Everything below is measured against 798 ranking bar listings across 40 US cities, not against a generic local SEO checklist. The full measurement sits on the bars benchmark page.
Where these come from. Google Maps results for “bar in {city}” across 40 US cities, collected 1 September 2026, signed out. Re-measured quarterly. Full method and findings.
This is the counterintuitive one. 55% of ranking listings sit at 4.5 or above, which means one in two does not. Rating is not the gate it is sold as.
And at the top it stops separating anyone. 19% of ranking listings are already at 4.8 or better, so the move from 4.7 to 4.9 buys you almost nothing in position.
What it does buy you is the click. Two listings side by side at the same position, one at 4.9 and one at 4.4, do not get the same share of calls. So we stop reporting rating as a ranking risk and start reporting it against calls, direction requests and booking clicks, where the effect is real and measurable.
Alerting on the trailing average rather than the lifetime one, because the lifetime figure hides a bad quarter behind years of history and by the time it moves you have lost two.
Routing a negative review to whoever can actually resolve it, within hours, with the response written by someone who knows the case.
Watching for the patterns that precede a suspension: a burst of reviews from one location, edits that keep reverting, a competitor filing redressal reports against you.
What the engagement actually contains, and what each part is measured against.
| Included | What it means in practice |
|---|---|
| Trailing-average alerting | Alerts on the trailing average rather than the lifetime figure, which hides a bad quarter behind years of history until it is too late to act. |
| Escalation with an owner | A negative review routed to whoever can resolve it, with a deadline, rather than into a shared inbox. |
| Rating against conversion | Rating tracked against calls, direction requests and booking clicks, because that is where it demonstrably moves. |
| Suspension risk watch | Review bursts, reverting edits and competitor redressal reports, all of which precede a suspension and none of which are visible without monitoring. |
| Reported against the benchmark | Every metric shown next to the measured figure for bars, so a number is always "against what" rather than floating on its own. |
| Per location, never per group | A group average hides your weakest site. Everything is reported per location first and rolled up second. |
Switch reporting off the lifetime figure in week one. It is the number that lets a bad quarter go unnoticed until it is two quarters.
A negative review with a named owner and a deadline, not a shared inbox and good intentions.
Move rating out of the ranking report and into the conversion report. With 19% of ranking listings already at 4.8 or better, it is not explaining your position.
Review bursts, reverting edits, competitor redressal reports. All invisible until something is looking, all slow to undo.
The median ranking listing carries 351, with the middle half between 101 and 896. That spread is the real answer: there is no single threshold. The full measurement, including how it differs by city size, is on the category page.
Lower than you have been told. 55% of ranking listings sit at 4.5 or above, which means one in two does not. Rating decides who clicks you, not whether you appear.
Only if it breaks a policy: it is spam, off-topic, a conflict of interest, or names an individual. Reviews that are merely unfair stay, and the response you write underneath is what future customers read.
Not directly, and anyone claiming a measured ranking effect is guessing. It helps the thing you actually want, which is the share of people who read the profile and then call you.
The free audit reads your live Google profile and puts it against the measured benchmark for your category.