[ad_1]
Confluent, Inc.’s (NASDAQ:CFLT) losses and inability to expand its customer base have been points of concern over the past 1-2 years. The company recently shifted the focus of its sales efforts to try to rectify this and reaccelerate growth, with early results appearing promising.
Confluent’s margins should also improve significantly over the next year as the last stock-based compensation costs related to pre-IPO options are recognized. Despite this, Confluent is still relatively inefficient and is only just now approaching cash flow breakeven.
I last wrote that the rebound in Confluent’s share price in early 2024 was overly optimistic, in large part because Confluent was struggling to expand its customer base and had an efficiency problem. The stock is down approximately 34% since then and may now be approaching a reasonable entry point.
While Confluent’s revenue multiple isn’t overly high given the company’s potential, the macro environment appears precarious. Confluent could easily fall by another 50% in the event of a recession as it doesn’t have profits or positive cash flows to backstop the share price.
Market Conditions
Consumption within Confluent’s large digital-native customer base remains volatile, with weakness observed in June. This was attributed to headwinds from customers optimizing infrastructure usage rather than cutting projects or switching to competitors, though. The revenue growth of many of Confluent’s larger digital native customers continues to moderate, supporting the notion that weak consumption is due to cost-cutting rather than something more nefarious.
OpenAI recently acquired a real-time database (Rockset) to enhance its retrieval infrastructure. Rockset is a relational database that supports dynamic, semi-structured data through special indexing of these files such that they can be used for SQL-based queries. Rockset is useful in areas like streaming, text search, analytics, and vector search. This could suggest streaming will be an important component of generative AI architectures and that Confluent will see a tailwind at some point. Based on a Confluent survey, 63% of IT leaders believe data streaming supports AI progress by building a real-time data foundation.
Outside of any potential AI tailwinds, there is a large and growing data streaming opportunity, with Confluent estimating that its TAM will be $100 billion USD in 2025. The questions at this point are how broadly will be streaming to be adopted and how much of the data streaming market will be captured by Confluent.
Longer-term, my biggest concern is Confluent’s ability to expand its customer base beyond a small core of power users. There are over 100,000 companies using open-source Kafka, and yet Confluent only has around 5,000 customers and is struggling to increase this number. Confluent is largely positioning its managed offering as a lower cost option relative to Kafka, with up to 60% Lower TCO. This is potentially problematic as managing Kafka reportedly becomes more difficult with scale, making Confluent appeal primarily to larger organizations.
Confluent Business Updates
Confluent’s Connect, Process and Govern products continue to grow at a faster rate than the overall cloud business. Connect is the most mature of these solutions, and as a result, is contributing the most to the company’s topline.
Flink is growing at the fastest rate though and should become a core part of the business over the next few years. Confluent’s service is based on Apache Flink, which is the emerging standard for data streaming. In 2023, there were nearly one million unique downloads of Flink and a 43% increase in open job requisitions for Flink developers. Flink was designed specifically for streaming and has the same processing power as a database. Flink now has 1,000 customers, up from 600 at the end of Q1, which could become a significant tailwind as more use cases move into production. Stream processing could also help Confluent to land more customers by increasing the capabilities of its platform.
Outside of product development, Confluent is working with partners to both strengthen its business and drive sales. This includes enabling SI (systems integrator) partners like KPMG, Accenture and EY so that they can help customers to implement data streaming. Confluent also has Connect Confluent, a partner program that makes it easier for partners to build data integrations. This program has driven triple the amount of data traffic from partner integrations since the start of the year.
Financial Analysis
Confluent’s subscription revenue increased 27% YoY to $225 million USD, with Confluent Cloud up 40% to $117 million USD. Confluent Platform growth remains lumpy as approximately 20% of TCV is recognized upfront. While this provided a boost in Q2, it may not be repeated going forward.
Confluent expects $233-234 million USD in subscription revenue in Q3, an increase of 23-24% YoY. As a result, Confluent’s revenue growth in the third quarter will likely be similar to Q2. Roughly $910 million USD revenue is expected for the full year, an increase of 25%.
Confluent’s total customer count increased by 320 in the second quarter, the largest sequential increase in 2 years, which isn’t really surprising given the shift in sales compensation strategy. The acceleration was driven by smaller customers, with Confluent trying to land customers earlier so that they build data streaming into their architecture from the start. Confluent must now demonstrate that these customers can grow into larger accounts.
Confluent’s net retention rate was 118% in the second quarter, below the company’s 120-125% target. The company’s gross retention rate remains in excess of 90%, though. Expansion of Confluent’s product portfolio could support an increase in NRR going forward, as multi-product customers have NRR substantially in excess of 130%.
The number of job openings mentioning Confluent in the job requirements is still fairly solid. A range of indicators suggest that demand growth weakened in the second half of 2022 and is yet to show signs of rebounding.
Confluent’s non-GAAP subscription gross margin was 80.8% in the second quarter. The result was driven by a relatively strong platform contribution and improving Confluent Cloud unit economics. Services gross margins remain depressed, although services revenue is a relatively small contributor to total revenue.
Confluent’s margins and cash flows have generally been improving over the past 3 years. The company remains a long way from GAAP profitability though, due in large part to SBC. Stock-based compensation is a lagging indicator though and the final tranches of Confluent’s pre-IPO options will be recognized as SBC through the first half of 2025. As a result, SBC will soon begin to moderate, moving Confluent towards its net dilution target of ~3%.
Conclusion
Confluent’s shift in sales incentives to encourage customer acquisition appears to be creating results. The company still needs to demonstrate that these new customers can drive consumption over time, though. If this proves to be the case, there is reason to be optimistic about Confluent’s forward growth prospects. Particularly given that stream processing has the potential to be a large business in its own right and is showing early signs of traction in the market.
There is also reason to be optimistic on the margin front, with Confluent’s SBC set to decline significantly over the next 12 months. Confluent is also in the process of transitioning to positive cash flows, positioning the company to begin returning capital to investors. Something that is supported by Confluent’s strong balance sheet.
Confluent’s share is now back to the lower end of its typical trading range over the past 2 years. The company’s valuation is also now beginning to look more compelling, with its revenue multiple considerably lower than the average of its peers when adjusting for growth and profitability. I am still fairly neutral on the stock though as more time is needed for the company to demonstrate it can drive consistent growth of its customer base and that this will lead to improved growth.
[ad_2]
Source link