At least without context! As fascinating Big Data is to me, there is a danger that people will stop thinking about the results and take everything as truth. I witnessed this last weekend at the fitness gym. I was able to have a look at my record in their “CRM” and a big red bubble instantaneously sparked my interest. I asked the trainer what this means and she replied “That your are 100% likely to leave us”. I was shocked! I’m going to leave … but wait! That’s not the case. I asked further “why are you supposing this?”, her reply “maybe because your come here once a week”. She couldn’t add more criteria and was more guessing than knowing. Aha. That’s the truth and further more this year is not a good year in terms of fitness for me (I`m not going into details), but I’m not having any intention to leave them. Two points are obvious here:
- Data are nothing without context: They only had to ask me, about my motivations or reasons for not visiting more often
- The decision for marking me as “likely to leave” was not transparent
And the result: They are 100% wrong!
But it showed one thing again: Validation of results is necessary. Not only for Dr. House. Don’t blindly believe things only because they appear on a computer screen.
Oh, this story has no happy ending, because they are not able to correct their data. So in their eyes will still 100% likely to leave. But I will stay, because CRM and computer systems are not their core competency. They are a lot better in helping me work out.