Software's Top 10 2005 Trends: #2 Open Source

Open Source is one of the most important and perhaps the most controversial trends within the software industry.   While some see Open Source as a kind of communal catalyst for innovation and creativity others see it as a wildly destructive trend that threatens to undermine the economic viability of the entire software industry.

To date, Open Source projects have largely been focused on broad horizontal platforms within infrastructure software.  These projects have been so numerous and so successful that many pundits now talk about building applications on top of the all-Open Source LAMP (Linux, Apache, MySQL, Php/Python/Perl) stack.  And that stack continues to get bigger as Open Source projects start migrating to more complex infrastructure platforms, such as applications servers (J-Boss, Zope, JonAs, Enhydra).  In fact, there are Open Source projects currently underway for just about every major infrastructure software technology you can think of from Content management, to BPEL, to you name it.

One of the biggest questions facing the software industry in 2005 is whether or not Open Source will “jump the species barrier” and start to become a major factor in the enterprise applications space.  A quick look at Sourceforge confirms that to date most Open Source efforts in the applications space have been confined to niche applications in the academic space, but there are now numerous efforts underway, such as SugarCRM, to try and create enterprise-ready Open Source applications.   Whether these applications efforts are successful or not is one of the key issues keeping software company executives up at night.

For software executives, Open Source presents three major choices:  beat it, join-it, or co-opt it.  The conventional wisdom suggests that the way to beat Open Source is to “out-engineer” Open Source by providing a more stable, secure and feature rich product while at the same time “out-servicing” Open Source by providing robust 24/7 support.   This strategy appears to be keeping many Global 2000 customers “in the fold” for now, but even at these customers Open Source software is increasingly showing up in non-mission critical areas.   

Joining Open Source is an option many companies, especially many small infrastructure software companies, increasingly appear to be taking.  These companies typically develop their software in-house and then once finished (or almost finished) declare to their software to be “Open Source”.  The strategy is essentially to use “free Open Source software” as a way to acquire customers that can then be charged services and maintenance fees.   This strategy sounds great but has several drawbacks, chief of which is that simply declaring a software product to be Open Source does not automatically create a large community of diverse programmers who are willing to devote substantial free time to improving the product, but it definitely does make the company’s IP public domain.  Thus, many of these companies face the prospect of getting little or no development leverage from Open Source in return for giving up the copyright to all of their code.

Attempting to co-opt Open Source is probably the most popular and most complicated approach to dealing with the issue.  IBM is the poster child of Open Source co-opting.  IBM primarily sees Open Source as a way to commodify the main revenue sources of its key competitors.  IBM’s bear hug of Linux was largely part of a (arguably wildly successful) strategy to undermine the value of Solaris and Windows NT and thereby stall their push into the glass house.  But IBM’s love of OpenSource only goes so far.  You don’t see them pushing J-Boss at the expense of Websphere or MySQL at the expense of DB2.  Indeed IBM continues to lay out big bucks for “proprietary” software, so their love of Open Source appears to only go as far as they can co-opt it.   Taking IBM’s lead, most of the other major software vendors (save Microsoft) now appear to be selectively co-opting Open Source to use as a weapon against their competitors.

For software VCs, Open Source creates a number of tricky investment problems.   If one continues to invest in proprietary software, you run the risk of getting “Open Sourced” if the space becomes attractive enough to create either a developer community groundswell or interest from an elephant like IBM.  If one invests in Open Source, you run the risk competing with 5 other companies selling the same product and turning software margins into services margins.

However, there is some middle ground.  For example, companies that offer software as a service are somewhat protected from Open Source pressures as customers are buying the whole service not just the code.  Indeed companies that offer software as a service can leverage a lot of Open Source software to lower their own costs.  In addition, while Open Source projects may make some headway into a few large horizontal applications, chances are that it will be very difficult for them to penetrate specialized vertical niches because there just aren’t a lot of developers out there that understand the business logic required for those niches, thus making it much harder for an Open Source project to attain critical mass.

Ultimately the biggest issue for VCs is whether or not Open Source has effectively “capped” the home-run potential of software deals by guaranteeing that any new horizontal software platform that achieves critical mass will inevitably face intense competition from a free, Open Source equivalent.

Personally, I’d like to think that the answer to that issue is “no”, but I also believe that creating a next-generation “home run” platform company is now a whole harder thanks to Open Source.

For a complete list of Software's Top 10 2005 trends click here.