Someone reminded me today of the Abraham Wald story.
If you can't be bothered to click the link, he was a World War 2 statistician who helped improve the armour on planes to protect against enemy bullets.
His major contribution was to tell everyone they were looking at it the wrong way.
They were putting more armour on the spots where planes, that made it home from battle, had the most bullet holes.
The right answer was to put more armour where there were NO bullet holes - because you had to assume the planes that didn't make it home got shot somewhere else.
There's some obvious learnings here when it comes to research about researching the whole sample.
But it also shows us that data is a waste of time if you don't have the skill to analyse it properly, and sometimes work out what it isn't telling you.
Sometimes we forget that planners were invented to make sense of research, because research can sometimes be very dangerous in the hands of researchers.
Step forward New Coke. Where, decades ago, they changed the Coke recipe to be sweeter, more like Pepsi. But they tested the new recipe on one sip, when many sips, which is how people actually drink, made the sweetness too cloying.
Or that's one side of the story. The other is that even getting the 'functional' research methodology right can be flawed, as the real reason everyone kicked off about New Coke, forcing Coca Cola to change the recipe back, costing millions, is that most of our relationship with brands isn't built on function.
It's built on belief and not having to think about it.
They kicked off about New Coke because we don't like change.
But then sometimes we do, we just don't know we do.
There's the cliche that Henry Ford telling us that if he asked people what they wanted, they would have asked for a faster horse.
There's the way Apple made the Ipod when MP3 players weren't getting much traction, because he realised it just needed to be simple and nice to use one.
There's the fact that the data from social media tends to contradict the data from search.
Why Should you Care?
It's easy to ignore data, or leave it to the data scientists. But there's gold in this stuff if you know how to look.
Just as it's unfashionable to see any use in focus groups, but loads of clients still use them, and believe what people tell them, or even worse, believe the analysis of their moderators.
Research doesn't make intelligent decisions for you, it helps you make more intelligent decisions.
Planners, strategists or detectives? Maybe a bit of both.
Comments