Last updated:
Quick answer
Build six starter templates: 5-star food, 5-star service, mixed 3-star, 1-star food, 1-star service, and suspected fake. Personalize each with one real detail before posting.
At a glance
All outcome figures are estimates from cited sources, not independent verification by us.
- What it is allowed to do
- Drafts for you
Writes drafts (emails, posts, plans, replies). You review and send. Nothing goes out without you.
- Time saved (estimate)
- ~2 hours per month
- Difficulty
- Intermediate
- Tools needed
- ChatGPT, Claude
- Typical cost
- Free tiers of ChatGPT and Claude. No paid tools needed.
Reviewed Oct 19, 2026
Basis (estimate): Estimate basis: assumes ~20 reviews a month at ~8 min each by hand vs ~2 min personalizing a template.
What does this setup do?
Create a template bank for restaurant and cafe review replies that handles praise, mixed feelings, and kitchen disasters without sounding like a chain.
Who is it for?
Restaurant, cafe, bar, and food truck owners who get reviewed constantly and need replies that sound like a human who was actually there.
How do you set it up?
Tools
- ChatGPT or Claude (free tier is fine)
- Your Yelp, Google, and TripAdvisor pages
- Your menu standouts and staff names you want to spotlight
- Your real policies: reservations, refunds, comp meals
Steps
- List your six situations: rave about food, rave about service, mixed 3-star, bad food, bad service, suspicious or fake.
- Ask AI for one template per situation, each under 70 words, in your restaurant’s voice.
- Add your specifics: signature dishes, your name, the phone number for offline follow-up.
- For each new review, pick the closest template and rewrite it around one detail from the actual review.
- Post it yourself. Never auto-post, and never let the template show: change the opener every time.
- Once a month, read your last 20 replies in a row. If they blur together, refresh the templates.
Example output
Illustrative. Your results will differ.
- 3-day trip sketched: flights, stay, and daily plan
- Side-by-side with 2 cheaper alternatives
- Nothing books until you approve it.
Where to paste this prompt
ChatGPT: ChatGPT → Project instructions, or ChatGPT Tasks for scheduled reminders (once a day on Free)
Copy-paste prompt
Adapt bracketed fields to your situation.
Write restaurant review reply templates for me. I will personalize each one before posting.
Restaurant: [name, cuisine, vibe: e.g. family Italian, casual]
My name and role: [e.g. Maria, owner]
Contact for unhappy guests: [phone or email]
Signature dishes: [2-3]
Real policies: [e.g. we redo a dish on the spot; no cash refunds after the visit]
Write six templates, each under 70 words:
1. 5-star praising the food.
2. 5-star praising the service or a named staffer.
3. 3-star mixed (liked X, did not like Y).
4. 1-star about food quality.
5. 1-star about service or wait time.
6. Suspected fake or never-visited review.
Rules:
- Warm, specific, plainspoken. No chain-restaurant corporate voice.
- Bad-review templates: apologize once, invite offline, no arguing about taste.
- Leave [brackets] where I insert the reviewer's detail.Newsletter
Get the free starter pack
Free starter pack with five beginner copy-paste prompts after you confirm, plus three sourced setups in your inbox each week. Unsubscribe anytime.
What usually goes wrong?
Where agents tend to stall on this job. More patterns in Where AI gets stuck.
- Never argue about taste in public. "Our pizza is authentic" against "pizza was bland" helps no one.
- Do not get defensive about health or cleanliness claims. Address calmly, move offline fast.
- Identical replies across Yelp and Google get noticed. Vary the wording per platform.
- Comping a meal publicly in every reply trains readers to complain for free food. Offer it privately.
Which chat tools work with this prompt?
Paste the prompt into any of these assistants (free or paid tiers).
- ChatGPT
- Claude
Common questions
- Should restaurants reply on Yelp differently than Google?
- Slightly. Yelp reviewers write longer reviews, so a slightly fuller reply fits. Google reviewers skim, so keep it tight. The tone stays the same: warm and specific.
- A reviewer lied about what happened. Do I correct them publicly?
- Briefly and factually, once, without heat. Then move it offline. Long public corrections read as arguing, even when you are right.
- How do I handle reviews in another language?
- Reply in their language if you can, even briefly. A short genuine reply in the reviewer’s language beats a polished English one they cannot read.
What tips make it smoother?
- Name the dish. "Glad the carbonara hit the spot" beats "Glad you enjoyed your meal" completely.
- Pass compliments to the team by name. It costs nothing and staff notice.
- Yelp readers skew detailed; Google readers skim. Slightly fuller replies on Yelp are fine.
- For 3-star mixed reviews, acknowledge both halves. Ignoring the complaint reads as deaf.
Where do the claims come from?
Summaries use public sources linked below, not independent verification. How setups are picked.
- Read HubSpot’s write-up (blog.hubspot.com)
HubSpot: Yelp response practices with positive and negative examples by scenario.
- Read MARA Solutions’s write-up (www.mara-solutions.com)
MARA Solutions: Restaurant-specific positive and negative reply examples, including staff mentions.
Related setups
- How to Use AI to Reply to Google Reviews (Even the Bad Ones)
Draft on-brand replies to Google reviews.
- Reply to 1-Star Reviews Without Sounding Like a Robot
Draft calm, human replies to angry 1-star and 2-star reviews that apologize once, offer one real next step, and take the fight offline.
- Ask Happy Customers for Reviews at the Right Moment
Write short, natural review requests for text, email, and in-person handoffs, timed to the moment the customer feels good.