One nice thing about testing thousands of ads is that we collect loads of data, and sometimes that data turns out to be useful in ways we hadn’t imagined it would.
For instance, we generally measure emotional reactions to video ads from moment to moment – so we can see where in a commercial happiness peaks, and whether it tails off towards the end. This is vital for working out how ads can be improved – spotting flat parts or unresolved negative emotions.
What we realised, though, is that it also gave us a way of measuring the emotional movement within an ad – in other words, how dynamic and dramatic the ad is. If we added up every time the emotion within an ad changed, we’d end up with a cumulative measure of how emotionally dynamic it is.
We called this measure Flux. If an ad takes you on a rollercoaster ride from happy, to fearful, to sad, then back to happy again, it would score a very high Flux rating. And if an ad just made you evenly happy from the first second to the last, with no other emotions coming in, its Flux would be really low.
What’s your flux capacity?
Investigating this stuff is fun, but it’s only useful if we can relate it to real-world behaviour. We know that if you feel nothing, you do nothing – ads which make people happy are better than ads which don’t, no matter how much dynamism and Flux we detect. So what sort of behaviour might the emotional movement in an ad predict? What is Flux good for? Continue reading