Problems don't respect administrative boundaries - neither should your data

At Nesta we have noticed an emerging trend nationally to tackle community problems at a sub-regional level. Increasingly councils at a sub-regional level are joining up across public sector silos to gain wider insights, unlock additional value from data and tackle community problems that don’t respect administrative boundaries, such as alcohol harm in the North East. We’ve seen this from Humberside to Essex, the West Midlands Combined Authority to the North East and Greater Manchester to London.

If you are involved in a project to do something similar, what steps do you need to take to prepare?

From my many years of working on public sector data projects, and more recently at Nesta, I’ve learnt that there are some surprisingly simple steps you can take now to ensure you get results. And it’s not about project management methodologies but taking a balanced approach. The biggest pitfall is to think that these projects are exclusively about technology, and not address the human elements.

Ask yourselves these questions to get off to the best start.

What do you want to change?

  • This needs to be a problem which you have the ability to influence, with significant and specific impact in communities - which can be measured. Collectively, across your city combined authority or sub-region, do you have the influence, resources and policy levers to make this change happen? For example, if you discover that some communities are not well served by existing services, do you have the ability to commission alternative services, or stimulate the market?

  • Do you have the enthusiasm and leadership backing to stick with a project that will last several months and potentially get difficult and boring at times?

  • Do you have an existing partnership with shared common objectives to steer this at a senior level? Perhaps a LEP board.

Who wants it to happen?

  • Make sure all partners are motivated and will benefit. This can be done by getting everyone to commit resources to the project. Each partner needs to put something in - be it financial, time or staff - in order to get something out of it.

  • Show brave dynamic leadership with a proportional appetite to risk and willingness to put their weight behind the project when things get tough.

  • Take a relentless outcomes focus for both the project and the service transformation. For instance, in Nesta's office of data analytics pilot in London, we will only know if it works once inspectors on the ground have checked the properties we identified through data. Don’t forget to use the insights you have gained.

How are you going to get the data?

  • Talk to your legal teams and information governance teams and start a culture change where information sharing is undertaken through a risk assessment approach using privacy impact assessments. Not sharing data may be more of a risk to vulnerable people than sharing the data. In our experience, disproportionate information sharing risk aversion is the single biggest barrier so far to data projects and it does not have to be like this.
  • Do you know what each partner’s starting point is in terms of how they currently use data? Ask each partner to candidly carry out a data maturity readiness assessment, Wise Council Report Appendix 2, so that work planning and ambition can be accurately calibrated

  • Cross reference and link geographic and people data linking your people records within your organisation. As well as bringing data together, bring your data analysts together across business areas.

Who can help?

Most councils will want to develop internal capacity - some will want external support, depending on their level of data maturity. Here’s some practical advice from our experience of running a number of workshops around the country, on what to focus on.  

External organisations who can help

There are a huge range of companies and products that can help - from SMEs to multinationals, to universities and not-for-profits.

Our advice is:

  • If you buy tools make sure they are simple to use, sustainable with low ongoing revenue impact, and deliver cashable savings. You don’t want to pay a consultancy to add every additional data set, for example.

  • Data science partners need to be able to communicate and demystify concepts for other partners, and bring modern cutting edge technologies such as machine learning.

  • Data visualisation partners need to be dynamic/creative/engaging and ensure that their work is useful rather than merely interesting or visually appealing.

What do you think? I’d like to hear your views, or about how you’re getting on with your data projects. Please comment or respond to me on Twitter for further information.


Hilary Simpson

Hilary Simpson

Hilary Simpson

Hilary worked in the Government Innovation Team at Nesta, helping to unblock information sharing issues for the Office for Data Analytics pilot projects.

View profile