I love end of year predictions. Whether they're right or wrong is one thing, but it's valuable to imagine various scenarios of what might happen. Taking the time to consider and connect what has happened with what might happen, opens new insights.
Having said that, I'm not going to make a 2010 prediction about Free and Open Source Software (FOSS). I do, however, want to share some stats about demand trends for enterprise FOSS platforms between 2008 and 2009. I want to see how those jibe with predictions from some other analysts. And maybe, if you squint real hard at the changes between the '08 and '09 stats, you’ll get some ideas about what will happen in 2010.
A little googling turns up evidence and arguments demonstrating that much of the fear, uncertainty, and doubt (FUD) about FOSS adoption have been put to rest. Remaining squabbles about the concept of obtaining and using FOSS in the enterprise appear to stem from lack of information. So what are enterprise software users demanding in terms of FOSS?
Since TEC's main objective is to help companies evaluate and select software, we collect a lot of data about what people are looking for. I look at the data from 20,846 inquiries (2008 and 2009) to see what kinds of changes appeared in the demands of companies evaluating enterprise systems. The graph below shows the percentage of demand for FOSS-related issues in terms of high-level requirements.
(click the image for a larger version)
The vertical bars represent the following six requirements (these are the criteria people told us they had for determining which enterprise software applications they wanted to evaluate for a potential selection):
- FOSS-licensed App (i.e., an application licensed under an official Free Software Foundation or Open Source Initiative-approved license)
- FOSS Database (i.e., an application that may or may not be licensed under a FOSS license but that supports a FOSS-licensed database)
- MySQL (i.e., one of the two FOSS Database options)
- PostgreSQL (the other FOSS Database option)
- O/S GNU/Linux (i.e., an application that may or may not be licensed under a FOSS license but that supports the GNU/Linux operating system as a platform)
- All FOSS Demand (i.e., a combination of all the above options—in other words, out of all requests for enterprise software evaluations, the number that had at least one of the above items selected)
Note that I used data from inquiries on ERP, SCM, EAM, HR, financial systems, PLM, accounting systems, pharmacy information management systems, CRM, ECM, BPM, MRO, PPM, POS, and SFA systems (I've included data from many permutations of these as well--for example industry-specific solutions, and applications that could fall under larger headings like warehouse management systems).
Let me break that down. User demand for all things FOSS increased by roughly 2% from 2008 to 2009. The greatest increase came from people seeking applications that supported FOSS database platforms, followed by those supporting Linux, and finally at an increase 0.5% those seeking enterprise applications that were themselves FOSS.
Over the past year, pundits have wondered whether the recessionary economy would lead to a great increase in the number of enterprises adopting FOSS (the rationale being that these applications would be available at a lower price point). I'm not certain that the increase in demand we've seen this year justifies that hypothesis. It may be the case that the FOSS message has only recently begun to be grasped. meaning that we could see a greater increase in demand next year. However, I think that there is still a fair amount of work to do in educating the marketplace (in spite of my belief that the FUD arguments are no longer very serious).
So how do the demand trends above, fit with the predictions bubbling up from the blogs of analyst firms covering open source in the enterprise? I've read a number of fascinating predictions for 2010 related to FOSS from the blogs at Gartner, 451 CAOS Theory, RedMonk, OStatic, and ZDNet. I'd like to point those out and comment on a few of them, especially in relation to our demand trend stats.
Let's start with the OStatic blog in which Sean Dean thinks that "more big companies will adopt/accelerate open source strategies." He believes there will be increased enterprise adoption of open source next year (and thus, I presume, increased demand). What I find interesting is the nature of his justification.
Dean points to Android's success as a key factor leading toward accelerated adoption. OK, sounds reasonable. As a mobile platform, Android (Google's Linux based mobile OS) has been gaining a lot of market share, very quickly. If for no other reason than a lot of people buy Android devices, a lot of the mobile apps used in business will be running on a FOSS platform and could very well be FOSS themselves.
Gartner's view interested me because Mark Driver's blog post largely revolved around approaches to providing open source software. He considers how adoption could drive three categories of providers. He suggests that adoption of "Affero-style" reciprocal licensing will increase (these licenses close potential gaps in the redistribution of FOSS source code). I wouldn't know his reasoning behind this but I could hazard a guess that it's due to the increase in SaaS availability, which was shown to sidestep some of the traditional source code distribution requirements.
Finally, most interesting to me is his comment that
"By 2012, at least 70% of the revenue from commercial OSS will come from vendor-centric projects with dual-license business models... This is may [sic] true today but the lack of revenue among broader market OSS products compared to Linux isn’t large enough yet to make this one a done deal..."
I think this may reveal something about the mindset of those adopting FOSS applications. Though again, I don't know the details behind Driver's prediction, I assume he's referring to the model in which users can download an application (or its code) at no charge and implement it in their companies. The application may not have as many features or come with the support that an alternately licensed version would. Thus, the company would eventually (if not outright) pay for the alternately licensed version.
That model, as Driver points out, already exists and if it's truly an increasingly successful model then it suggests to me the companies adopting FOSS applications tend to accept it. Driver's post suggests that it is the "more-conservative open-source adopters" likely to want the support that this model more easily affords a vendor to provide. It's not without its downsides, which some, like ZDNet's Dana Blankenhorn would characterize as a "phony open source" trend for 2010:
"When there’s an open source "community" version and a paid "enterprise" version of the same software, what is the difference between writing a check for enterprise support and just buying a closed source license? As open source increasingly becomes an enterprise mandate, you can expect such questions to gain new relevancy. How open do you have to be? How closed must you be?"
I talked about some of these business models a few years ago with various open source ERP companies. And indeed, there ought to be consideration given to the value in a community version versus other versions.
Aside from the Gartner blog post, there was also a recent Gartner press release predicting that open source BI tool adoption will grow five-fold through 2012. They also mention that it was once considered for "cash strapped" organizations but is now mainstream. (TEC's Anna Mallikarjunan wrote about this previously in her article, Open Source Business Intelligence: The Quiet Evolution.)
According to TEC's research, 2.5% of the people seeking BI solutions in 2008 wanted something available under an official FOSS license. In 2009 that number rose to 4.2%, an increase of 1.7%. Not only is the demand for FOSS BI systems slightly higher than demand for FOSS enterprise apps in general, but it increased at a much faster rate.
If Gartner's BI prediction holds true, I imagine we'll see roughly 20% of the total demand for BI systems hitting FOSS vendors by 2012. Now, I'm talking about demand but Gartner talked about adoption. The number of companies that actually follow through and adopt a FOSS BI system is likely lower than the number that seek it for evaluation in a selection project. In other words, demand would likely rise to higher than 20%.
Matthew Aslett's 451 CAOS Theory blog post about the war being over corroborates Gartner's predictions in the areas of licensing models. Aslett also has a point about the overall market, he cites IDC's prediction that the open source software market would hit $8.1B (or grow at 22.4%) by 2013. Aslett makes an educated guess that this is still only 2% to 4% of the overall software market. Aslett's point is that in spite of the growth, there is still a long way to go for penetration. Our research shows demand for FOSS enterprise apps to be 4% of the total, which further supports Aslett's assertion.
I enjoyed seeing that Stephen O'Grady's tecosystems blog on RedMonk counters a few of the predictions above. His five FOSS predictions include the idea that dual licensing will decline (countering the Gartner and 451 Group predictions). O'Grady predicts that permissive licensing will increase rather than reciprocal styles such as those predicted by Dean at Gartner. I'd think that if permissive licensing models increase, some vendors will have less motivation to use a dual license model. I can't say that I've ever fully understood why a vendor would want to use a permissive license as opposed to one that requires more reciprocity.
I said previously that if dual licensing is truly on the rise, it may mean those demanding FOSS are more willing to accept it. If instead it's declining but permissive licensing models are increasing, then that also signifies acceptance for potentially more restrictive terms. In either case there are two predictions inching toward a business model that involves FOSS distribution but with increased restrictions in one form or another. I wonder which (if any) will be more palatable to those who are demanding free or open source software.