter and look set to trend a little bit higher in the back half of the year as well. With some cost actions and a slowdown in hiring that you've talked about, how important is it for you to show incremental margin improvement from here? And are there any -- are there additional cost levers you would look to pull if macro conditions were to soften, and get even tougher.

We're fully committed to our growth and profitability framework that at the time of our IPO, we outlined to the investment community. We are a big believer and you have to manage both sides of the equation. And this company has never been about growth at all costs. It's always been about showing growth and margin improvement. So as we think about the trajectory the company is on, what we've said to folks is that we're going to be showing annual operating margin improvement in 2022 and '23 and '24 with exiting '24 at roughly, call it, breakeven on an operating margin and free cash flow basis. So Q4 '24 is what we've set out there. Now that is not some sort of C change the company went through because we're trying to react to the market dynamic that's out there. That has always been our plan. So this is a well thought out, like deterministic approach that we're taking to managing both top line and profitability. Now we have to go like earn our ability to like hit all of the numbers, right? So every day, every week, every month, every quarter, we need to go out and deliver. And we've had a track record, as a public company, of delivering on our commitments for 5 quarters in a row. We're going to continue to make sure that we are set up to do that going forward while delivering both, what we call high growth, which is greater than 30%, which is our midterm target model for top line growth, and delivering annual operating margin improvement. So it's doing both of those things in concert. And as far as any levers are concerned, we mentioned that we control, obviously, the rate and pace of hiring. We're going to continue to invest in the business. I think it makes sense to just remind folks, we've raised forever capital. We have close to $2 billion of cash on the balance sheet. By the time we turn cash flow positive, we're going to have well -- excess cash well in terms of the level that we're going to need. So we feel like we're in a position of strength. And when we think about growing the base and growing into the opportunity, we feel very good about -- it's about delivering top line and bottom line improvement.

[indiscernible] Average of raise money when you can, not when you have to. You guys have done a nice job. Maybe just in terms of the '24 goal, anything you could say to underscore your confidence in hitting it. And just as we think about the leverage from the various expense lines even from a gross margin perspective, what are the levers that get you there specifically?

Yes. So starting with gross margin. We have invested a lot of time, effort and energy in improving the unit economics of our cloud gross margin. And the reason why I underscore that point is, as cloud becomes a bigger part of our business, the gross margin mix gets impacted by that. Three years ago, before I joined the company, the cloud gross margins were very poor. Let's just put it that way. We've made tremendous improvement. And so some of the things we've done on this, again, being deterministic about it, we have done a lot of engineering and operationalization of how to optimize the engineering environment. We've gone back and renegotiated contracts with all the 3 main cloud service providers. We've been disciplined on pricing and discounting. And we've seen that pay off in a meaningful way. So even as cloud has continued to increase in terms of revenue percentage, our gross margins have been roughly, call it, 69%, 70% on even in the face of...

Mix.

Mix. Exactly. So that's 1 thing. On the sales and marketing side, if you look at the areas that we showed the most improvement -- where we need to show the most improvement. Sales and marketing is definitely the biggest one. So we look at internal metrics, profitability metrics and unit economics, LTV to CAC is a big one. We look at magic ratio. But one of the most important things is productivity per sales head. And for folks who are familiar with the story, in FY '21, it was a year of catch-up investment for us, specifically in sales. And it takes about a year for a salesperson to become fully productive. And so exiting 2022, we will have proportionally more fully ramped and tenured salespeople than ramping sales people, which will be a catalyst for FY '23 and '24 in terms of having more productivity per person that will help with improving sales and marketing leverage. There are other things we're doing on the sales and marketing side of the house such as where -- I've mentioned this before, we're still in the early days of building a real partner infrastructure, and we're going to be getting leverage over time out of those partnerships that we strike with the large SIs, also our relationship with the cloud service providers and other partners in the ecosystem. That will be a way for us to get leverage. We've also made strides at making our product easier to use on the engineering side of the house. We've come up with recipes that make it easy for companies to plug and play. As I was mentioning before, some companies don't know what the art of the possible is. And so we've come up with some recipes that's what we call them, for KSQL as an example. So to get further adoption of that core product for us, we make it easy for companies to go ahead, pick a recipe then start and get going. And we're going to see additional product expansion within accounts, and that's going to be important for us to show leverage there. On the R&D and engineering side of the house, there's definitely room for improvement. I can tell you that R&D -- you can't build a technology company without like investing in innovation. And we've been investing both in platform and in cloud over the years. We're devoting more proportionately to cloud going forward. And so we're going to start to see benefit of that as well. And then on the G&A side of the house, it's just really about scale. And so I would refer people back to our -- both our midterm target model and our long-term target model, which we published that shows the progression there. And then you can marry that back to our comments around exiting Q4 '24, roughly breakeven just to see the progression by line item.

Awesome. Steffan, with that, we're out of time. Congrats on all the success. You guys are killing it. Thank you again for being here. It's always great to see you, but even better in Las Vegas for the D. B Tech Conference.

Thank you so much for having me.

For sure.

All right.

Cool.