June 2, 2021
Photo: Index Ventures
On April 22, Confluent confidentially filed its S-1 in anticipation of an IPO on the Nasdaq Stock Market, and released the full details of its S-1 statement earlier this month. The company is planning to offer 23 million shares, priced between $29 and $33, with the goal of raising $759 million, according to a regulatory filing on Wednesday June 16th. That would value Confluent at roughly $8.3 billion. The stock will trade under the ticker symbol CFLT.
Founded by Apache Kafka creators Jay Kreps, Jun Rao and Neha Narkhede, Confluent empowers enterprises to better capture event data on a real-time basis. This data could encompass feeds from internet-enabled sensors on the manufacturing floor, mentions of a certain word or phrase on Twitter or up-to-the-second retail inventory updates so websites accurately reflect what's available in stores. That information is then dumped into a data lake or warehouse for analysis or to feed into machine learning algorithms.
Such a system could prove vitally important to organizations that are all angling to use their own stored information to drive some sort of benefit for the company, like better product demand forecasting or improved customer insights. It's just one of several startups that have gained prominence by helping to replace the on-premises databases of the past with new cloud-based infrastructures.
Confluent's main services are more of a complement to services provided by the likes of Snowflake, as well as AWS, Microsoft and Google Cloud, given it doesn't provide the analytical layer — although Confluent does compete directly with some products from those cloud providers. Its other main rivals are legacy relational database providers like Oracle, messaging vendors like MuleSoft, ETL providers like Informatica and on-premises event data services from Red Hat and Cloudera.
This story has been updated with new financial information.
In the past, companies built their applications above on-premises databases from the likes of Oracle. While the de facto standard for decades, those types of systems are starting to fade in prominence because the data tends to be more static, among other reasons.
Right now, enterprises want real-time data injections to run analyses and get quicker updates on everything from the status of the business to the health of their machinery. As more companies move to the cloud, it's becoming easier to query that data on a constant basis. But the problem is actually getting the data into those repositories.
Many of the older, different applications or systems in use by an enterprise haven't been integrated with each other, which means data is held in lots of disparate parts of an organization. Confluent places something akin to a sensor within apps that helps connect the exact web of information a customer wants and stores it in one location; this allows customers to constantly query the most up-to-date data.
Like other tech providers, Confluent relies on a low-code approach to app development, which makes it easier for non-technologists to use. As the number of systems built on Confluent grows, so does the amount of information flowing through the service, which means the company can quickly expand its presence within an enterprise. It offers Confluent Cloud, a SaaS product that runs on all the major hyperscalers, and the self-managed Confluent Platform.
Confluent is still very much in its growth stage, which is a nicer way of saying it has generated significant revenue growth accompanied by even faster growth in net losses.
Between 2019 and 2020, for instance, Confluent's revenue grew from $150 million to $237 million. While revenue nearly doubled in that period, the net loss more than doubled — Confluent lost $95 million in 2019 and $230 million in 2020, representing a 142% increase.
While Confluent has never turned a profit, it's trending in the right direction. The most recent financial figures in the S-1 show that Confluent generated $77 million in the first three months of 2021. This represented a 51.3% increase from the revenue generated in the same period a year prior. At the same time, the net loss grew only 32% in that time frame, from $33.6 million in 2020 to $44.5 million in 2021.
Sales and marketing expenses have typically constituted the largest operating expense for Confluent. In 2018 and 2019, sales and marketing respectively accounted for 60% and 58% of total operating expenses. But in 2020, general and administrative operating expenses skyrocketed, jumping to $122.5 million from just $24.7 million the year prior. This increase is almost entirely due to the $111.9 million in stock-based compensation expense Confluent incurred in 2020. Some investors might take this as a positive sign, since stock compensation expenses are somewhat removed from direct operations.
With all that money allocated for sales and marketing, Confluent will be closely scrutinized for its ability to attract and retain a diverse client base. The S-1 shows that Confluent has done a good job so far — the number of customers who contributed at least $100,000 in annual recurring revenue jumped from 337 at the end of 2019 to 513 at the end of 2020, a 52% increase. At the end of March 2021, Confluent had 561 such customers, representing a 50% increase from a year prior.
There are still some signs of trouble in the realm of client retention. Dollar-based net retention rate — a metric that shows the relative spend percentage of a set group of clients from one period to the next — has steadily declined over the years. It went from 177% in 2018 to 134% in 2019, and finally to 125% in 2020 — which is considered about average for the industry. Confluent attributes the decline as well to "large initial deal sizes that incorporate potential growth, the impact of the COVID-19 pandemic, and the initial impact of existing customers transitioning to our usage-based Confluent Cloud offering."
One main theme sticks out in Confluent's S-1 filing: competition from many different parts of the industry.
The company said it operates in four markets: application infrastructure and middleware, database management systems, data integration tools and data quality tools. Together, those total an estimated market size of $149 billion. Confluent says it owns roughly $50 billion of that.
While some areas may be easier to steal share from, others are going to be more difficult, and the key area of competition is likely with the hyperscalers.
Confluent has partnerships with Google Cloud, Microsoft and AWS, but each of those companies also offers its own suite of rival products. AWS, for example, has Kinesis, its own event data service. It also runs Redshift and SageMaker, and integrates with providers like Snowflake. Customers who use AWS may be disinclined to add Confluent as another vendor in the stack.
The Confluent S-1 disclosed the following ownership stakes of common stock, as they stood on March 31, 2021: