Data-Informed vs Data-Driven

We hear a lot these days about the need to be more ‘data-driven’. And I’ve used this phrase a lot in the past to describe the heightened imperative in just about every business to put data more front-and-centre in strategies, processes and decision-making. But more recently I’ve started to use the term ‘data-informed’ instead of ‘data-driven’ because there is a meaningful difference between the two phrases and what they imply.

The latter is far more prevalent (it has 62m results on Google compared to only 7m for ‘data-driven’ – see what I did there), but there is a key distinction between them and good reason to change our terminology, expressed well here by Andrew Chen.

Metrics, says Andrew, are of-course typically based on existing or past circumstances (a reflection of a current product status or strategy, for example, and an existing audience) so they may help you to iterate towards what he calls the ‘local maximum’ (the point at which you hit the limit of the current structural, customer or design foundation) but it’s easy to take this too far.

Evidence based decision-making is great, and still under-utilised in most businesses, but there is a limit to the gains that we can derive from optimisation, and it de-prioritises the bigger picture – the broader aspects of a problem or the potential bigger solution. Data is often systematically biased and the kind of data that is most easy to collect and interrogate is often not the kind that will show us what’s possible or where our end vision could be.

Being data-informed means setting a course (creative, visionary, compelling) and then recognising that you are likely to have only a subset of the data you need at any given time to be ultimately successful. So it is important to break the work down into smaller pieces and use data to help validate and navigate towards the realisation of that end goal.

Like the visual at the top of this post suggests, data may be the new oil (or even the new coal in the sense that, like the early days of the steam engine, innovations are far more useful to those who have copious amounts of the raw material to work from) but as Andrew says, not everything is an optimisation problem and being data-driven rather than data-informed means that we are in danger of over-simplification.

As I say in the book, we need to design with vision and optimise with feedback.

For more like this, order your copy of Building the Agile Business Through Digital Transformation, or you can join our community to access exclusive content related to the book.

Image source


Also published on Medium.