Data interoperability- a teamwork challenge
You may well have missed it amongst the truly newsworthy and titillating morsels that came from the Dominic Cummings Select Committee appearance this week, but he also bemoaned the lack of data interoperability within government. He related tales of people frustrated in their attempts to join data together so that social care could be better supported.
I’m sure it was tricky, because as Jeni Tennison remarked at the ODI summit last year, there was precious little aggregated or aggregate-able data about the care sector because it was never monitored. The need for interoperability had never really been considered.
So that’s the thing, you can have all the cleverest geeks but it will be super hard for even them every step of the way if the need for interoperability has never even been considered before. It will be even worse if you ignore the professional data scientists and try and do all your calculations with the wrong tools, you might even think that it’s another argument in favour of data infrastructure.
Unfortunately, I see the argument for data interoperability all too often fall exclusively to the people who love to talk technical, let’s lovingly call them geeks. These technophiles bring it upon themselves to an extent, swept up in the pure intellectual joy of solving a really hard problem and jousting with their fellow geeks about relatively small differences. As a result, the non-geeks, walk off, happy in the delusion that data interoperability is solvable and solved and believing that there is no change for them to make.
I fear that we non-geeks are wrong. Glorious (and hard) data science requires data which has been treated like an asset and invested in. We all know that sound investments are based on good judgement and hard choices. It requires decisions to be made and I suggest the following as questions to ask about what data should be prioritised to be made and maintained as interoperable;
- How useful, really, is that data?
What is the data specification and has it been consistently applied?
2. What constraints exist on the data use for us and for others?
What is the data governance?
3. What are we expecting to happen if we share this data?
What is the business model and risk appetite for the data transaction?
These are tricky questions that require the geeks and non-geeks to interact. We need higher levels of data and digital transformation literacy and a shift away from restricting data discussions to the infrastructure on which the asset sits, the technology. I’m not sure if these are the right questions, but successful enterprises hold people accountable for creating value from assets and data assets should be no different.
