Recently I was invited to the CASCADOSS symposium to give my views on open source business models, specifically in the GIS market. For the presentation I developed some themes that might be of interest to others working on open source and/or GIS. I therefore decided to (finally) start putting this blog to some good use by publishing a series of posts that elaborate these themes. This is the first installment.
The topic of this post is the following thesis: Open source business models don't matter. Obviously they do matter to companies pursuing such models, but the sustainability and vitality of open source doesn't depend on the sustainability and success of open source business models. (And by extension, it doesn't depend on the existence of successful open source companies).
The thesis evokes Carr's (in)famous paper "IT Doesn't Matter". Carr's central argument is that IT is an infrastructural technology and infrastructural technologies tend to become commodities: they become standardized, interchangeable, and ubiquitous. On the basis of a commodity a company cannot build a strategic competitive advantage. How could it when all its competitors can acquire (substantially) the same commodity? When an IT function is commoditized it ceases to be useful as a means to competitive advantage, and instead becomes a necessity of doing business. Just like electricity and the telephone, it becomes a cost factor. That is bad news for software vendors that like to charge hefty licensing fees: as soon as companies realize that IT is mainly a cost factor they will exert pressure to lower that cost.
I think Carr's analysis is essentially correct and that it explains the current success of the Professional Open Source business model - the model adopted by companies (or company divisions) like Alfresco, MuleSource, MySQL and JBoss. These companies rapidly gain market share because they offer substantial cost savings relative to the competing proprietary offerings. By supplementing open source software with the support, dependability and accountability that can only come from contractual relations with a solid commercial company, they succeed in creating offerings that are (at least) adequate for many customers but at a much lower price point.
Now suppose that the Professional Open Source companies succeed in forcing the proprietary software vendors out of the market. If Carr is correct there is no reason for thinking that the drive to commoditization would then stop, and with it the downward pressure on prices. Since open source has dramatically lowered the barriers of entry into the software market, open source companies would need to keep prices relatively low. This suggests that in the long run these companies will not be so highly profitable as an Oracle or Microsoft currently is.
Would any of that matter to open source? Would open source loose its attractiveness and the community its vitality if it should turn out that Professional Open Source, or any other open source business model, is only moderately successful? I don't think so. The fundamental economic value of open source lies in its potential to lower the IT costs of its users. This potential is, and will remain, a strong motivator for the commercial development of open source software. Open source companies are but one organizational form for this development. Many other forms are possible such as ecosystems of cooperating IT-service companies, foundations such as the Apache and Eclipse foundations, and enterprise collaboration through open source communities (see this story).
So even if open source business models turn out to be financially less successful than many think (hope), I believe open source will continue to thrive. And because of that it will continue to provide business opportunities.