I once asked an audience of data and marketing specialist from different industries: how many of you have a churn model? Almost all of them raised their hand. Now, I asked, how many of you are making more than a million euros with this every year? Chuckles. No hands in the air.
The problem is that when you know who is likely to churn, you can get a higer churn rate when acting upon it. It is liking waking them up. Newsday seems to have cracked that code. They reduced churn by 58%in a certain group by using persuasion modelling, which identifies groups of target customers with high or declining engagement who may be re-engaged with gifts or incentives. They avoid disturbing what they call ´self-selectors´. These people like to make decision on their own and retention efforts could have an adverse effect.