Hang on that must be true because IBM said so back in 2013. That article is worth a read – not because the basic premise stands up, but because it correctly identifies three of the most important data management components that are foundational and integral to any organisation recognising that data is an asset. We’ll come back to that because those four small words have suffered a huge amount of misunderstanding, misdirection and misuse. First though let’s return to the less trumpeted parts of data management. Down in the operational trenches, data stewards and other practitioners are in a war to maintain and – if possible – improve their data asset. They face many combatants who seem determined to fight any concept of an integrated an coherent data management strategy. Three weapons at their disposal are data governance, information asset ownership and data quality.

The last one of these should be an output of the previous two. Data quality is however rater more often deployed the bazooka to blast away at the wrong types of issues. There’s regularly collateral damage when DQ is somehow held up as the solution to an organisations chaotic data supply, and it rarely achieves the stated outputs. This is because what we’re looking at here is the wrong end of the dog.

The problem I have with the ‘data is oil’ concept is oil has a finite quantity. There’s only so much in the ground. Data isn’t like that at all, it’s increasing at what would be an unsustainable rate were it not for Moores’ law and the availability of cheap compute power and extremely low cost storage.  Throwing all your data into a datacenter or, as is ever more prevalent, a cloud service can exacerbate the problem of what to do with these endless terabytes of noughts and ones. Information – data with context – is more like gold, you have to mine through a whole load of worthless stuff before hacking out a nugget.

I keep seeing ‘BI’ or ‘Big Data’ or ‘Advanced Analytics’ proudly represented on an organisations portfolio. But for all the advances in these areas, the basic data management tenet holds: ‘Garbage In, Garbage Out’. BI especially has the goal of creating a single version of truth or trust, but is hamstrung before it starts by the poor data management framing its data sources. Data can only become an asset with more than transactional value, if culture, process, activities and technology follow a rigorous management discipline.

So the frustration – especially – senior staff have with data is they do not see it as any kind of asset, whatever the public declaration. It has such a low trust value it cannot replace instinctive decision making with empirical analysis.  We live in a world of workarounds and data silos because no one cares to make the link between genuine business benefits and the practice of data management. Well, they do actually because it’s these forgotten data warriors in those trenches who absolutely get it. They just aren’t very good at selling it.

For the last four months, I’ve been working on a data capability toolkit for HEDIIP. We’re really proud of what we’ve created because it’s taking an Enterprise Architecture approach to improving data management. The Higher Education sector is going through some genuinely transformational change in many different areas, and most of these can be very well supported by getting your data house in order. It’s a slog though – and like any change programme without a clear line of sight between organisational wider aspirations, risk mitigation and benefits and specific activities and frameworks to improve data management, it’s not going to get the priority, funding and focus it needs.

If organisations truly value data as an asset, then they see gold in the careful analysis of rich and varied data sources. More and more do and this is a good thing. But the response of firing up a BI project and riding in on the technology solution horse is not the right one. The analogy just about holds if you think of buying a £5m mining digger, setting up camp with a hundred workers and then drilling into entirely the wrong mountain. Obviously we’d never do that. But we do it with data because it’s largely invisible, generally poorly managed andnot supported by a simple governance structure and given insufficient priority and resources to improve it.

Good data management is hard to do. But my view is many of the practices I see (both within HE and outside) are unsustainable. Data is not like oil. It’s far more important than that.