Contextualisation – the Known Knowns & the Known Unknowns

Suddenly there is a lot of noise  about contextualising data. I was trying to explain this to someone when I realised there was a very personal explanation, and it highlighted the dangers of doing this badly

Prejudice. Good and bad.

Prejudice occurs when you take a (often small)  piece of initial data, generally visual, about another person or thing and then add previously acquired data to form a view, right or wrong. We all do it on a daily basis, we’re human and it’s how we work. The problem comes from the quality of the data added to the only bit of absolute visual- in this case – data. Our brains are stuffed with opinions and experiences related to the image and we happily add them. The result takes a piece of raw data, and by contextualising it with additional – possibly poorly sourced and referenced –  data can lead to a very skewed and often dangerous conclusion.

For an example of good and well intentioned prejudice think of the example from the Highway Code about seeing a football bounce into traffic?  The point is to influence how you contextualise that data. The aim is for you to see the ball and immediately think that it means that there is a higher than normal chance that a child will appear chasing the ball. The idea is that inexperienced drivers make not make that connection soon enough so it is a blatant and worthy  attempt to prejudice their thinking. Ball=kid=cover the brakes.

Businesses are slowly waking up to the fact that it’s not how much data they hold but what they do with it that matters The skill as a person or a business is understanding more about the data that we are turning into facts. The integrity of the data source, how many iterations it has gone through, if it’s been tested, is there enough of it, are all factors that can influence the quality of the contextualisation process. This means not only does the base data need to be clean, well collated, de-duped and generally of a high quality the same holds true for the additional data and processes used to contextualise it. As sentient beings this is an example of how much more sophisticated our brains and programming are over any software program. Think how fast you can take a piece of visual data and contextualise it and then in an instant test your hypotheses, inbuilt prejudices, additional inputs and prior experience and that of others. In an instant you can re-contextualise that initial data and form an entirely different view.

We are seeing the first attempts to do and sell this ability with reference to customer data. The aim is to understand your customer at that moment in time so you can target your offerings or even choose whether you even compete. This is so cool and the people that nail this are going to be v. rich. Why do you think Facebook and Google are valued so highly? It’s not only the data but what the market believes they’ll be able to do with it.

In the meantime, I miss Bush and Rumsfeld mangling the language. It was an easy way to raise the spirits and now they are gone.

Advertisements