Last week I was invited to talk at the South Tyrol Free Software Conference which took place in northern Italy, in the city of Bolzano (disclosure: a paid gig.) As its title indicates, this was a more local, specialised conference than some of its more famous international siblings, but I was impressed just how much activity was going on. It was also interesting to see that open data was already a hot topic here – it's not just national holdings that are being opened up.
My talk was called "Open Source, Open Data and Open Innovation" (the programme went with a more general placeholder), and drew on several themes that readers of this blog will be familiar with. For example, I pointed out how we are experiencing a unique transition from a world dominated by scarce analogue artefacts to one in which abundant digital artefacts are not only present but are crucially important for many of the most rapidly-advancing areas of innovation.
My main focus was on two strands: how Linus Torvalds invented open innovation – something that I've written about before – and the business of open data. What's interesting is that open data has now moved on from the "how can we possibly open up all this data that we kept secret before?" to "what's the best way of giving people access to all this data?"
In other words, the open data movement has won the basic argument that it makes a lot of sense to release certain kinds of (non-personal) data as freely as possible, especially for governments – not least because the public paid for those holdings to be created in the first place and therefore has a good claim on being able to see and use them.
In my talk, I started to look beyond those achievements. Just as the success of free software led to the founding on companies based around that freely-available code, so I think we are about to see a wave of exciting new startups based around freely-available data, some of which could well grow to significant size (Red Hat is just breaking through the psychologically-important figure of $1 billion annual turnover.)
Such companies are few and far between at the moment. A good example is OpenCorporates.com, which I wrote about before. As I noted there, I've no idea how it will make money, but this seems such an incredibly rich set of data about some rather important aspects of modern life, that it's hard to believe that there isn't going to be something that can be done in this regard.
That's one kind of open data company; as I mentioned in my presentation last week, it's an example of the direct model of making money from open data. Alongside this, I predict that there will also be indirect models. The distinction can be seen in the world of open source, where there are companies like Red Hat that make money from open source directly – selling various kinds of services – and others like Facebook that simply use open source for its operations, thus saving serious amounts of money without sacrificing power, scalability and flexibility. It's striking that practically all the most exciting players like Facebook, Twitter and Google employ open source in this way.
Another example of this indirect model is Google again with Android, which it gives away in order to create a vast ecosystem so that it can then make money in other ways – from advertising for example. Again, I don't really know exactly how this will translate into the world of open data but I'm pretty sure it's going to happen.
Finally, as I explained in my Bolzano speech, open data is not only of interest to companies that want to make money from it. Just as open source has become an accepted tool for most businesses today, I predict that more and more of them will routinely draw on the growing bodies of open data for their businesses – after all, given that it's freely available, they'd be mad not to.
Putting together open source tools with sets of open data also allow companies to move towards open innovation. This means that those outside the business – partners, customers and even competitors – become part of the process of developing new products and processes. Open source code means that software enabling open collaboration can be distributed freely, while open data means that there are no issues about trade secrets or corporate secrecy.
To be sure, this kind of open innovation will take time to be adopted and used effectively, but the example of open source shows that the culture of business can and does adapt when presented with opportunities that make sense. As companies move towards this thoroughgoing use of open resources, they will turn into open businesses, where openness permeates their entire operation.
A slightly fuller version of the above argument is available in the slides of my presentation, which I have embedded below. It is also freely downloadable under a cc-by licence.