My organization is moving from a 10 point satisfaction scale to a 5 category Likert scale for all satisfaction surveys. Our final metric will be the percentage of "Happy Customers" we have (those who will say they were either "somewhat" or "very" satisfied with our service).



I need to calculate baseline data for next year. At first I assumed:

1-2 = Very Dissatisfied

3-4 = Somewhat Dissatisfied

5-6= Neither satisfied nor dissatisfied

7-8= Somewhat Satisfied

9-10= Very Satisfied.

So.... 7-10 is a "Happy Customer".

This gave us some really high numbers. We tried to set targets a little lower to allow for some wiggle room in our assumptions. But leadership is now arguing that if our numbers are so high, surely our targets must be high too. Re-calculation now to assume that 8-10 is a "happy" customer.... But...



Has anyone else made this switch? What sort of assumptions did you use to set baselines? How accurate were there? Anyone know any articles on how respondents interpret the different scales?