Is the Hortonworks IPO a referendum on open source?

This past Friday, Hortonworks made good on its announced intention to present an initial public offering to the market. The shares were initially priced at $16, opened with an initial bid of $24, and closed their first day of trading at $26.48. The shares were down to $24.97 as of Tuesday’s market close. We’ll see how the company does post-holidays.

In the run-up to the IPO, some criticism of it emerged, based (understandably) on the company’s startup status and current lack of profitability. Revenues may be substantial, but if costs exceed Hortonworks, you can’t blame people for being skeptical around its shares.

Suffice it to say, there are plenty of companies in the data sphere whose business model seems to be all about the exit. These companies are less built to be profitable than to be bought by profitable companies that don’t feel like building competing products themselves. That premise feels non-compliant with the laws of business gravity I learned over my own career, but it does at least follow a certain logic. And Hortonworks’ pre-profit IPO doesn’t seem any more deviant.

OSS + IPO = ?

The area where I do have questions revolves around Hortonworks’ operational model of bringing 100 percent open-source software to market and monetizing exclusively on training, services and support to its customers (including other vendors).

If we think about recent splashy data IPOs, we quickly arrive at those of Splunk and Tableau. Like Hortonworks, both companies are in the data and analytics space, and Hadoop is a very influential part of their businesses. Tableau is a profitable company whose prospects seem ever-expanding, and Splunk reached profitability several years ago. Both companies deal in commercial proprietary software as their business model (even if Splunk is involved in a number of open-source projects). This makes for a straightforward pitch with shareholders: We make software, we sell licenses for that software, and then we update it and add new products to the portfolio, for which we also sell licenses.

Sometimes the pitch is more nuanced. I remember when two other open source companies, Pentaho and Jaspersoft, first came on the scene. The model both companies adopted was to develop core software and make it available under open-source arrangements but also build and offer higher-end extras for more conventional licensing and sale.

Even with that hybrid approach, back then, in the largely Enterprise world of business intelligence, such a model seemed suspect. Eventually Pentaho and Jaspersoft made their way, though, and as incumbent vendors’ license pricing rose, the efficacy of the commercial open-source model seemed even more apparent. Regardless, Pentaho does not today particularly identify as an open-source company. Jaspersoft, meanwhile, has been acquired by the very commercial software-oriented Tibco.

Hadoop chutzpah

The bottom line of all this is that recent analytics IPOs have orbited around commercial software and even open-source analytics companies who have not gone public have become more commercial. So with those precedents in place, where does Hortonworks, a pre-profit startup dedicated to 100 percent open source, get its nerve? Before offering a hasty response to that question, consider some counterpoints.

To begin with, Hortonworks does make software and it does, even if indirectly, make money from that software. No, Hortonworks doesn’t sell licenses. Instead, it takes a leadership role in developing key components in Hadoop, including Hadoop 2.0’s YARN cluster management layer, the Tez framework that offers interactive services on top of it, and the Stinger initiative that has converted Hive from a SQL abstraction layer over MapReduce to one that utilizes the aforementioned Tez and YARN. And, lest you forget, Hortonworks is a spinoff of the team at Yahoo that drove Hadoop’s creation in the first place.

Elephants in a row

Since YARN and Hive ship with virtually all Hadoop distributions, and as support for Tez grows, Hortonworks’ position as “the” support organization around Apache Hadoop looks increasingly credible.

If Hortonworks made proprietary extensions to Hadoop, those components would not ship with competitors’ distributions. And if Hortonworks were less committed to open source, it might have less influence in seeing some of its projects (like Tez) on-boarded to the Apache Incubator and then reach Top Level Project status. Suddenly, Hortonworks’ open-source strategy seems less naive and less altruistic — in fact, it may be pretty darn shrewd.

Hadoop is becoming a universal data layer, increasingly embedded in other software. Open source may not be the fastest road to monetizing software, but it is a super highway for establishing standards that gain rapid industry-wide support. And since Hortonworks wants to be, quite literally, the standard bearer, its 100 percent open-source mantra actually makes a lot of sense.

Let the market decide

That doesn’t mean it will work. There’s an increasingly pervasive attitude in the analyst community that the open-source model hasn’t been financially rewarding for most companies that have bet on it. Red Hat is offered as the exception, though that offer is usually followed with disparaging remarks uttered under the critics’ breath.

But Hortonworks is bullish on Hadoop, on open source, and on the business model of supporting the big data technology that it helped build and helped establish as an industry standard. Perhaps ticker symbol HDP will serve as a tracking stock for that go to market approach.