Open Data Knots

Jeni Tennison

Drawing on poems by R. D. Laing, ODI Technical Director Jeni Tennison explores ‘knots’ that open data owners and users can get trapped in when they don’t take time to develop a common language, build trust and identify shared goals


To break free of open data knots, owners and users need to see their relationship as symbiotic rather than parasitic. CC BY 2.0, uploaded by gabe popa.

The patterns that the owners and prospective users of data fall into with each other remind me of the poems by R. D. Laing in his book Knots.

These poems explore self-perpetuating patterns of thought, action and conversation, usually ending in immobility and despair. Read some extracts to get a sense of the style.

The basic open data knot goes something like:

OWNER: I do not know what data you want
       So I do not know what to give you.

 USER: I do not know what data you have
       So I do not know what to ask you for.

Digging a little into the thinking behind this knot:

OWNER: If I give you data
       And you do not build anything with it
       It will be a waste of time and money.

       I do not want to waste time and money
       So I need to know what you will build
       With the data I give you.

 USER: If I come up with an idea
       And you do not have that data
       It will be a waste of time and money.
       I do not want to waste time and money
       So I need to know what data you have
       That I can build with.

At this point, prospective data users spot a way out of the knot, namely getting a list of datasets. My observation is that this seldom lives up to its promise because the interaction goes something like this:

 USER: If you make a list of the data you have
       I will know what data you have
       And will know what I can build
       And will know what to ask you for.

OWNER: Here is a list of the data I have
       So now you know what data I have
       So you know what you can build
       So you know what to ask me for.

 USER: I am mad because
       The list includes so much data I do not want
       I cannot work out what I can build
       And do not know what to ask you for.

OWNER: I am mad because
       I gave you a list
       And you have not told me what you can build
       And what data you want.

The other solution I’ve seen users suggest is for publishers to just publish everything. Again, this tends to fall into a bad state:

 USER: If you give me all the data you have
       Then I will build something.

OWNER: I will give you the data I can give you.
       If you build something with this data
       Then I will give you more data.

 USER: I am mad because
       You have given me a lot of data I do not want.
       I cannot build anything
       With the data you have given me.

OWNER: I am mad because
       I have given you data
       And you have not built anything.
       I do not believe you need any of my data.

Both data owners and prospective data users get trapped in these knots. Both are frustrated by the other’s actions or lack of action. They talk past each other and feel powerless to change the other’s behaviour.

What makes for successful open data publication?

When I look at successes in open data publication, where both data owners and users feel they are benefiting, they tend to be around:

  1. Where users have very clear ideas about what they want to build and are persistent in pursuing the release of that data. This may be because the data is already accessible in some form under a restrictive licence (eg for money or only for research purposes), so users know it exists and understand its utility.

  2. Where a data owner both publishes data and engages a wider community of potential users (or other parts of their own organisation) to use that data with their own very clear strategic goals in mind, in order to satisfy the owner’s public task more efficiently, for example.

  3. Where data owners and data users are the same types of organisations, who work together collaboratively to maintain a data asset that everyone benefits from.

Identifying shared goals

What’s common in all these situations is the close collaboration between data owners and users. This requires data owners and users to have shared goals, to see the relationship as symbiotic rather than parasitic. It’s rare that everyone starts out with shared goals. It takes time to develop a common language, to identify common problems and to build trust that everyone is committed to solving them.

The Open Data Challenge Series provided a very successful model for bringing together a group of stakeholders around a shared challenge, by taking a sector focus. It had a predicted ROI of 5–10x through getting startups and small businesses to create new products using open data. It wasn’t as successful at getting data owners to publish new or higher-quality data.

This year, we’ll be investigating other ways of bringing together people and organisations across different sectors to bring data to bear on their shared goals.

Jeni Tennison is a Technical Director & Deputy CEO at the ODI. Follow @JeniT on Twitter.

If you have ideas or experience in open data that you’d like to share, pitch us a blog or tweet us at @ODIHQ.