7.33 out of 10. That is the median across 137 COLLAB measurements where a brand rated its agency, collected by IKAg since December 2019.
What that number means
A 7.33 feels like "good but not exceptional", and that is exactly what it is. In practice this scale runs from 6 to 8.5: below 6 a relationship usually ends by itself, and almost nobody gets above 8.5. So a 7 is not a bad report and an 8 is exceptionally good.
If you measure your own agency and land on 6.5, you are below the middle but not in the red. The question then becomes: on which axis?
The three axes, because that is where the difference sits
- Attitude: 7.68. The highest of the three, year after year.
- Process: 7.47.
- Expertise: 7.30. The lowest, and the most striking finding from six years of measuring.
Agencies are therefore judged best on how they are and least on what they know. At question level it is sharper still: across 166 measurements agencies score high on reachability (8.36), trust (7.91) and enthusiasm (7.82), and low on using data (6.49), steering timing (6.76), "we learn from the agency" (6.81) and value for money (6.85).
Six of the ten lowest-scoring questions sit in the Expertise axis, while "is an expert in their field" tops that same axis. Clients do consider their agency skilled; what they miss is that skill reaching them.
What to do with it
Do not use the number as a report card but as a starting point. Below 7 on process is usually fixable in a quarter with agreements. Below 7 on expertise is a conversation about the team and about seniority, and that takes longer.
And measure both sides. The other way round, when the agency rates its client, the median is 7.57: agencies are slightly more positive about their clients than the reverse.
See also: how to evaluate your agency and does my agency think things are going better than I do.