Ticket to Free Ride

From the Archive: New research shows how open-source software and commercial firms can work for one another.
July 30, 2012 | Research Feature
Print this page

Firms often go to great lengths to protect the intellectual property that is the source of their competitive advantage and profit margin, creating massive security systems and requiring staff to sign noncompete agreements that prevent them from working for the competition for a set number of years after they leave the firm.

Yet the rise of open-source software has bucked the conventional wisdom about protecting intellectual property. Open-source products — including but not limited to software — are typically produced on a pro bono basis or by expert volunteers and made available for free to the public.

One popular open-source alternative to the Windows operating system for PCs is Linux. Despite the radical transparency of open-source software, and Linux in particular, the software development firm Red Hat has built a healthy business based on the operating system. Red Hat does so in part by effectively piggybacking on Linux by providing supplemental services that customers still value but that are not provided by the open-source developers, including documentation, technical support, service contracts, and complementary software that helps manage the Linux software.

At first glance, the ability to profit from open-source products seems difficult at best. Open-source software licenses typically require that any time a commercial venture makes changes to the software, those changes also be made available free of charge to the public, along with open access to the code itself. If the firm’s intellectual property thus becomes public, any competitor can easily free ride on its enhancements — so how is a firm to gain any competitive advantage?

“Why is Red Hat willing to put so many resources into improving a product that is made freely available to anyone else?” asks Professor Brett Gordon, whose research interests include competitive pricing and product innovation. Gordon worked with Vineet Kumar of Harvard and Kannan Srinivasan of Carnegie Mellon to understand how firms like Red Hat are able to compete in this unique market, and how a broader market based on open-source products can work. To do so, the researchers created a model reflecting the behavior of firms and their competition with one another and how open-source software is created.

The latter is no simple matter. Software developers’ motivations to contribute to open-source platforms vary greatly. Some do it for altruistic reasons, some simply enjoy contributing or for the challenge, some do it to enhance their reputations — and some do it to signal their skill to potential employers.

Developers are linked to the pieces of open-source software code that they write, giving employers a completely transparent way to verify their contribution to the product. “That’s a huge advantage to a developer because it is otherwise very difficult or impossible to show prospective employers exactly what work you’ve done,” Gordon says. “Software firms simply don’t open their code up to inspection by competitors. But if you work on open-source software you can carry your CV with you wherever you go.”

Firms can then put these skilled developers to work on either the public features of the open-source product — the software’s functionality — or private commercial enhancement such as service contracts, documentation, support services, and easy-install programs — which enhance the software’s usability.

But where should a firm primarily concentrate its resources — on the features or on usability? Both, according to the researchers’ model: when firms make better products in the public dimension — that is, when they contribute to the open-source software — they make their private products more valuable and vice versa. This means that firms should not focus on one or the other but should split their resources to invest substantially in each.

“Firms can free ride on the contribution of other firms, but unless they themselves contribute to public components of the software, they aren’t likely to reap much benefit from it,” Gordon notes. This is borne out by Red Hat’s place in the open-source software market — historically it has provided more contributions to Linux than any of its commercial competitors, placing it in the strongest position to profit from the software.

In addition to helping show why Red Hat and firms like it can succeed in the open-source space, the model predicts how much open-source software is created as a result of commercial firms’ entry into open-source collaborations. One common open-source licensing agreement, the GNU Public License, deems that firms must share their contributions publicly. Another, known as the Berkeley Software Distribution license, allows firms to keep some of its enhancements private. Under the former license, firms tend to contribute to open-source software primarily to make their own products better — this mirrors the experience of Red Hat. But the model predicts a different outcome under the Berkeley license: more open-source software gets created for public consumption when firms are able to privately appropriate anything they do on top of the open-source software itself. This happens when the market is large enough and when firms are able to effectively screen good from bad developers.

Why does more open-source software get created under the Berkeley license? Because firms are able to charge a higher price for the private components of the product, other firms can’t copy the product or build on it, and price competition among firms is reduced. The frontrunner firm earns a higher return on its contributions and can attract better developers for higher salaries. Developers are attracted to the highest-paying jobs so they contribute even more to open-source software to better signal their skill to potential employers.

“You can think of open-source development as not unlike a charitable contribution,” Gordon says. “More seems better. And open-source contributions are higher when licensing arrangements allow firms to keep some of their contributions to open-source software private.”

Brett Gordon is the Class of 1967 Associate Professor of Marketing at Columbia Business School.

Brett Gordon

Brett Gordon was a Columbia Business School faculty member from 2007 to 2014.