Numbers that look good and mean little.
WHAT LOOKS GOOD AND MISLEADS
Tickets closed Average handling time First-contact resolution, measured loosely Satisfaction scores with low response rates
WHY TICKETS CLOSED MISLEADS
Closing is not resolving, and it is trivially inflated.
WHY AVERAGE HANDLING TIME MISLEADS
Fast is not good if the problem returns.
WHAT TO PAIR IT WITH
Reopened rate.
WHY FIRST-CONTACT RESOLUTION MISLEADS
It depends entirely on how it is defined, and definitions drift toward flattery.
WHAT TO DEFINE PRECISELY
Whether a follow-up question counts as a second contact.
WHY SATISFACTION SCORES MISLEAD
Extremes respond, and small samples swing wildly.
WHAT AVERAGES HIDE GENERALLY
The slow tail, which is what customers remember.
WHAT TO USE INSTEAD
Percentiles.
WHY
They describe the worst experiences, which drive reputation.
WHAT TO BE SUSPICIOUS OF
Any metric improving while complaints rise.
WHAT THAT INDICATES
The metric is being optimised rather than the outcome.
WHAT TO TRIANGULATE WITH
Reopened tickets Repeat contacts from the same customer Cancellation reasons
WHAT TO NEVER TIE TO INDIVIDUAL PAY
Satisfaction scores.
WHY
It produces rating requests and corrupts the data.
WHAT TO REPORT
A small number of measures, honestly interpreted.
