ODCamp 7: Ten years of open data. What have we learned?

Tim Davies has written a book: The State of Open Data (@stateofopendata) with the support of international development and open data organisations and the OD4D.net initiative.

So he wanted to run a retrospective of the first ten years of open data at Open Data Camp 7 in London: and to find out from campers what their experience had been: “the good, the bad, the in between.”

First, the book. “We recognised sometime last year that we were coming up on a decade since President Obama made a splash on open data in the US and the UK launched the Open Data Institute. So we put up some Google documents and looked at seven areas to brainstorm thoughts about what had happened in them.”

More than 200 people pitched in ideas, and commented on events and initiatives from different perspectives. Then, the authors – 40% of whom were women, and many of whom come from the global south – were asked to comment and revise what became 6,000 word chapters.

“One thing we found is that the open data community is a very self-critical community. It wants to do stuff and then when it does it it wants to change it. So sometimes it can forget the good things that have happened,” Ian said: adding that a lack of effective archiving did not help with this.

“The other message is that we are well past peak-hype; but different areas are at very different stages. Transport, for example, was an early area of interest, but a lot of data has gone back to being proprietary. Telecoms, there is very little going on. And we have missed some tricks. Data literacy is a real issue.” But a lot of interesting things are still happening.


Retrospective (because it’s not Open Data Camp without sticky notes)

Open data campers who attended the session at Geovation were then asked to write down their own perspectives and experiences on sticky-notes, to form their own retrospective.


What came out was: a cluster of responses about “people”: a first time Open Data camper said that “if this camp is indicative of the open data community, then it is a really good place and a really good thing” because it is far more friendly and inclusive than other tech and data conferences.

Tim said this was interesting, because in the US the open data community is fragmenting, but “in the UK it seems to be holding together.” Another participant said Open Data Camp was also far more inclusive than it had been when it started. People matter, Tim added, because “movements like these depend on people.” In the UK, some aspects of open data “have become more institutionalised” but others “very much depend on people going out and doing this themselves.”

Data management

Another theme was data management and the need for ongoing investment. Participants agreed that there was less government focus on open data than there had been in the early days. So there is an issue about how interest and investment is going to be maintained. Another issue is governance. Even if people have good and well-maintained data, initiatives like the General Data Protection Regulation have made some organisations more worried about releasing it as open data.

Use and value

If data is released as open data, it also needs to be used. One participant suggested that there seem to be fewer examples of open data being used around these days. Tim said he thought there was a lot of private sector use of open data “but that is not always fed back.” Another camper suggested that there needed to be a renewed focus on use cases and open data being used for good.

“How many, not necessarily business, end-use cases can we point to and champion, so that we can build towards there being more business and social enterprise involvement in events like this, and more demand for data to be released as open data,” he asked. An example of something that had worked, it was suggested, was prescription data, which has been used to analyse prescription information and drive clinical change.



The bad, the good

Sticky notes that focused on bad and good initiatives put the BBC’s initial interest and subsequent lack of interest in open data into the disappointing category, and Open Street Map into the success category. One participant suggested that Open Street Map had made it “much easier for people to add data” and organisations were interested in using it – libraries have trained users to add local features of interest, and the National Trust may use it to map paths.

Others felt even more could be done: local government could be encouraged to use Open Street Map. Open data experts might go to public sector conferences “and park tanks on their lawn and say look, you can spend a fortune on a commercial licence or you can use this.” Tim agreed that there was definitely still a sales and marketing job to be done: and a business opportunity for small, local companies that could provide the expertise and support that might be required.

Ecosystems and silos

This led back to the use and value point. One of the areas in which open data might have struggled is in creating an ecosystem that supports not just the publication of open data, and some use-cases, but paid work opportunities. Also, another participant suggested, “the dream of open data was that it would break down silos” and this has not happened. In the UK at least, open data may be referenced as a source of information about a sector, but it is rarely used across sectors.

Camping is great, but where is the follow-up?

This led to another discussion about community. One camper pointed out that while Open Data Camp works well, there is no Slack channel or WhatsApp group that does the same job: and no way of following up on any decisions taken at Open Data Camp and “holding people to account.” (Actually, there is a Slack channel on the website, but nobody in the session had heard of it or used it).

One participant suggested this may be cultural: in the US, it’s common to exchange telephone numbers and use social media tools: in Europe, not so much. Also, in Europe there is some suspicion of “institutionalisation”; so there is a challenge to create a community with some formal communications channels without getting too formal. Another participant suggested using Twitter: but someone else pointed out that Twitter and other social media tools have become much “less cosy” spaces than they used to be, and might not be as useful as intended.


Asked for their key points, campers said: open data had made some activity easier for business, but there were still issues around literacy and using tools like SPARQL. Leadership and finding people interested in open data is a challenge, although Open Data Camp is helping to keep the community together. There is a knowledge and management information problem, with some data sets missing, like Post Codes. And value is key. “We need to capture the value and put a number on it.”

[Session Notes]

2 thoughts on “ODCamp 7: Ten years of open data. What have we learned?

  1. Thanks to Tim and to Mor Rubinstein for running this session. I really enjoyed it. Also, thanks to Jez Nicholson for suggesting the retrospective format which worked really well.

    I suspect we will be equally self-critical in ten years time and that this is (usually) A Good Thing.

Comments are closed.