Back in the 1990s, the technology world seemed alive with companies that had chosen to monetise an open-source project and get behind it. However, for Intel Capital vice president and managing director of services, open source, and machine to machine Lisa Lambert, the startup landscape today is quite barren, despite the number of open-source projects being higher than ever.
Lambert told ZDNet at the Intel Capital Global Summit that any startup looking to succeed with open source needs to break new ground and be the first to claim it.
"You've got to be on the leading edge, and out there, and right," she said. "I think [the successful startup] is going to be the first that disrupts. You're not going to see a whole bunch of companies win, and honestly, in the LAMP stack era in the '90s, that was probably the case too."
The change needed has to be revolutionary, not an incremental evolutionary change, according to Lambert, who said that open-source companies often struggle to differentiate themselves, especially given that revenue is generally generated from open-source projects in three ways: Providing support and services, selling a proprietary version of an open product, and through a subscription model.
"Those are the three ways you monetise; there's not a whole lot of differentiation," she said. "And so if it is an open Apache or GPL or BSD software licence, anybody can access the code. How do you differentiate? Those are the standard business models, and why can't anybody else do that?
"It is more difficult to differentiate in an open model than it was before. It was very new back then, and so people didn't even know how to implement the business model, to say nothing about the architectural model complexities."
Lambert also sees large businesses using open source to solve their own technology issues as an explanation for the high number of projects without companies looking to monetise directly from them. She cited the example of the BigTable and Hadoop ecosystem being spawned out of Google, Yahoo, and Facebook.
"BigTable, which was the forerunner to Hadoop, was born largely to address the problem that they were processing the internet every other day. We're talking lots and lots of data, lots of variety types, mostly unstructured, and they needed a solution to do that; a relational database and data warehouse just wasn't working. Structural data set methodology doesn't work, SQL queries don't work in that kind of environment," she said.
"For them, what makes the most sense is, 'We're solving a business problem for us. Why not benefit from an open development model and get input from a broader, global community?'
"They had no interest in monetising, they had an interest in solving their business problem so that they can monetise in the way that they monetise. I think that is a good reason for why there are fewer companies behind [open source]."
Compared to the dotcom era, where Lambert said business founders saw an opportunity to monetise a project and did so, today, open source cuts across many verticals and isn't a monetisation opportunity in and of itself.
"The number of projects are exponentially larger than they were back then, but I think people are viewing open source as a tool set for delivering a function, in a crowdsourced method, which is very valuable in today's world — you need multiple points of view, it's a great development methodology," she said.
"It is more a development methodology than it is a business method."
Disclosure: Chris Duckett travelled to Intel Capital Global Summit as a guest of Intel.