Two numbers describe a tribe's reviews, and they answer different questions.
| Score | What it is |
|---|---|
| Average rating | The plain average of the stars. Simple, and easily distorted by a small number of reviews. |
| Bayesian score | The average adjusted for how many reviews there are, so a tribe with a handful is not ranked above one with hundreds. |
Why the adjusted score exists
A tribe with three five-star reviews has a perfect average and has proved very little. A tribe with four hundred reviews averaging 4.7 is demonstrably good. Ranking on the plain average would put the first one top of your network table every month, which nobody believes.
The adjusted score pulls scores with few reviews towards the middle, and lets them rise as the volume grows. It is the fairer basis for comparing locations.
What to look at
- Total reviews — volume is the thing most locations are short of.
- Average rating — the number customers see.
- Response rate — how much of it you are answering.
- The adjusted score — for ranking tribes against each other.
VOLUME FIXES ALMOST EVERYTHING
A tribe worried about one bad review is usually a tribe with twelve reviews. Asking every customer for a month does more for the score than any amount of agonising over the bad one.