Ten point scales are very popular ways for measuring customer satisfaction or customer experience. It used to be fashionable to make an average of customer ratings and give this the status of an index. However, this suffered from a few cosmetic issues.
The average of 10 and 2 is 6. The average of 8 and 2 is 5. Clearly the second value is worse than the first and it’s obvious why. So everything is fine. Apart from the fact that no customer actually gave a rating of 6; our index value. Nevertheless, it looks like a 6; not great, but acceptable. This also overlooks the fact that there is one very unsatisfied customer, who is likely to be lost. So a second bit of information from the rating is that we will lose half the business. Does the average value say any of this? No, it alters the image.
We’ve already learned this from the Net Promotor Score (NPS). Average values are confusing, so we define three areas for good, bad and inconspicuous. The problem with NPS is the mixer in which these values are spun, but reality is a little closer. Just because we have 10 fingers, doesn’t mean we have to use a 10 point scale. Or have you heard someone say, “I think the new member of staff is really 9.2 on a 10 point scale”. We think in categories of good, better, best, not scales. If necessary, we put ourselves somewhere in between, maybe on a 3 point scale, but this has nothing to with averages. It sticks with different categories. Three times bad is not good, also when three times one “three” is good.
We find these categories in all languages. They are a human characteristic. A 10 point scale however is interpreted in different ways in different cultures. When you ask for something to be interpreted (what is “bad” for you?) and you squeeze several cultures together under a 10 point scale, what do you get? A 3 point scale.