Beware impostors in open-source paradigm shift, says Hortonworks COO


Startups might yet make good on so-called broken promises of open-source big data software if they spike the punch with a proprietary ingredient. But then why call themselves “open” at all, asked Raj Verma, president and chief operating officer of Hortonworks Inc.

“Why are you doing open-source? What is your intent of doing open-source? To be seen as open or to be truly open?” Verma asked at the DataWorks Summit in Munich, Germany. Companies seeking to make money from open-source software by more traditional means do help signal its potential to investors, he said.

Verma spoke at length about Cloudera Inc. (valued at $260 million) and its IPO with John Furrier (@furrier) and Dave Vellante (@dvellante), co-hosts of theCUBE, SiliconANGLE Media’s mobile live streaming studio. (* Disclosure below.)

“It’s testimony to the fact that companies can be built,” he said. Hortonworks is increasingly being compared to Cloudera, but the former is pure open-source, whereas Cloudera has sprinkled some proprietary software into its mix. This may give Cloudera an advantage in the short-term, but customers will ultimately value Hortonworks’s easy-exit policy more, Verma stated.

“If you add a slim layer of proprietariness, why are you doing that?” Verma asked. His 23 years of experience working in proprietary software provides a clue. Talks with customers who had locked themselves in and suddenly wanted out often led to a phenomenon known as “switching cost.”

“I would say, ‘Look, you really don’t have a choice. Because if you want to exit my technology, it’s going to probably cost you 10 times more than what you’ve spent till date,'” Verma said. In other words, Cloudera and others add a proprietary layer so that the cost of leaving makes the cost of staying seem less painful, he said.

While most analysts would argue that this is just horse sense for businesses that want to retain customers, Hortonworks’ mission is to prove them wrong. By not locking customers in, Hortonworks keeps the open-source purity pledge, and customers appreciate this, he said.

Open source’s dirty secret?

But then how does one define customer-lock-in in the real world? One simple answer is: “Proprietary code equals customer lock-in. Period,” said Verma. There is more than one way to skin a cat; what if “truly open” companies are locking customers in through more insidious means than proprietary code that they can see upfront?

Hortonworks enjoys partnerships with cloud providers like Inc. and Microsoft Corp., companies that may lock in their customers to varying degrees.

“You could argue that Hortonworks, through their relationship with partnering, has a lock-in spec,” Furrier said. The point was not refuted by Verma, who ultimately agreed that if the value of proprietary software exceeds the lock-in risk, then it’s worth it.

At any rate, being free of overt lock-in is not the only advantage that Verma said Hortonworks has over Cloudera. For one, he said it’s more efficient than Cloudera, probably owing to the fact that it has been public for two and a half years now. “We’ve had a thousand-watt bulb being shone at us, and we’ve been forced to be more efficient because we were in the limelight,” said Verma.

Now that Cloudera is going public, the market may well force it to tighten its belt, but it’s quite unpracticed at this, Verma argued. “Some would say that they were actually spending money like drunken sailors if you really read their S-1 filing,” he said.

And why not? Cloudera was generously funded and had other plush cushioning underneath it, Verma said. But here’s the kicker: “You’re talking about a company that had a three-year head-start upon us, had a billion dollars of funding at very high valuations, and yet they’re only one year ahead in terms of revenue, and they have burned probably three times more cash then we have,” he said.

To put it in numbers, Cloudera’s margins are 82 percent in proprietary software; Hortonworks’ margins in open source are 84 percent, said Verma.

However, Cloudera’s new machine learning ambitions will buoy it up to Amazon, IBM Corp. and Google Inc.’s rank, right? “With a billion dollars of funding, with the Intel ecosystem behind them, they couldn’t effectively compete against Hortonworks. What do you think are their chances of competing against Google, Amazon and IBM?” Verma said.

Staying the course with pure-play data

Focusing deep on data monetization rather than chasing a variety of trends has placed Hortonworks at the forefront of big data, Verma stated.

In the past couple of years, chief information officers have become less interested in structured data, which cannot really drive innovation, he explained. “Innovation and insights into unstructured data was the seminal challenge of most of the executives that I was talking to,” he added.

And there is no better technology right now for making sense of unstructured data than Hortonworks, Verma claimed.

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of DataWorks Summit 2017 EU(* Disclosure: TheCUBE is a paid media partner at DataWorks. The conference sponsor, Hortonworks, does not have editorial oversight of content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE