rewrite this title Confluent Cloud Growth Becomes All Consuming – Nanalyze

The business model that perhaps best reflects ‘Merica consumerism is the all-you-can-eat buffet. For a flat fee, one can gorge on everything from E coli-tainted salad bars to pizzas with crusts made from reconstituted sawdust to Indian curries swimming in more oil than the Exxon Valdez spilled. The economic theory behind this bacchanalian business is that these so-called restaurants can maximize profit by carefully managing food costs (feature loads of breads, rice, and other cheap fillers), customer psychology (offer the illusion of abundance and variety), and operational efficiency (fewer employees who require minimal training). Sure, a few customers might eat their weight in profit, but most will actually under consume compared to costs.

Unless, of course, we’re talking about an all-you-can-eat shrimp buffet. The restaurant chain Red Lobster famously lost $11 million in Q4-2023 alone when it launched its endless shrimp deal that allowed customers to consume as many crustaceans as they wanted with no limits. The debacle accelerated the company’s financial meltdown, leading the seafood chain to file for bankruptcy last year.

The Consumption Model 

So, what does soggy shrimp have to do with emerging technologies? Many software-as-a–service (SaaS) firms traditionally operated on a similar all-you-can-eat business model. Salesforce (CRM), for example, was a pioneer of this approach, offering per-user licenses for everything on the menu – charging a fixed, recurring fee to customers to use their software and services. However, not every customer wanted the stale breadsticks or soft-serve ice cream, and it was difficult for companies to monetize their most gluttonous users. Over time, the paradigm shifted to tiered pricing (multiple packages with different features/limits) and hybrid models (combining subscriptions with usage-based elements). Eventually, with advances in cloud computing and data analytics, SaaS providers gained the ability to precisely track and bill for every click and compute. 

The model is mostly seen as a win-win for both sides. Customers get more flexibility, paying only for what they need and want. SaaS providers cash in on high-usage customers while enabling smaller clients access to their platform, potentially expanding their total addressable market (TAM). Snowflake (SNOW) is considered the gold standard of the consumption-based model, charging customers based on the compute, storage, and data transfer resources they consume. The consumption-based model is now believed to be the dominant one in the SaaS landscape, with more than 60% of SaaS companies adopting some form of usage-based pricing as of 2023, up from 27% in 2018. 

Confluent (CFLT), a leading SaaS firm that specializes in managing, analyzing, and utilizing streaming data in real-time, joined the revolution in 2019. Back then, about 10% of its revenues were based on usage. As of Q1-2025, 55% of $261 million in subscription revenues are cloud-based and rely solely on compute, storage, and data transfer usage of its data streaming platform. For all of 2024, subscription revenue reached $922 million (up 26%) and cloud revenue hit $492 million (up 41%). Let’s dive further into the latest and greatest moves by Confluent since our last article on this growth stock.

Confluent Data Stream Platform

First, let’s briefly recap what Confluent does. It provides a platform that turns data into real-time streams, allowing businesses to analyze and act on information as it’s generated, rather than waiting for it to be stored and processed later. At the heart of Confluent’s technology is Apache Kafka, an open-source tool for streaming data that was originally developed by the company’s co-founders. 

Confluent has since bolted on additional components, such as Flink, which lets users analyze, transform, and enrich data in motion, supporting complex real-time analytics and AI use cases. Last year, the company introduced TableFlow, which helps operational data flow seamlessly and securely into analytics engines like Databricks and Snowflake. This reportedly results in more accurate insights, reduced data processing costs, and the ability to power next-generation AI applications. Flink, TableFlow, and other tools are part of the company’s Data Streaming Platform (DSP), a unified solution for real-time streaming, connecting, processing, and governing of data.

Since the end of 2024, management has insisted that DSP will be the engine driving the company’s future growth. Last year, it accounted for 13% of cloud business and is “growing substantially faster” than overall cloud.” In Q1-2025, 13 of the 16 new $1 million-plus annual recurring revenue (ARR) customers adopted DSP components. Overall, Confluent has 210 customers with an ARR of more than $1 million as of the first quarter of 2025, up from 168 a year ago (representing 25% growth). This follows a strong 2024 in which the company ended the year with about 5,800 customers, after adding 840 new accounts, nearly double the total increase from the previous year.

Confluent Customer Case Studies

So, who are some of these customers and what are they doing with Apache Kafka and the more advanced tools within the DSP suite? You’re probably familiar with some of the more consumer-facing use cases, such as Netflix recommendations that update based on whatever Liam Neeson film you’re watching right now. Or social media feeds that constantly update with new content to make you more angry and lonely. Behind the scenes, banks and other financial institutions can use the tech to detect fraud in real-time or retailers can manage inventory in real-time to ensure Cabbage Patch dolls never go out of stock ever again. Never again.

Take the case study of Citizens Bank, which transitioned from open-source Kafka to Confluentʼs DSP. The platform enables the bank to connect data from sources like consumer checking accounts, credit cards, and FICO fraud scores, generating real-time actionable insights. Citizens Bank was able to slash IT costs by 30%, saving $1.2 million annually in fraud reduction, and speeding up loan processing by 40%. In another example, a top-three Fortune Global 100 telecom leverages Confluent to stream data from 70,000 cell towers, optimizing network coverage and potentially giving people covid with 5G unlocking new revenue streams.

Expanding TAM & Decelerating Growth

While Confluent now boasts about 5,800 customers, that represents less than 4% of the estimated 150,000 organizations that rely on Apache Kafka as the real-time backbone of their business. That vast untapped reservoir is the foundation for the company’s now $100 billion TAM. The TAM calculation also factors in adjacent markets that Confluent is moving into, such as stream processing, data governance, and real-time analytics. Of course, the company anticipates AI as being a major market in the future. Management says it is positioning the platform as the “connective tissue” for AI and GenAI applications, feeding real-time data into AI models for context and decision-making.

Confluent is projecting total revenue of about $1.1 billion in 2025, which would represent just 1% total market penetration. However, it’s worth noting that this year’s guidance represents about a 20% increase over 2024, a significant deceleration from 24% total revenue growth the prior year. The projected slowdown underscores one of the big downsides to the consumption-based models: Revenue is more sensitive to customer demand – or lack thereof. In March, Confluent noted reduced consumption among some of its larger cloud customers and expects the trend to continue without a near-term rebound. This reflects the broad slowdown we’ve seen across the SaaS landscape, including decelerating revenue growth for Snowflake.

In addition, the net retention rate has slipped to 117%, down from about 125% when we checked in with Confluent stock a year ago. If DSP Cloud is the future, then we would expect to see more cross- and up-selling, not less. Is there some product cannibalization happening here? Grok perused all the data and concluded, “DSP appears to be driving incremental growth by expanding use cases and attracting new workloads, particularly in real-time AI and analytics.” No evidence of cannibalization is apparent. Maybe the focus simply shifted to all the new customers that the company has added in the last 15 months or so. But the relatively modest increase in total cloud subscription revenue mix to 58% by Q4-2025 suggests some overall softness this year as the macroeconomic headwinds blow. 

Financial Metrics Look Good

We see little reason to panic here. Confluent Cloud growth continues to outpace core Confluent Platform growth. The former, which consists of the company’s on-premises or self-managed offering, actually jumped 18% year-over-year in Q1-2025 to more than $118 million – the strongest Q1 growth in three years. The performance seems partly driven by success with OEM partnerships, particularly internationally. While cloud remains Confluent’s growth engine, the resurgence in the original platform business helps offset turbulence from consumables chop. In addition, about 40% of the company revenues come from outside of the United States, which offers further resilience amid current economic uncertainty, especially domestically. 

There are certainly other positive metrics outside of revenue growth. Gross margins reached 73% last year. More importantly, Confluent is now both free cash flow positive and posting positive EBITDA (based on non-GAAP operating margin). What that basically means is that the company is figuring out how to become profitable and generating more cash than it needs just to keep the lights on. Despite losing more than $345 million last year, Confluent still has nearly $2 billion in the bank, so plenty of runway for takeoff.

Finally, we classify Confluent under “Augmented Reality / Virtual Reality” because it’s a play on the metaverse. The idea is that once a company’s operations are replicated as a digital twin, all the data being thrown off in real time will be fed to hungry AI algorithms for us in areas like predictive analytics. Consequently, a key element of our thesis is the growth of artificial intelligence. As hungry AI algorithms gobble up data so fast that they’re running out of it, the next area of focus will be on how fast they can process new data. Competitive advantages will go to those companies processing new data in real-time and then using it to make better decisions quicker than the competition. Eventually, every company should be utilizing a real-time data processing platform to optimize all aspects of their operations.

Conclusion

Even though some investors are a bit spooked, Confluent’s looking pretty good value-wise, sporting a simple valuation ratio (market cap/annualized revenue) of 7. That’s darn close to our catalog average, especially for a SaaS stock with bonafide upside. The company’s platform is coalescing into a comprehensive offering with opportunities to upsell to existing customers, plus tons of potential with all those Apache Kafka users out there. And with AI driving more demand for real-time data, Confluent is set up for even more growth ahead.