Open Source Goes Vertical

LinuxWorld added something new this year: vertical tracks. These day-long events were dedicated to exploring how open source can contribute innovation and cost-efficiency to specific industries.

I had the opportunity to moderate a panel in the Healthcare track. My panel focused on “Opening up Healthcare Markets with Open Source” and it provided some interesting discussion. Members of the panel ran the gamut: a healthcare expert from IBM, the CTO of Agfa Healthcare, a software engineering manager from HP, an open source software provider to small clinics and medical practices, and the sales VP from Medsphere. The latter company is a fascinating open source story: it provides Electronic Health Record (EHR) software based upon VISTA, an EHR system developed by the US Government’s Veterans Administration. Medsphere received the software under the Freedom of Information Act, since the US Government does not copyright anything it publishes, including software!

I asked the panel to address two questions: (1) What does open source mean for the established proprietary vendors of healthcare software; and (2) how does open source enable new vendors to compete for healthcare IT?

On the first question there was general agreement. Open source will challenge the pricing of proprietary entrants and force lower prices and margins. As far as this goes, not so controversial. However, given that both of the keynote speakers for the Healthcare track pointed out that the industry trails mainstream IT practices by five to ten years, the expertise in Healthcare IT organizations is very thin and their ability to hew a different path than that provided by vendors seems like a real challenge.

Still, there is intense price pressure in the entire healthcare system today and it looks likely to become even stronger in the future (statistics cited by Dr. Ken Kizer, CEO of Medsphere and keynote speaker for the Healthcare track: the US currently spends $1.7 trillion dollars on healthcare and is on course to spend $3.4 trillion by 2013 — clearly something’s got to give). That price pressure will undoubtedly come through to software providers and open source will provide a stick with which to beat up vendors.

So, the presence of open source software will cause price pressure on healthcare IT providers. So far, so good.

It was the next question that proved even more interesting. All participants shared a perspective that open source will be the route that new participants enter the market. James Kaufman and Phil Robb of HP emphasized that the growth of standards in the healthcare industry will change the ground rules of competition — in the past, proprietary vendors competed to impose their own file formats and communication protocols on the market, but standards allow anyone to enter the game.

However, vertical open source differs from traditional open source in one significant way — the requirement that significant domain expertise be part of the product. In the past, open source was created by engineers for engineers; in fact, the cliche is that open source developers scratch their own itch, meaning that they create products that solve problems they themselves face in their day-to-day activities.

I posed a challenge to the panel: how will domain experts be included in development of vertical open source products? I noted that while it’s easy to imagine an engineer going home after a day of work and programming during the evening on a pet project, it’s very difficult to imagine a, say, an oil-and-gas guy going home and putting in some additional hours on creating a new piece of vertical open source. Incorporating vertical expertise into a domain-specific open source product seems to be a difficult nut to crack.

Several of the panelists felt that venture capital would provide funding to support development of vertical software and thereby address the problem of incorporating domain expertise. Instead of an oil-and-gas guy toiling on open source after a full day of work, venture money would make his day job be contributing his expertise to vertical open source.

There’s one shortcoming to that scenario. To this point, I haven’t seen any enthusiasm from the venture community to invest in vertical open source. They are too busy steadily marching through the IT software industry, investing in one or more open source startups in each segment.

On the other hand, it makes a ton of sense that venture money would eventually flow to vertical open source. It’s a truism that the entity closest to the customer reaps the lion’s share of total value chain margin (although the PC industry doesn’t seem to work that way), so top-of-stack open source software would seem to be a very attractive investment proposition.

My opinion is that we’ll see a mix of startup strategies in the healthcare software industry. Venture capital will probably power some startups. However, it’s a misperception that venture funding is a prerequisite for getting a startup off the ground. Vertical open source will also be funded by a model typical of non-high tech industries — retained earnings garnered during the slow but steady growth of a bootstrapped company. In other words, many vertical open source software startups will start to resemble construction startups, media startups, etc., etc.

As software becomes more mainstream in its revenue and margin expectations, software companies will become more mainstream in their growth strategies. Vertical open source has a bright future, but it probably won’t resemble its proprietary forebears.