The dilemma of customer support ratings

I am really impressed with 37signals’ implementation of their Custumer Support Happiness Report. But we have found a general problem of this kind of systems after talking with a colleague of our service/support department about it.

Quick summary

  • Every answer from the 37signals team includes an immediate opportunity for the customer to rate the support answer. There are 3 smileys with a label to classify the answer as “great”, “ok” or “not good”.
  • The report page presents statistics based on all those ratings, including an average in the box on the right and the latest 100 ratings at the bottom.

So far so good.

The general problem is, that a customer tends to rate the suggested solution of his problem, not the quality of the support answer. So it doesn't matter how well the supporter has written that something is not possible. Next thing? The customer is upset about the fact, doesn’t care about the nice answer and erroneously rates the answer as “not good”. So when it comes to a supporter’s rating (a rating average based on all his/her answers), despite the fact he/she is generally nice and polite, his/her ratings are bad because of given facts which aren’t even in his/her power.

veröffentlicht am 14.03.2011 um 18:15 von Martin Labuschin in ,

blog comments powered by Disqus
The Ruby on Rails Link Library Spanning Sync SEO-Dokumentation Gowalla
ProWebApps WellDone (β) Pinboard (α) Netzwerk Münsterland
Blogs Bücher Frameworks iPhone Mac Web-Apps