Data Governance is not Data Science. Which is a shame because it is the latter that excites the senior leadership of most organisations I work with. This is understandable as these individuals are interested in high quality information which informs the decisions and actions they are making on a daily basis. Much of this information is purportedly an output of Data Science.

What is less understandable is how the good management and governance of data is somehow divorced from the use of it. This is not a new problem; in the last ten years ‘Data Science’ could be replaced by ‘Analytics’, ‘Predictive Modelling’ and ‘Machine Learning’. The outcome is the same, a focus on the thing, not on what makes the thing work.

So rather than continue to whinge about it, we’ve come up with a new visual metaphor to cast light on the old adage of ‘Garbage In, Garbage out’.  This is the idea that a Data Governance capability builds an enabling environment – a lab if you will – to ‘perform good science’. Further the Data Governance team is analogous to a lab technician in ensuring those scientists have what they need to be successful.

If you take this metaphor a little further, you will find:

  • Building one lab with all the capability you need is far more effective than building multiple smaller specialist labs (Data is in silo)
  • Extending and expanding the capability of a single lab creates new opportunities for everyone not just a single organisational function (Data is an asset)
  • The role of the lab technician is understood and respected (Data is being professionalised)
  • Problems with the lab are reported centrally but solutions are likely to be – at least in part – implemented by specialists (Data ownership and stewardship is embedded)
  • Lab capability is a central function, but usage is federated (Data is managed)

The takeaway should be that “Data Governance as being the lab and data science the work being done IN that lab

While this may feel like a metaphor which works specifically for Education, it really isn’t. What we’re trying to do is create a value chain which visibly ties the ‘wanted’ outcomes to building a sustainable and scalable capability to support those outcomes.

This is so important, especially now with data  under the spotlight. The decisions I talked about at the top of this article are both time sensitive and significant. Futures of organisations and peoples roles within them are being decided every day. In uncertain times, we need certainty in our data.

It’s easy to see how the value of Data Governance is getting lost in all the noise of the world we’re operating in right now. There’s pressure to focus on narrow and unsustainable data quality cleansing exercises for single uses. We need a message to cut through this to showcase the opportunity to build a professional data framework to support both what you need now, and what you are definitely going to need in future.

That message is simply ‘Stop building more rubbish labs, instead build one good one’. A good one has all the capability you need and will operate at a lower cost.  Sure, there will always be cases where specialist ‘labs’ are needed, but these are the outliers of the bell curve not the centre of it!

Back in the day when I used to ‘sell’ Enterprise Architecture for one of the Big 4 consultancies, we often lamented that senior staff ‘didn’t get it’. That was absolutely the case as it is for data governance. The difference between then and now was we thought that was their fault!

So let’s not repeat that mistake. We need to establish what our outcomes are, so we know what kind of capability to build. We need to build and measure that capability based on the professionalism of managing and governing our data. We need to make that capability available to everyone and keep improving it to fit changing organisational needs.

The Lab concept came out of discussions with my colleagues in the Higher Education Data Governance Network. This network is made up of practitioners from over 50 UK universities. So we have prime ground to test our metaphor.

Obviously, I’ll report back. Until then though, I’m always interested in your challenges and feedback!

If you found this article interesting, it’s not the first time I’ve attempted to address the issue. Here’s something I wrote nearly two years ago: