Confluent Inc
NASDAQ:CFLT
If you don’t study any companies, you have the same success buying stocks as you do in a poker game if you bet without looking at your cards.
Utilize notes to systematically review your investment decisions. By reflecting on past outcomes, you can discern effective strategies and identify those that underperformed. This continuous feedback loop enables you to adapt and refine your approach, optimizing for future success.
Each note serves as a learning point, offering insights into your decision-making processes. Over time, you'll accumulate a personalized database of knowledge, enhancing your ability to make informed decisions quickly and effectively.
With a comprehensive record of your investment history at your fingertips, you can compare current opportunities against past experiences. This not only bolsters your confidence but also ensures that each decision is grounded in a well-documented rationale.
Do you really want to delete this note?
This action cannot be undone.
| 52 Week Range |
18.67
34.53
|
| Price Target |
|
We'll email you a reminder when the closing price reaches USD.
Choose the stock you wish to monitor with a price alert.
This alert will be permanently deleted.
Confluent Inc
You don't have any saved screeners yet
Hi, everyone. Welcome to the Confluent Q2 2022 Earnings Conference Call. I'm Shane Xie from Investor Relations. And I'm joined by Jay Kreps, Co-Founder and CEO; and Steffan Tomlinson, CFO.
During today's call, management will make forward-looking statements regarding our business, operations, financial performance and future prospects, including statements regarding our financial outlook for the fiscal third quarter of 2022 and fiscal year 2022. These forward statements are subject to risks and uncertainties, which could cause actual results to differ materially from those anticipated by these statements.
Further information on risk factors that could cause actual results to differ is included in our most recent Form 10-Q filed with the SEC. We assume no obligation to update these statements after today's call, except as required by law.
As a reminder, certain financial measures used on today's call are expressed on a non-GAAP basis. We use these non-GAAP manage measures internally to facility analysis of our financial and business trends and for internal planning and forecasting purposes. These non-GAAP financial measures have limitations and should not be considered in isolation from or as a substitute for financial information prepared in accordance with GAAP. A reconciliation between these GAAP and non-GAAP financial measures is included in our earnings press release and supplemental financials, which can be found on our Investor Relations website at investors.confluent.io.
With that, I'll hand the call over to Jay.
Thanks, Shane. Welcome, everyone, to our second quarter earnings call. Confluent delivered another strong quarter, exceeding the high end of our guidance on all metrics. Total revenue grew 58% year-over-year to $139 million. Confluent Cloud revenue grew 139% year-over-year and represented 34% of total revenue in the quarter.
Confluent Cloud continues to increase as an overall mix of our business and is seeing rapid adoption across our customer base as reflected by strong consumption trends. We're especially proud of our performance given the uncertain macro environment we're currently operating in.
I'll start by touching briefly on this topic, why Confluent continues to see strong demand despite economic headwinds. There are 2 reasons for this durability: First, our product sits in the operational stack, powering applications that directly serve critical business operations and real-time customer experiences. Given this criticality, it can't be switched off without a complete disruption to the operations of the business.
Our 2022 state of data in motion report underscores this, finding that of the nearly 2,000 IT and engineering leaders surveyed, more than 80% said real-time data streams are critical to building responsive business processes and rich customer experiences.
Second, one of the key value propositions of a managed cloud service such as Confluent Cloud is cost savings. Using Confluent Cloud has significant TCO advantages compared to trying to build out internal teams of engineers to attempt to build internal services around open source.
SaaS Institute is a customer that perfectly illustrates these dynamics. SaaS is a marketing analytics powerhouse helping more than 80,000 businesses like Discover, Honda, Levi's and Nestle transform data into real-world intelligence, making their marketing campaigns more targeted, more personal and more relevant.
SAS initially built its real-time data platform on open source Kafka, but they soon ran into scalability issues, from self-supporting open source -- that made it difficult to adjust to changing demand. Plus, the operational overhead and complexity were driving significant costs.
So SAS turned to Confluent for a complete data streaming platform that scales both compute and storage on demand even amid unpredictable ebbs and flows of traffic. With Confluent now as the backbone of their next-generation Customer Intelligence 360 platform, SAS can easily stitch together data from multiple sources to find and act on fresh and timely insights for their customers.
Key to this ability to serve mission-critical use cases and to help customers recognize the cost and agility advantages I described are the underlying capabilities that Confluent Cloud provides. Kafka has become ubiquitous and is the de facto standard for data in motion used by over 70% of the Fortune 500.
But Confluent Cloud is not just a matter of putting Kafka in the cloud. In building Confluent Cloud, we rethought virtually every layer of the stack from how data is routed over the network, how it is processed, where it is placed and how it is stored. This deep engineering investment is necessary to provide a truly cloud-native service that can meet the needs of the most mission-critical use cases and can help customers truly step back from the operations of the service and focus on their applications.
To achieve this, over the last 5 years, we've poured more than 3 million engineering hours into Confluent Cloud. Today, it represents a 10x better Kafka service with a deep competitive moat of hard technology. By making a service that is 10x better than open source Kafka, Confluent lets organizations avoid investments in low-level operations, monitoring and scaling and be able to instead rely on a service that can scale elastically with their needs.
This is what drives the substantial cost savings customers see when they adopt our service. As we shared last quarter, a recent Forrester study identified TCO savings of more than $2.5 million for businesses that use Confluent, translating to an ROI of 257% in less than 6 months.
Another example that demonstrates both the mission-critical nature of our use case as well as the economic value of Confluent Cloud is ETC. A leading electronic toll collection company. To support next-generation congestion management services, ETC collects real-time sensor input for millions of cars and IoT devices across city transportation corridors, totaling 2 billion toll transactions per year.
By collecting and processing this data continuously and in real time with Kafka, ETC produced the first truly predictive dynamic pricing algorithm in the industry. But as their use of Kafka skyrocketed from onboarding new customers, increased traffic congestion and expanding toll and smart mobility projects, so did their total cost of operating and maintaining open source Kafka.
After conducting an internal TCO analysis, ETC moved to a fully managed Kafka on Confluent Cloud. By making the move to Confluent Cloud, ETC saved an average of 20% on infrastructure costs, significantly reduced their downtime risk and was able to reallocate about 50% of their engineering and development talent that was dedicated to managing Kafka to more strategic projects that accelerate innovation.
Our relationship across the software and data landscape remain core to our everywhere pillar of differentiation and are a key part of our go-to-market. We made a few notable announcements on the partnership front that deepened our key partner relationships.
First, we are thrilled to announce the launch of a Confluent cloud reseller program. Organizations can accelerate their adoption of data in motion by purchasing Confluent Cloud directly from the consulting partners they already work with who know their business and can offer localized support.
To start this program, we expanded our strategic collaboration agreement with AWS by joining the marketplace channel program, consulting partner private offers. Now we can work with 17 leading data streaming partners, including slower, mega zone and SBA to make it easier for our customers to unlock the full value of data streaming throughout their business.
This quarter, we were also recognized by both Microsoft and MongoDB as one of their top partners for 2022. We're incredibly proud and thankful for our strong partnerships with cloud service providers and technology partners.
I'd also like to spend a few minutes on work we're doing to accelerate usage for customers at the early stages of their data in motion journey. We've previously discussed our customer growth go-to-market model that builds a product-led consumption-oriented journey down the data and motion adoption path.
The early stage of this journey is particularly critical for customer acquisition and for making Confluent Cloud the default starting point for developers. This early stage of adoption often starts with developers experimenting with pilots and proof of concepts or simply learning the new technology.
At this stage, it's critical for the onboarding process to be low friction. So a developer can instantly gain full access to the power of our platform with minimal disruption. To make this process even easier for developers, I'm very pleased that towards the end of our first quarter, we removed the requirement of entering credit card information for the free trial of our product.
This paywall removal is a strategic move that aligns well with our customer growth go-to-market model, allowing us to reduce the friction for developers to test our product, grow usage and progress to the production stage. And we are already seeing strong returns at the top of our funnel, as evidenced by the accelerating growth in Q2 sign-ups, which are up more than 130% year-over-year and up more than 50% sequentially.
This paywall removal has been incredibly successful in increasing sign-ups, but it has also created some short-term noise in our total customer count metric. Users who would have incurred small amounts of spend had been previously counted as customers in their initial trial phase will now show up as just sign-ups, not paying customers, which impacts our customer count growth in Q2.
This means that the change has eliminated a large chunk of preproduction customers paying us an average of less than a few hundred dollars per quarter, creating a reset of our pay-as-you-go customer count. Reset of customer count aside, it's unquestionably the right strategy for our business as our customers can now test drive Confluent risk-free. And for us, the reduction in developer risk and friction drives easier land and ultimately more paying customers as the larger cohort of trials leads to sticky production applications that grow and expand at scale.
And finally, I'd like to share that after a 4-year impactful run at Confluent, Ganesh Srinivasan will be stepping back from his role as Chief Product Officer. I will be acting as interim Chief Product Officer as we search for a new leader. Ganesh, we wish you all the best, and thank you for your many contributions.
Thanks again for joining us today. We remain very confident in our market opportunity and positioning headed into the second half of the year. And we look forward to seeing many of you on the road in the coming months. including at Current, a new data streaming industry event we're hosting in October.
With that, I'll turn the call over to Steffan to walk through the financials.
Thanks, Jay. We delivered a strong second quarter, exceeding the high end of our guidance for the fourth consecutive quarter. Key highlights include strong Confluent Cloud growth, best-in-class net retention and significant margin improvements, which are a testament to the team's performance.
RPO in the second quarter grew 81% year-over-year to $591.3 million. Current RPO estimated to be 62% of RPO was approximately $364 million, up 62% year-over-year.
Total revenue grew 58% year-over-year to $139.4 million. Subscription revenue grew 62% year-over-year to $127 million and accounted for 91% of total revenue. Within subscription, Confluent platform revenue was $80 million, up 36% year-over-year and accounted for 57% of total revenue.
With Confluent platform, we're positioned to address the broader opportunity around hybrid cloud where customers bridge between on-prem and multi-cloud environments using Confluent platform and adding Confluent Cloud over time.
Confluent platform subscription model also adds more visibility to our revenue streams. Confluent Cloud revenue in the quarter exceeded our expectations, up $8.1 million sequentially, representing our largest net revenue growth to date. On a year-over-year basis, Cloud revenue grew 139% to $47 million and accounted for 34% of total revenue compared to 22% of revenue a year ago.
And for the third consecutive quarter, Confluent Cloud accounted for more than 50% of new ACV bookings. Our strong cloud performance was driven by healthy consumption across industries with particular strength in technology and financial services.
Turning to the geographic mix of revenue. Revenue from the U.S. grew 54% year-over-year to $87.6 million. Revenue from outside the U.S. grew 64% year-over-year to $51.9 million.
Turning to customers. The growth in our large customer base continue to be robust. We ended the quarter with 857 customers with at least $100,000 in ARR, up 39% year-over-year and 107 customers with at least $1 million in ARR, up 53% year-over-year. Our 100,000 customer cohort contributed more than 85% of total revenue in the quarter, in line with historical trends.
Our diversified customer base spans across various industries, including financial services, technology, retail, telecom, public sector, health care, media and entertainment and many more. Total customers ended at approximately 4,120, up 46% year-over-year and flat sequentially. There are 2 components of total customer count: pay-as-you-go and committed contract customers.
As Jay discussed earlier, paywall removal has driven strong growth acceleration in sign-ups, which is a key indicator for the overall health of our customer funnel, and it's a testament to the strong demand for our cloud product. While paywall removal had a short-term impact on the number of pay-as-you-go customers in our total customer count, we're very pleased to see the continued momentum in our committed customer base.
We added 128 net new committed customers in Q2 compared to 113 in the previous quarter and 80 a year ago. The vast majority of our revenue is attributed to our committed customers, which provides a high degree of visibility into our revenue in any given quarter.
Dollar-based net retention rate in the quarter remained above 130% for the fifth consecutive quarter driven by 90%-plus gross retention and strong expansion across both of our product offerings. NRR for cloud was substantially higher than the overall NRR for the company, and NRR for hybrid customers continue to be the highest.
Moving on to gross margins and profitability. I'd like to note that I'll be referring to non-GAAP results unless stated otherwise. Total gross margin was 70.6% and subscription gross margin was 76.8%. Our platform gross margin remained steady and strong.
Our cloud gross margin improved substantially driven by our continued efforts to optimize hosting costs and improved pricing with our cloud service providers, which offset some of the headwind of a higher cloud revenue mix to total gross margin.
The concerted efforts and focus on improving the unit economics of the cloud business have been paying off and will continue to drive efficiencies in the future. In the near term, we continue to anticipate total gross margin to fluctuate near our midterm target of approximately 70%.
Turning to profitability and cash flow. Operating margin improved 8 percentage points year-over-year to negative 33.5%. The improvement was driven by revenue outperformance, improving sales efficiency and our focused efforts to proactively manage spend across the organization, such as controlling the rate and pace of hiring.
Net loss per share was negative $0.16, using 278.3 million basic and diluted weighted average shares outstanding. Free cash flow margin improved approximately 25 percentage points year-over-year to negative 26.5% driven by strong collections. We ended the second quarter with $1.96 billion in cash, cash equivalents and marketable securities.
Now I'll turn to our outlook. We are raising our revenue and profit guidance for Q3 and the year, but the magnitude of the raise has been tempered by the current macro dynamics. Towards the back half of June and through July, we saw an increased scrutiny on deal approvals.
We believe this is driven by customers' cautious view on the current macro environment balance with their need to continue with the digital transformation initiatives with data in motion being a must-have capability. We're assuming this dynamic continues through the rest of the year and have estimated approximately $2 million to $3 million negative impact on our Q3 revenue guidance and approximately $4 million to $6 million negative impact on fiscal year 2022.
We've adjusted our spending levels in the second half to ensure we meet our operating margin targets and incremental investments we make will be in the highest ROI segments of the business. We'll continue to monitor and course correct if the conditions change materially and action is warranted.
Our goal in the midterm remains delivering high growth with annual margin improvements and turning non-GAAP profitable exiting Q4 2024.
Turning now to guidance. For the third quarter of 2022, we expect revenue to be in the range of $143 million to $145 million, representing growth of 39% to 41% year-over-year. Due to our Q2 sequential outperformance, we expect Q3 sequential cloud revenue net adds to be between $8 million and $8.5 million, and we expect non-GAAP operating margin to be approximately negative 33%, non-GAAP net loss per share to be in the range of negative $0.19 to negative $0.17, using approximately 282 million weighted average shares outstanding.
For the full year 2022, we expect revenue to be in the range of $567 million to $571 million, representing growth of 46% to 47% year-over-year, non-GAAP operating margin to be in the range of negative 35% to negative 34% and non-GAAP net loss per share in the range of negative $0.73 to negative $0.69, using approximately 280 million weighted average shares outstanding.
Turning to free cash flow. We changed the structure of the payout for the annual bonus from 1 lump sum payment in Q1 '23 to 2 payments, one in Q3 2022 of approximately $14 million and the remaining payment in Q1 23. Due to the timing of these payments, Q3 and fiscal year 2022 free cash flow margin will be lower than originally anticipated, while FY '23 free cash flow margin will be better than previously anticipated.
Before turning to Q&A, I'd like to invite you to join our investor session at Current 2022 on Tuesday, October 4, in Austin, Texas. We'll provide an update on our strategy, product and customers. To join in person, please contact IR for the registration information. The program will be webcast live on our IR website beginning at 2:00 p.m. Central Time.
In closing, our second quarter results underscore our ability to drive high growth with increased operating leverage. While the near-term macro environment is uncertain, the secular tailwind for data in motion is firmly intact. With a market-leading data streaming platform and a track record of delivering on our commitments, we're well positioned to drive durable growth and improve profitability ahead.
Now Jay and I will take your questions.
Thank you, Steffan. [Operator Instructions]. And today, our first question comes from Jason Ader with William Blair followed by UBS.
So yes, I wanted to ask you about the macro impact, just starting out implementing Kafka requires some planning and requires some process change. Is that where the impact is as customers sort of maybe slow down a little bit on the pace of some of those projects? And then what -- how do you expect that to play out as we go through Q3?
Yes. I mean what we actually saw was continued strong demand, and I would say some sporadic delays in deals. So I would say maybe more finance inspection, a little bit longer, another round of review kind of inspection of TCO. We saw that across geos and segments. So it wasn't really a localized thing, but it wasn't hugely widespread. And I think all that is factored into the guidance.
So we kind of took a look at that and we said, "Well, what do we expect going forward?" I think probably more of the same. I think we haven't seen any of those deals that kind of went through further review be lost, but they have been delayed. And I think that's a testament to the kind of overall work we've done on TCO and making sure that, that story is strong.
And then a quick follow-up just on use cases. You talked about some of those on the call. Are there any examples that you can think of in the quarter where you guys were like, "Wow, that's a really cool use case that we've never seen before," or maybe a vertical. Maybe in the past, it wouldn't be obvious that there would be a use case for Confluent. Just kind of curious about that.
Yes. I would say the trend I'm the most excited about is kind of that move to some of these really central use cases where it's kind of broad across the company. We've seen a number of organizations doing that, including some of the customers that started on premise as they graduate to the cloud, taking that on. So that was probably the most exciting trend I noticed was just some of these early on-premise folks now going big in the cloud and seeing some of the results of that.
We'll take our next question from Karl Keirstead with UBS, followed by Morgan Stanley.
So maybe 2 questions. Steffan, you mentioned that there was outperformance on the cloud side in 2Q, and that's why you were assuming a roughly similar sequential improvement in 3Q. Was there anything unusual, Steffan, you'd call out on the cloud performance in 2Q? Any pull in, et cetera? And then I've got a quick follow-up on the platform business.
Well, we're very pleased with the record performance in Q2 sequential for cloud. We didn't see anything unusual or a pull forward. We just saw a broader adoption in consumption patterns across really our entire customer base, which is very good to see.
So we also gave that additional guidance for Q3 to help calibrate and level set folks' expectations. And we're still calling for sequential growth for each of the orders between Q2 and Q4 of this year. So very healthy growth, and we're well positioned to deliver that.
Got it. Congrats on that number, Steffan. And then on the platform side, when you initially set guidance for '22, your implied outlook for the platform business I felt like was a little bit conservative where perhaps you were prepping us for more moderate growth on the platform side yet goodness on the cloud side.
Yet, your platform business actually grew at a pretty healthy clip. It grew sequentially by $5 million, which is pretty similar to what it did in the year ago quarter despite that pretty strong cloud growth. So we really didn't see any sort of migration or trade-off there. Anything that you would call out as maybe helping that platform number in 2Q, where maybe we just need to keep that in mind as we model the platform segment in the second half, Steffan?
I think it's an important point here. The implications of us building a central nervous system for our customer base really means that we have to be wherever our customers' data resides. So that means on-prem, in the cloud, in hybrid cloud environments. And we feel like we are very well positioned to do that.
Our CP business in particular, is going to continue to grow. But as you look at the new ACV percentages that are coming in and how much is cloud versus platform, over 50% of new ACV has been cloud for the last 3 quarters in a row. So by definition, we are going to see somewhat more moderated growth coming out of Confluent platform.
But the biggest takeaway is the Confluent platform business is important because it gives us an opportunity to upsell and cross-sell Confluent Cloud on top of those CP-only environments. And you look at our NRR numbers for total NRR and then also for hybrid and cloud. And the hybrid and cloud and our numbers are greater than the total company average.
Got it. Okay. Well, nice job, Jay, and Steffan. I know it's a tough demand backdrop, so well done.
Thank you.
Thank you.
We'll take our next question from Sanjit Singh with Morgan Stanley, followed by Wells Fargo.
Can you hear me?
Yes.
Okay. Sorry about that. I'm not sure I can see the video, but that's okay. There we go, start my video. I really appreciate the framework on the guide, Steffan. That was super helpful and gives us a way to sort of map to how you're thinking about the environment.
Jay, I wanted to take -- I want to get your opinion on how this softening demand environment, this downturn might compare to what we saw in sort of 2020. And there were some -- in some industries, some sporadic customers sort of downgraded to Kafka. And given what you said on your earnings in your script around Confluent Cloud being 10x better than Kafka, what do you think the risk is in that is downturn that customers would choose to potentially move from paid Confluent either platform or cloud to open source Kafka? How has the dynamics have changed in 2022 versus some customers choosing to go that route during the height of the pandemic?
Yes I think the dynamic has changed significantly, and I would say really for 2 reasons. Like the first one is what I touched on, which is actually our product is much better. And you would think of that was just a few years ago. But Confluent is a young company. So maybe as a percentage of our life, we've done a lot of development in the last few years, those kind of like pillars of differentiation around being cloud-native and complete and everywhere. Those are very real, and it's meaningful to customers.
And that analysis of TCO and our ability to communicate that effectively and believably to customers has gotten much better. So I think that's the first reason is actually it doesn't make sense to get off of a cloud platform and go try and hire a team of engineers. And that's not get a pencil out positive by any means. So I think that helps us significantly.
The second thing that I think has helped us is really just the maturity of the field organization, start to finish from account management to customer success. We're just much better at helping to support our customers and getting them through to the production use cases that are the thing that's going to be sticky. And I think that shows up in the results as well.
Got it. And then my follow-up was sort of around the announcements on getting rid of the paywall. It makes a ton of sense in driving that -- those sign-ups. And so I think your top of funnel is probably going to see a lot of healthy growth.
Historically, in terms of customers making it to that $100,000 threshold, what does that sort of conversion look like? And what sort of your view or optimism that increasing the sort of top of funnel will help you source some customers that can make it to that $100,000 cohort?
Yes, it's very high. I mean the -- in the area we're in, it's very important that we have the open source. And in many ways, the kind of self-service cloud is like the open source as that kind of frictionless thing you can build against.
So I would say a significant portion of our customers either come from that open source usage or come from self-service with cloud, which we see as kind of the future of that open source adoption. So it's actually quite important to plant those seeds.
Sometimes that path is a little bit winding. That first use case may just be some developer playing with it and trying some stuff out. It may be several hops in a totally different person who signs up and does it later, but it's still actually really important.
And that's why we have that continued investment at the top of the funnel versus just kind of going after enterprise, where there's a lot of dollars. You have to have that broad spread. And the push for us is there's so many open source Kafka users out there. We want to go get them all. And so these are the investments that we think can set us up to do that over time.
We'll take our next question from Michael Turrin with Wells Fargo, followed by Cowen.
I don't know if the video has been turned off. But you added record levels of new cloud revenue this quarter. It sounds like the commentary suggests you're expecting that can hold even with some of the impacts you're characterizing in June and July.
I'm just wondering if there's any -- if this is just sort of a function of the broader mix shift towards cloud, or if there's any change you're observing in how customers are adopting Confluent technology you'd call out. I'm imagining the flexibility could prove more attractive. So I'm just wondering how much of this is just the direction the market was likely to head anyways, or if there's anything more recent to call attention to the...
Yes, it's a combination of both. I mean I would say, quarter-to-quarter, it's not like the dynamics of the product or the market changed dramatically. So there is just some seasonality to how this flows if you look at Q1 versus Q2, and we called that out previously.
But yes, if you step back and you look over a multi-quarter trajectory, I think there's a significant mindset shift in a lot of these teams that would be our customers to move away from kind of hiring a lot of very expensive engineers around the internal infrastructure to like trying to find really high-quality cloud services and build against us over the long term.
And I think that the customers who are doing that are actually getting the value back much faster. And they're able to build and develop quicker. And they're able to apply the resources to the thing that builds real unique competitive advantage for that company versus kind of lower-level infrastructure.
So I think that trend is absolutely a tailwind. And as we've gotten to scale and have really checked a lot of the boxes on security, reliability, et cetera, I think there's a pretty high bar there for this kind of infrastructure you have to cross before you can start to pick that up.
As we've done that, I think we've started to benefit from that as well where it becomes far less risky to work with a company like Confluent. I think that's another tailwind. And then the adoption of data in motion overall continues, right? And those are longer trends around just the use of data and the integration of software and how that's driving these more operational kind of run the business use cases.
That's great. Just a quick follow-up, similar line of questioning. You had some useful comments around the hybrid use case and the expansion retention rate there holding even stronger than the 130% you're seeing overall. Is that something you're finding repeatable, the openness of Confluent and this ability to work across multiple environments? Is that something you can...
Yes, absolutely. Yes, we keep coming back to it because I would say out there, there's definitely a set of people who see our Confluent platform to Confluent Cloud as a kind of transition, where customers will delete the Confluent platform and install the Confluent Cloud.
So it's actually a little bit different from that. Like this has to span their different environments. So it's typically an expansion out into these cloud environments as projects are happening there in each of these customers.
And that ability to bridge between is really strategic. Like one of the key problems you have to solve if you want to get meaningful amounts of workloads into the cloud is how those can tie back into all the legacy data systems and applications on premise. And we think that we have a key strategic role in that.
And that's actually quite important for a lot of these organizations. It's just not practical for them to turn over their full suite of software investments into cloud native applications in any kind of short period of time. Some of these customers have significant investments that are some generations back on mainframes, right?
And so it's all going to make it to the cloud, but that time period is fairly elongated. And how it all works together as one company in the interim period is incredibly important to them.
So this ability to both bridge across, be able to connect to the new and the old and move to the more real-time streaming event-driven architecture they want anyway is a big deal.
We'll take our next question from Derrick Wood with Cowen followed by JPMorgan.
Great. I guess my first question, Jay, we get a lot of questions around the puts and takes of your cloud consumption business. And what we've heard from Snowflake as a comparison is that a lot of customers will kind of pump a lot of analytic workloads into the system and drive a lot of consumption and growth and then go through this kind of optimization cycle where consumption flattens and then start to see more workloads grow again.
And maybe we're in a little bit more of an optimization cycle right now. But just curious, do you guys have a similar dynamic where you go through these optimization cycles? Or because you're more centered around transactional workloads, do you not tend to see that kind of behavior?
Yes. I mean certainly, any customer, as their spend gets to scale, looks at, hey, is this well spent? And are we using it inefficiently? We have a lot less of that, I think, than things that don't serve production use loads -- work loads.
Like production workloads usually kind of come optimized as it were. They have development days, where you're kind of building it. Some of the data warehouse usage is more ad hoc reporting, things that are kind of thrown together. And so I think you do kind of accumulate more and then prune it back.
So maybe in that respect, we might be more comparable to something like MongoDB, which is serving more permanent applications in our usage patterns. And I haven't looked enough at the dynamics of Snowflake to say if that comparison is accurate. But in terms of us, that is -- this idea of production workloads is actually quite important.
And in conversations with investors, there has been this idea that like, oh, as soon as there's some economic pressure, like consumption is going to drop -- that hasn't been what we've seen and it's certainly not what we saw this quarter.
Okay. And Steffan, one for you. Just looking at the Q3 guide and knowing that you're expecting $8 million to $8.5 million in cloud quarter-over-quarter build, that would suggest downtick in another line item, whether it's platform or professional services. And platform is obviously the bigger revenue piece.
So what would cause platform revenues to be down sequentially, especially kind of in light of that being a stronger government quarter, which tends to be -- purchase more platform licenses?
It really comes down to how the mix of business is coming in. We have one sales force, and they sell -- they're selling a solution, right? The fact that we've had 50% of new ACV coming in that's been cloud for the last 3 quarters in a row is a dynamic at play.
We still think that there's going to be just a healthy business coming out of Confluent platform for Q3. But the way that the models are working, we're anticipating that Confluent Cloud will continue to grow very helpfully. Confluent platform will be a contributor for sure, but it won't be as pronounced in Q3. Typically, we would see a little bit of a bounce back in Q4 because that's the end of the year buying cycles and that's when Confluent platform typically has a very strong sequential quarter.
We'll take our next question from Pinjalim Bora with JPMorgan, followed by Goldman Sachs.
Can you guys hear me?
Loud and clear.
Okay. For some reason, I can't start the video. Oh, here it is. It seems like everybody has the same problem. I wanted to ask you guys about the margin guidance. It seems like OpEx is in the second half coming down more the year-over-year growth, at least when I look at it. I think you talked about a little bit on hiring. Help us understand the levers that you're pulling in the second half to optimize OpEx.
Yes. So we haven't done anything overly aggressive. I mean we see a huge growth opportunity ahead. And so for us, it's about high sustained growth and increasing improvements in operating margin and other unit economic metrics.
So we did make some adjustments on kind of the timing of hires. We did the kind of normal readjustment we would do in the second half of the year on just moving spend to higher ROI programs and efforts.
Overall, that did add to the savings we would expect. And we've seen already some of that play out, which is great. So we're happy with the improvement that we had kind of plotted out our trajectory. And this gets us a little bit towards our goal, even a little faster than we do.
So just to be clear, in 2020, when COVID hit, you had kind of paused hiring completely, right? So this is not anything close to that.
Yes. No, nothing like that.
Understood. One technical question, I guess, for Jay. The ZooKeeper dependency, I thought was gone from Kafka, 2.8 or something like that. I look at Amazon MSK. They still talk about ZooKeeper.
So I'm trying to understand, are they running a few steps behind in terms of Kafka version? And does that -- ZooKeeper dependency now gone, does that help you in your cloud gross margins?
Yes, yes. So it's -- that was a dependency in the open source that people don't like. And so we've done work to take a lot of these workloads. This is about how do you manage the internal meta data and how you help that scale up for very large use cases. So that's kind of how it helps as well as just the simplicity of our users getting started.
As to Amazon's product, they do tend to be a few versions behind. The current stage is we've built a self-hosting Kafka with no dependency on ZooKeeper. That's going through a set of iterations that make it kind of increasingly production ready, and it's on a great path. So we're seeing already kind of early folks who are adopting and playing with that. And we expect that will become kind of the more mainstream approach over time.
Will that help us? Yes, it will definitely lower the cost to serve a particular use case because we won't have that dependency running in our cloud.
We'll take our next question from Kash Rangan with Goldman Sachs, followed by Bank of America.
Jay and Steffan, great job dissecting the quarter and your outlook. I'm curious, during an economic downturn -- let's just say what we're going through is an economic downturn. Companies redefine value proposition, data architectures. It generally takes a bit of leap of confidence to change these things. How are you adopting, if at all, change in your go-to-market approach in order to get people over the fence?
Clearly, you have done a great job with the technology that it's less of a contention Kafka versus Confluent commercial. But I'm curious if you are thinking about changing go-to-market strategy, assuming that these tough times will last for a long time, higher interest rates, higher inflation, higher hurdles, that sort of thing. You see what I'm getting at.
Yes. It's a really important question. So we show a little bit of that customer growth go-to-market graph of kind of the stages for our customer. This is why the early stages are so important. When we land with the customer, we're not landing for an architecture change. It's not like, hey, let's install a new central nervous system.
We're landing for use case. Like we're landing for one useful, valuable application. And then we want to grow from there in an organic way that also solves for that bigger picture. That's something that we've put effort into from the early days of the company to make that kind of gradual pack easy to realize.
And I think it's really important. I think if we didn't have that, we would suffer from what you described, which is like, hey, do we want to make some big IT investments and some change in data flow? Maybe, maybe not. It's very important that each step on that journey, deliver value projects by project. That's critical for us. And so I think that's the first thing to understand.
The second thing is, yes, how do we adjust in a more difficult environment. I would say probably the biggest adjustment, we haven't seen the need for this. We're still seeing lots of new use cases happening. But if we see a slowdown in net new use cases, we would shift to more of the conversion of open source Kafka, where there's a very significant installed base of Kafka. I shared some stats on that in the last earnings call.
And so the migration of that over to Confluent Cloud is a significant pool of usage to fish out of. And so if we see there's some slowdown overall in software projects, obviously, we'll still get the ones that are happening, but we would compensate by focusing on that more.
Both for the new use cases and for those conversions, they rely on that underlying TCO analysis of like, hey, is this better? Should I build on this cloud service? Is that going to be both a better product but also something that's cheaper and a better pay off to me when I consider all my costs?
We'll take our next question from Brad Sills with Bank of America followed by Deutsche Bank.
Great. I wanted to ask a question around -- sorry about the echo here.
No problem.
Is that better? I think there's something wrong with the audio.
Yes. We can hear you okay.
Okay. All right. So my question was on net revenue retention, cloud outpacing net revenue retention in total. What is it about cloud deals such that the velocity to expand is greater than overall airport platform?
Yes. Yes. So for anybody who missed the last bit, I think the question was why is cloud NRR higher. It's really both sides, right? So the retention is better. If you're building it against a running cloud data service, as that gets to larger and larger scale, it gets harder and harder to move off of it, especially for something like Confluent where there's many applications in different parts of the business that have attached to this. And you would have to all agree to be able to switch together.
And in that respect, it doesn't suffer from as much competition from the open source. That kind of becomes a very difficult path to pursue. So that's on the retention side.
On the expansion side, I think it's mostly just about the ease of realizing the value. One of the things for a large enterprise that's often hard is spinning up a net new production data. Some big distributed system, hiring the team of very specialized people, getting all the servers, being able to scale that, doing that reliably enough that you can build against it and then all the teams can kind of count on it as a utility, that's actually really challenging.
And even if they master that core bit, getting to kind of the full ecosystem for data in motion, what we would offer in our product is usually out of reach. So all the connectors and stream, all the stuff that teams need to be successful, they typically won't have that.
So then why does cloud expand faster? It's because it's much easier to just go and build your application and get the value up. And so those are the 2 factors that I think drive it.
That's great. And then one more, if I may. Any color on greenfield versus replacement-type work loads that you're seeing or applications. I think historically -- applications that drive the growth. Like you talked about increased connectivity to legacy infrastructure. Does that mean you're starting to see more of that replacement cycle come in...
Yes. We've always had an element of replacement. It's typically not exactly 1 to 1. But I would say that there's a replacement of legacy messaging and middleware layers, replacement of more batch database processing, replacement of some legacy ETL or data movement products.
When we're doing analysis with customers, that's typically where spend and budget is freeing up from. I don't think we've seen a big shift in that mix. It is some new stuff and some freeing up of old stuff.
And oftentimes, the new stuff is replacing some of the old steps. I would say broadly, that mix is similar to what it's always been.
We'll take our next question from Brad Zelnick with Deutsche Bank, followed by D.A. Davidson.
Congrats on the strong growth, guys, and nice margin progression. Steffan, the change in timing on bonus payments, can you just share what inspired that? And are there any changes in your philosophy around equity versus cash comp? And how is that evolving maybe just given what seems to be an easing labor dynamic out there?
Good question. Thanks for that, Brad. Well, we are looking at overall compensation for our employees This year, we originally had decided to move to a semiannual payout from a bonus -- we've decided to move to an annual payout from a bonus standpoint for this year.
And given kind of the dynamics that we've seen just with stock price and with other things that our employee base is going through, we wanted to be employee-friendly and effectively go through and have a semiannual payment structure for this year, instead of an annual payment structure.
And that was a conscious thing that we did. And that was well received by our employee base. And it has really no impact to free cash flow over, call it, a couple of year basis because it's really a timing of cash flow payments. So we feel good about that.
And then as far as equity versus cash, we've been very selective in terms of making sure that our top talent is taken care of. And we also are making sure that our employees are compensated according to market.
So that's something that we will -- we have been doing and that we'll continue to do going forward to make sure that we have a fully engaged employee base. And I think a follow-on question maybe around attrition. Attrition has been trending better than we expected. And we feel good about like the health and well-being of our employees.
Excellent. Congrats.
Great. Thank you.
Thanks, Brad.
We'll take our next question from Rudy Kessinger with D.A. Davidson, followed by Barclays.
The total customer count, I know there's -- the pay-as-you-go guys, that causes some messiness there. But the total count was flat quarter-over-quarter. So have you seen a slowdown in those new pay-as-you-go trials? I guess maybe just given the macro, people are more hesitant even if it's free to start new projects?
Yes. We haven't seen any kind of macro impact. There's enough changing in that area that it would be very hard to disambiguate. At the top of the funnel, we actually saw very strong growth, so like 50% more sign-ups. Those people getting into usage is very strong.
We think over time, that will lead to increasing customer adds at an even better rate. So we feel pretty good about where that will go. It's a new funnel, so we have to work through all the details of that. But it's actually, I think, a really important strategic move for us.
Got it. And then you mentioned -- just a follow-on to that, Rudy, just quickly. I know your question was more on pay-as-you-go. But I do want to just point out that from a committed customer standpoint, we did see really healthy growth, more ads versus last quarter and last year.
And why is that important? The vast majority of our business comes from our committed customer base. And so that's a dynamic that we just -- we wanted to reinforce as well.
Got it. You mentioned the more scrutiny on the new deals. I guess I'm curious with your existing customers. Have you noticed any notable amount of customers that have, I guess, slowed down their expansion plans with Confluent? I guess when you look at the impact to the year, the $4 million to $6 million impact to the guide for the year, is that entirely from slower new customer adds? Or is any of it from your existing customer...
Yes. I would say, generally, scrutiny can come on either expansion, new projects in the same company or new projects in a company we don't yet sell to. It's a little bit more pronounced probably on the new logo acquisition because that's probably when they do the most analysis. But yes, it's across both of those.
And that -- what Steffan gave on the guidance kind of wraps up our summary of, well, netted all out, what does it add up to, probably something around there.
Our final question today comes from Raimo Lenschow.
Quick question, more strategic question because I'm the last one in the queue here. Jay, if you think about your cloud mix, we're moving to more like the range, et cetera. What does it mean from an organizational structure in terms of how you're selling, how you're doing resources?
I know you talked earlier about the importance of having both. But like sales motion, et cetera, is there a change that kind of you need to start thinking about as you're becoming more and more like a proper cloud company? And then I have one follow-up.
Yes. Yes. There's a significant amount happening around that and a significant amount that has already happened. So often, these adjustments actually have to happen before you get the results. So it's not like we've been doing nothing and suddenly this happened and now we're adjusting.
But I do think there's continued advancement there. So in particular, I think really leaning into the consumption aspect of the business. The fact that customers can expand their usage 365 days a year without -- and do transactions only as needed to kind of take the commit up, that's actually quite important.
That makes it much easier to add workloads, to add use cases, to spread to other teams. Leaning into that and how our team think and what we measure and monitor in customer usage in how the sales team is motivated, that's incredibly important. We've done that in some of the segments already, but it's absolutely a journey across everything.
And I think that's an opportunity for us to both accelerate growth and drive more efficiency by really focusing on the key things that matter to our customers in like delivering use cases to production and getting the value out of it and then to us in driving consumption revenue.
Yes. And then last question for me was on -- you mentioned earlier, in tougher times, you might want to lean more into the Kafka base. Where are they at the moment in terms of deployment models? Are they kind of self-posted in the cloud? Are they kind of on-premise? And how can you kind of help to kind of drive that switch over?
Yes. Yes. So the use cases are where they are, right? Many of them are on-premise. Many of them are in the cloud. The ones that are -- we can convert over easily are the ones in the cloud. Those are the ones we can go get on Confluent Cloud.
We think there's a net migration of stuff into the cloud overall. And that's a natural cutover point to move to some different way of doing it. I would say there is a mindset shift just happening in all types of companies, tech companies, enterprise companies, away from kind of a do-it-yourself mindset on some of this infrastructure, which I think is just proven kind of expensive and unreliable and harder to scale and just a drag on productivity.
And so I do think we're just seeing that increasingly where companies a year ago who would have told us like, hey, we're doing it ourselves, we're happy are now like, okay, this doesn't make any sense. Like we have this large and growing team that kind of can't keep up with the needs internally. Like we need some solution for it.
Okay. Perfect. Makes sense. Congrats from me as well.
Yes. Thanks so much.
All right. That concludes today's earnings call. Thanks again, everyone, for joining us. Take care.