As of 13 August 2026, AI can draft responses to your UK customer reviews.
Most people should do this themselves; the prompt is on this page.
Can you do it?
5 minutesto a draft.
15 minutesto something you’d act on.
Cost, all in£0
Skill needednone
Who has to check ityou
What the alternative costsBotpress is an open platform for building LLM chatbots and agents, while manual drafting remains an internal staff task with no price stated in the supplied sources.
If this goes wrong: a reply sounds dismissive or promises something your business cannot provide, and you can correct it before publishing if a person checks the draft.
What to actually do
Do it yourself
The route this page recommends
A chat interface, no skill needed, and roughly 15 minutes until you can act on the result.
How to actually do it
- Open your review platform and copy the review text, removing names, order numbers, contact details and any other information the public reply does not need.
- Gather the relevant verified facts from your order, booking or support records, plus your current refund, complaint and escalation policies.
- Paste the review, business name, verified facts, tone and policies into the prompt, and ask the chatbot to draft one reply.
- Read the draft against the original review and confirm that every factual claim, offer, apology and timeline is supported by your records or policy.
- Route reviews involving safety, discrimination, privacy, legal threats, vulnerable customers or unresolved complaints to the responsible colleague instead of publishing the draft automatically.
- Edit the reply into your normal brand voice, remove private information, then publish it manually through the review platform.
Prompt
Draft a response to the customer review below for a UK business. Review: [PASTE REVIEW] Business name: [BUSINESS NAME] Approved facts about the customer's order or visit: [PASTE ONLY RELEVANT VERIFIED FACTS] Our tone: [DESCRIBE TONE, FOR EXAMPLE WARM, PLAIN AND PROFESSIONAL] Our response policy: [PASTE RELEVANT POLICY, REFUND RULE OR ESCALATION ROUTE] Write one concise public reply in British English. Acknowledge the customer's specific point without repeating private information. Do not invent facts, compensation, policy, staff actions or timelines. Do not argue with the customer or make legal admissions. If the review raises a serious complaint, safety issue, discrimination allegation, data concern or matter that needs account access, mark it "ESCALATE TO A PERSON" and write a neutral holding reply instead. If the available facts do not support a claim, say what needs checking rather than guessing. Return only the draft reply followed by a short "CHECK BEFORE PUBLISHING" list of any factual claims, promises or missing information.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
Use a tool built for this
Second choiceHand it to a person
The distant thirdA person who owns the outcome does this end to end, worth it when the failure is dear.
What it gets wrong
- AI cannot know whether an implied apology or promise is acceptable under your internal complaint process.
- AI cannot verify the customer's account, order history or the latest service recovery decision unless you supply those facts.
- AI cannot reliably judge whether an ambiguous review hides a safety, discrimination, privacy or legal issue.
- AI can make a generic reply sound specific by inventing details, so every claim still needs checking.
- AI cannot take responsibility for the relationship with the customer or the consequences of publishing the reply.
Even on a YES, the friction has a name: judgement under ambiguity, consent and privacy and stakes of error.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 2 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 9 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT reply to my customer reviews?
- Yes. It can draft replies from the review, your verified facts, your tone and your policies. Check every claim and publish the response yourself rather than allowing an unchecked reply to go live.
- Should I let AI automatically respond to customer reviews?
- Use automatic drafting, not automatic publishing, unless your rules and escalation routes are tightly controlled. A person should handle reviews involving safety, privacy, discrimination, legal threats, vulnerable customers or unresolved complaints.
- How do I get AI to write professional responses to bad reviews?
- Give it the exact review, the verified account facts, your response policy and a plain description of your tone. Tell it not to invent details, compensation or timelines, and ask it to flag anything that needs escalation.
- Can AI respond to Google reviews for my business?
- AI can draft a response that you then paste into your review platform, if that platform allows replies. Remove unnecessary personal data, compare the draft with your records and policy, and publish it manually.
Nearby answers
- Can AI analyse sentiment in my UK customer reviews?YES
- Can AI analyse feedback from my Google reviews?YES
- Can AI analyse open-ended answers in my customer survey?YES
- Can AI calculate my Net Promoter Score?YES
- Can AI categorise my customer feedback automatically?YES
- Can AI create a customer feedback dashboard?PARTLY
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.