How to respond to Google reviews with AI (without sounding fake)

Updated October 21, 2026 · App names checked against vendor documentation; vendors rename things, so search the step picker for the key word if a name differs.

Why replying to reviews is worth the effort

People reading your Google Business Profile see your replies as well as your reviews. A thoughtful reply to a happy customer reads as appreciation; a calm reply to an unhappy one often reassures the next reader more than the review itself worries them. Google also says that responding to reviews shows you value your customers' feedback.

The trouble is time. Writing a fresh, specific thank-you for each review takes a few minutes, and replying to a 1-star review while you're annoyed is a bad idea. AI fixes the first problem, and a simple rule fixes the second.

The one rule: AI never auto-posts negative replies

Split reviews by star rating:

• 4–5 stars with a comment: the AI drafts a short, specific thank-you. Once you trust the output, these can post automatically.
• 1–3 stars: the AI drafts a calm reply and sends it to you. You read the review, check what actually happened, edit the draft, and post it yourself.
• Rating with no text: a short, warm one-liner (or nothing, if you prefer).

The AI can't see your job history. It doesn't know whether the customer's account of events is accurate, and it should never argue about facts in public or reveal anything about the customer.

Step 1: The AI review reply prompt

Business: [Your business name]. [One sentence on what you do and where.]
Tone: [warm, plain-spoken, not gushing]. Sign off as: [Your name], [Business name].
Contact for problems: [phone or email]

Write a public reply to this Google review.
Reviewer: {{Reviewer Name}}
Rating: {{Star Rating}} out of 5
Review: {{Comment}}

Rules:
- 4–5 stars: thank them by first name and mention one specific thing they said. Under 50 words. Vary the opening; don't start with "Thank you so much for your kind words."
- 1–3 stars: calm, non-defensive, no excuses, no arguing about facts. Acknowledge their experience, apologise that it didn't meet expectations, and invite them to contact [phone or email] so we can make it right. Under 70 words.
- Never mention any details about the customer or the job beyond what they wrote.
- Never offer refunds, discounts or freebies in a public reply.
- No hashtags, no emojis, no keyword stuffing.

Don't stuff your city and services into review replies to chase rankings. It reads as spam to the next customer, and it's not how reviews help you.

Step 2: Build the automation

You'll need a verified Google Business Profile and an automation tool. Connecting Google Business Profile means signing in with the Google account that manages your listing.

Zapier:

1. Trigger: Google Business Profile › New Review, choosing your location.
2. Action: ChatGPT (OpenAI) › Conversation (or AI by Zapier) with the prompt above, mapping the reviewer name, star rating and comment.
3. Action: Paths by Zapier. Path A: rating 4 or 5 → Google Business Profile › Create Reply with the AI text. Path B: rating 1–3 → Gmail › Send Email or Slack › Send Direct Message to you with the review and the AI draft.
4. For the first two weeks, send every review down Path B so you can see what the AI writes.

Make: Google Business Profile › Watch Reviews (shown as Google My Business in some accounts) → OpenAI › Generate a response → Router: high ratings to the reply module, low ratings to an email or Slack message to you.

n8n: Google Business Profile Trigger (Review Added) → OpenAI node → IF rating ≥ 4 → Google Business Profile node (Review › Reply); otherwise Gmail or Slack to you.

If you'd rather not auto-post anything, skip the reply step entirely and send every draft to yourself. Copying a finished draft into Google takes about 20 seconds.

Step 3: Templates for when you reply by hand

5-STAR
Thanks, {{First Name}}. Really glad the [job] went smoothly. We appreciate you trusting us with it, and we're here whenever you need us again. - [Your name]

NEGATIVE (holding reply)
Hi {{First Name}}, I'm sorry your experience didn't meet the standard we aim for. I'd like to understand what happened and make it right. Please call or email me directly at [phone/email]. - [Your name], [Business name]

NEGATIVE (after it's resolved)
Thanks for talking it through with us, {{First Name}}. We've [what you changed or fixed], and we appreciate the chance to put it right. - [Your name]

What a good AI reply looks like (and a bad one)

Say a customer writes: "Showed up on time, fixed our water heater in an hour, and cleaned up after. Ask for Dan!"

• Bad (generic): "Thank you so much for your kind words! We strive to provide excellent service to all our valued customers." It could be a reply to any review, anywhere.
• Good (specific): "Thanks, Priya. An hour start to finish is what we aim for, and I'll make sure Dan sees this. Enjoy the hot showers! - Maria"

If your drafts sound like the first one, add an example of the second to your prompt. One good example improves the output more than another paragraph of instructions.

Mistakes to avoid

• Arguing in public. Even when the customer is wrong, the reply is really for the next person reading it. Take details offline.
• Revealing private information. Never confirm appointment details, addresses or anything the reviewer didn't share.
• Offering compensation in the reply. Make it right privately, not publicly.
• Identical replies. Twenty copies of the same thank-you look automated, because they are. The prompt's 'vary the opening' rule helps.
• Asking only happy customers for reviews. That's review gating, and it's against Google's policy. If you automate requests too, ask everyone the same way.

The other half: getting more reviews

Replies help with the reviews you already have. To get more of them, set up automatic, policy-safe Google review requests. It's one of the two automations in the free sample, alongside the missed-call text back.