Early in George Eliot’s great novel Middlemarch, the young idealistic heroine Dorothea Brooke marries, spectacularly unwisely, Edward Casaubon. Casaubon has spent his life assiduously travelling the world, buried in libraries, collecting every single source and scrap of information he can about the mythologies of various cultures which he will synthesise in his great opus, The Key to All Mythologies. Of course this work is never finished, there’s always one more piece of information, one more text to consult, one more scrap of data to uncover and then everything will suddenly fall into place, and the cultures of every civilisation, since the dawn of time, will be explained.
Against this backdrop appears the figure of Will Ladislaw. Will is a young, impetuous, fiery romantic, very much in the Keats / Shelley / Byron mode which was so fashionable at the time. He has little time for Casaubon’s great project. He trusts his own instincts. He knows not everything can be explained neatly and logically, but thinks the world is a better place because of it. He is more interested in imagination than information.
You can probably tell what happens next, and I won’t spoil the ending for you. Go read it. It’s ace.
But you’re probably asking what any of this has to do with a blog about communications planning, though the more astute of you will have realised that I’m going to talk about data. Sorry, Big Data.
Real time, unfiltered insight. Petabytes. Dashboards. It seems that data is the new black in our business. And though it may not be perfect at the moment, just around the corner, it’s going to be great. The Key to All Mythologies, in fact.
But data is a just a tool. Of course, it can help us make better decisions, and let’s not forget that our business has always been based on data (BARB, NRS, TGI etc.), and there are areas in which, within certain parameters, data is absolutely fundamental (search, for example). But the danger is we extrapolate from these scenarios and fool ourselves into thinking that the more data we have, the better our answers. In fact, we don’t need ‘strategy’ any more, we just need dashboards. If we can just get enough data, fast enough, all we need to do is react to it, optimise it and, hey presto, we’re sorted. This approach is wrong. Not only is it wrong, it sells short both the power of data and the power of our own imaginations. So here’s my manifesto for data….
More is not, necessarily, more
In The Black Swan, Nassim Nicholas Taleb discusses at length the fact that more information does not necessarily make us better at diagnosing situations and predicting outcomes (which, is of course, what we’re supposed to do). Citing numerous examples, and a whole series of experiments, he illustrates that increasing information does not increase understanding. In actual fact, the opposite is often true. What he does point to, though, is that increasing information does dramatically increase people’s confidence in their predictions. I won’t go into the reasons for this now (it’s a lot to do with belief persistence and confirmation bias), but suffice it to say that some data is important (in that it has predictive power) and some isn’t. And it’s vital to understand this distinction, otherwise you’re just drowning.
Deal with what we’ve got, not what we’d like
Despite what you may read in Wired or Fast Company, we will never find the Key to All Mythologies. Data is not suddenly going to ‘stop’ when Google announce they have produced an algorithm that predicts, with 100% accuracy, all of human behaviour. Of course, we will make advances, but data utopia is not just around the corner.
A helpful way of thinking about this is Gregory Treverton’s distinction between puzzles and mysteries. A puzzle can be solved with more information, mysteries require judgements and the assessment of uncertainty (Treverton used the whereabouts of Osama bin Laden as an example of a puzzle and the problem of what would happen after Saddam’s fall in Iraq as a mystery). The trap that we need to avoid is to treat mysteries as puzzles and to believe that if we just get that last final scrap of data all will become clear.
(If you want more, Malcolm Gladwell has written about this at some length – it’s very good)
Use data to inspire not just optimise
Lastly I’ve noticed a tendency to assume data is just about spreadsheets and optimisation. Whilst this has its place, I think it’s vital that we increasingly put data upfront in our planning, not just at the back end in executional optimisation. This requires rigour, but it also requires imagination – insight and understanding is often found in numbers, in data. As some of our creative agency partners have increasingly moved away from the sorts of rigorous analysis of consumer data originally envisaged bv Pollitt and King, this is an area in which we can add huge value to our clients plans.
Have fun out there.
