All news

Cloudera's IPO shows public buyers pay for growth, not private marks

A neat stack of storage drives with clear covers on a stone counter, one drive in cobalt blue

Cloudera listed at less than half its 2014 private share price, then rose 20%. Public buyers paid for subscription growth and ignored the label.

Fifteen dollars a share, less than half the $30.92 paid in its last large private round in 2014, is the price at which Cloudera went public. The data management and machine learning platform priced 15 million shares on 27 April, raising $225 million, and closed its first day on the New York Stock Exchange at $18.09, up 20%.

The listing is a verdict on the private mark, not on the business. Public investors bought above the $12 to $14 range and lifted the shares on day one, so demand for a platform growing more than 50% a year was real. What they would not pay was the $4.1 billion valuation set in 2014; at the offering price Cloudera was worth about $1.9 billion. The gap belongs to the late private round, and any AI company that raises at a price set by scarcity rather than by revenue carries the same gap until it lists or sells.

The prospectus shows what public buyers paid for. Revenue for the year to 31 January rose 57% to $261 million, three quarters of it from subscriptions, and customers from a year earlier now spend 42% more on subscriptions than they did then. About 500 of the world's 8,000 largest companies use the platform and bring 73% of revenue. Against that, Cloudera lost $187 million and used $117 million of cash in operations. At about seven times trailing revenue, the market priced a fast-growing enterprise software company with heavy losses. It gave no extra credit for the filing's description of a platform for data management, machine learning and analytics. A strategic buyer prices differently: Intel offered about 41 times revenue for Mobileye in March, for chips already in cars.

For founders the cost of a high late round sits in the capitalization table. A listing below the last private price hurts the investors who bought at the top and the employees who joined afterwards with options priced off that mark, and it turns a successful company into a disappointment for both. A founder who takes the highest price on offer sets the bar that a public buyer will later measure against revenue, retention and the cost of growth. Fanaura judges an entry price by that later test, investing over a horizon that runs to the listing and beyond, because a valuation the public market will not repeat is a loss that has only been postponed.

Critics would say Cloudera is a big-data company from an earlier wave wearing a machine learning label, so its price says little about AI companies. Most of its revenue does come from storing and processing data, not from models. That supports the point rather than weakening it: public buyers paid for subscriptions and expansion and ignored the label. The first quarterly reports will settle the rest. If the shares hold while losses narrow, the reset was about the private price. If they fall, the market doubts the growth itself, and the bar for machine learning revenue rises for everyone who lists after Cloudera.

A private round can set any price. A listing sets the one that sticks.

For founders

Building machine learning infrastructure with subscription revenue that expands? Send us the project.

Fanaura invests its own capital and the capital of its limited partners (LPs), usually through a separate SPV for each deal. If we see a fit, we get in touch.

Apply in five minutes
Subscribe

Get the news by email

One short note each weekday. No marketing, and you can unsubscribe at any time.