Google review analysis
A business's Google rating is a single number and it does not say what is inside it. A 4.3 hotel may be praised for breakfast and criticised for cleanliness; the average hides both. Review analysis is the work of splitting those reviews by topic and showing which topic is improving and which is not.
Why an average is not enough
An average hides volume and cause at the same time. If forty negative reviews in the last quarter are all about one thing, that registers as a few decimal places — while the operational meaning is single and specific. Reading by topic turns the same data into something you can act on.
- How many reviews mention each topic
- The positive rate for that topic and how many points it moved
- Which reviews are driving the movement
Mixed reviews should not be flattened
"Great location but the room was noisy" is neither positive nor negative; it is two judgements on two topics. nalyzed records it as positive for location and negative for noise, and counts each into its own topic rate.
No evidence, no insight
Behind every finding sit the reviews that produced it. For a sentence to count as evidence it must appear verbatim in the original review; if a model invents one, the result fails schema validation and is quarantined.
Coverage has to be visible
Collection does not always finish. A source may rate-limit, its page structure may change, or access may be blocked. In those cases the job stops and the dashboard shows the source-reported total, the number actually collected and the coverage status together.
Method
Every number on this page comes from the analysis pipeline nalyzed runs over publicly visible Google Maps reviews: normalisation, language detection, personal-data redaction, business-type topic classification and aspect-level sentiment. Results are stored with their evidence quote, sample size and confidence score; a result whose evidence does not appear verbatim in the source review is not published. When collection does not finish, coverage is shown as `partial` and no screen claims that every review was analysed.