Suicide and customer satisfaction

Sweden is the 7th happiest country in the world based on a Leicester University study, Denmark just a few miles away is no 1. But they both rank extremely highly in the international suicide rate table. For that matter where do you think Switzerland -not unassociated with suicide ranks? ... that's it, 2nd in happiness.

This research is based on a questionnaire along the lines of how happy are you (there are other more sophisticated ways of measuring happiness). But it illustrates a really important statistical point.

If you take all the unhappy people out of your happy survey or they exclude themselves by virtue of suicide then your national happiness score will increase. So any happiness survey needs to find a way of accounting for this, perhaps including their responses until they would have reach the average life expectancy, but that doesn't sound right either. Exactly the same logic illustrates why the FTSE 100 does not reflect the fortunes of the economy, it is because the companies that perform worse fall out of the index. It is called survivor bias, which takes on a whole new meaning when you relate it to happiness and suicide.

And the point. By the same logic we should not exclude customers who have moved to a competitor from our customer satisfaction scores; if we want to use customer satisfaction as an indicator of how we well we are servicing customers. The smarter way to measure this is to use a composite score underpinned by churn, customer sat, net promoter scores. Otherwise it is conceivable you would see customer sat increase side by side with customer churn.

I now have to explain to the family why i have bookmarked a couple of pages on suicide statistics.