# The complete Google review reply guide (2026 edition)

> The full playbook — from framework and tone to reply-rate benchmarks, edge cases, and the operations changes that stop bad reviews at source.

_Published 2026-07-22 · Updated 2026-07-22 · 18 min · By The HappyReplies Team_

Source: https://happyreplies.com/resources/review-reply-mega-guide

This is the guide we would have wanted in year one. It's long because the shortcuts don't work — every 'reply to reviews' shortcut we've tried has either lowered the star average or lost the brand voice within a month.

## The framework

Every reply, regardless of star rating, does four things: it names the reviewer, it names one specific detail from the review, it takes a position (agreement, disagreement, or ownership of a fix), and it offers a concrete next step. Four sentences, in that order, works for 95% of reviews.

## Tone by star rating

- 5-star: warm, brief, specific. Don't over-thank.
- 4-star: acknowledge the miss, thank the visit, invite back.
- 3-star: this is the trap — sound too grateful and you look tone-deaf, sound too apologetic and you invalidate the review. Match the reviewer's calibration.
- 2-star: ownership without over-apology. Move to email fast.
- 1-star: measured, named, offline within two sentences.

## Reply rate targets

100% on 1- and 2-star, 60% on the rest, within 24 hours for anything below 4 stars. See our benchmarks piece for the sourced numbers.

## Edge cases

- Fake reviews: reply publicly with a two-line neutral note while flagging.
- Ex-employees: neutral, non-defensive, flag with the conflict-of-interest report.
- Anonymous reviews: reply without personalisation but with the specific fix.
- Star-only, no text: single-line reply keyed to the star count.
- HIPAA / regulated industries: never confirm the person was a patient/client.

## The operations loop

Replies compound. But they don't fix the underlying pattern — three cold-food replies in a month is a kitchen problem, not a reply-craft problem. Run a monthly review of the top three themes in your 1- and 2-star reviews and change one operational thing.

## What to automate, what to keep human

Draft with AI, approve with a human, ship. Full auto-post is where every case study we've seen goes wrong — the wrong reply on a fake review or a suspected sabotage attempt can compound the damage.