thoughts on success factors for open data initiatives

Martin Tisné (@martintisne) asked at Open Knowledge Festival (#okfest) about success factors for open data initiatives (I'm paraphrasing the question).

My thoughts:

To some extent we are trying to describe a successful end or target state for an open data initiative, but all of the initiatives we have are in early stages - we have no mature model to look to. (This may indicate we need to identify similar initiatives that are mature as indicators of possible success factors.)

Scoping it to just open data is useful because we can talk about specific motivations and activities rather than the broader issue of open government. Many of the open data initiatives were, at least in part (and sometimes covertly) ways to drive a wedge of innovation into government (as well as a wedge of transparency). The motivators were to open government up to civil society, to open government up to modern technology approaches, and to unleash the creativity of civil servants who often don't have channels for their innovative ideas.

Initially we thought just creating the portals and getting the data would be the hard part. So a lot of effort was put into just getting data up in any format, using manual processes, and leveraging whatever data was released using apps contests and hackathons. This almost turned out to be easier than expected - maybe we succeeded too quickly in a way.

Now there is a struggle to define what a sustainable next generation open data initiative looks like, but confusingly this is happening at the same time as some countries are just embarking on their first open data steps.

The lessons I draw from the first generation of projects is:
* there is long-term value in pushing for specific types of data (geographical; budgets; health indicators)
* technical standards (eg machine readable formats) are important but shouldn't be used as an excuse for delaying data release
* community is more important than technology - there must first be outreach to developers, but ultimately there needs to be some direct connection to a community of interested citizens

There may be no one path to success; ultimately the sustainable and impactful sites will be the result of an ongoing conversation between the data owners and the community of interested users.