The clear answer: score plus evidence

A higher score means raters liked the taste more, but Deskbrew does not declare one universal cutoff that turns every workplace into 'good office coffee'. Context matters: compare the number with its rating count, confidence state and other offices in the same city.

An 8.2 based on many recent ratings is more meaningful than a single 10/10. The first is repeated evidence; the second is one person's opinion and remains heavily discounted by the scoring method.

Why the displayed score is not a raw average

Deskbrew uses a Bayesian prior centred at 6.5. Thin samples stay closer to that neutral centre, while repeated ratings move the score further. Recent opinion carries more weight so a machine or supplier change can eventually affect the result.

Provisional versus ranking-eligible

A provisional score appears before Deskbrew has enough evidence to place the office in an official ranking. This gives early contributors feedback without letting a handful of ratings claim a city lead.

A ranking-eligible office has cleared the current confidence and integrity rules. Only then can Deskbrew publish a genuine city rank or call the location a city leader.

Use the city and company comparison

The city ranking supplies the strongest local context because every listed office cleared the same eligibility rules. A company page answers a different useful question by comparing the employer's physical locations with one another.

Read the methodology for the scoring rules, then rate your office if the current number does not reflect the coffee being served now.