Reporting its first quarterly results as a public company, Cloudera reported Q1 revenues of $79.6 million, a jump of 57% year over year. But it was the next line that at first looked to us like a misprint: losses of $222.3 million that were up roughly 4x from a year earlier. But a look at the fine print painted the real picture: the Q1 GAAP losses included $191.1 million for one-time stock compensation charges, presumably to employees.
So it's almost tempting to pop the question: Mrs. Lincoln, aside from what happened, how was the play?
When you strip away the one-time expenses, Cloudera's numbers are going on the right trajectory, with net losses trending down 26% year over year. From the results, its revenue mix of subscriptions has stabilized at the 80% mark, and that net expansion rate from customers was up 142% for the quarter. That's positive evidence that Cloudera's land and expand strategy is working.
Nonetheless, after hours, Cloudera's share prices took nearly a 15% hit. This contrasts with a 9% bump for Hortonworks after it reported a better than expected Q2 last month. It's all about expectations.
It's been fashionable to downgrade Hadoop's prospects, with critiques that adoption has not accelerated as fast, with the reduced valuations of companies like Cloudera and Hortonworks as proof points. Hadoop remains a complicated platform to adopt, but that opens the path for managed cloud-based services that should overcome much of the obstacles. Admittedly, Cloudera is a bit late to that game, having only introduced Altus a couple weeks back.
Cloudera has some cool tools in its portfolio, like Navigator Optimizer that for now is used for tweaking database migrations to Hadoop, but could potentially be repurposed as a cost-based query optimizer for its cloud services. Its recent Data Science Workbench, which claws notebook back under the governance umbrella of the mother ship, provides another tool that could make its cloud services even more consumable to a demanding audience: data scientists.
Next week, we're off to DataWorks Summit in San Jose, where there are more customers speaking, but fewer sponsors showing up. Are there too many Hadoop conferences? Probably. Is it time to change the subject? Yes. Is Hadoop dead? Those rumors are greatly exaggerated.