Skip to content
PortuguêsPilot programme

Automatic replies to Google reviews for restaurants

Quick answer

In one sentence

Renver answers safe Google reviews automatically, at any rating, in the language the guest wrote in, using only facts the restaurant confirmed. Sensitive topics and cases with no confirmed facts go to the manager, with an alert. You can pause at any time and take a published reply down.

Every review passes checks before any reply is written

Nothing is written until the review has been read and classified. These are the steps:

  1. 1. Sensitive topics

    Rules in each supported language catch illness, allergies, legal threats, safety, discrimination and money.

  2. 2. Read by AI

    A classifier weighs the context, the tone and its own confidence.

  3. 3. Sarcasm and manipulation

    An ironic compliment hiding a complaint, and attempts to trick the AI, go to a person.

  4. 4. Enough facts

    If an honest answer needs something you have not confirmed, nobody invents it: it goes to the manager.

  5. 5. The text is checked

    A second agent checks every sentence against your confirmed facts: no promises, no admission of fault, no invented detail.

  6. 6. Limits and pause

    An hourly cap, and automatic replies pause when a wave of negative reviews arrives.

What always goes to a person

These topics never get an automatic reply, whatever the rating:

  • Illness and food safety
  • Allergies
  • Legal threats
  • Safety
  • Discrimination
  • Refunds and charges

And when facts are missing, or the language is different

If an honest reply depends on something you have not confirmed, nobody makes it up: the review goes to the manager. The same goes for reviews in languages other than Portuguese, English and Spanish, instead of getting a reply nobody has checked.

What a reply looks like

An illustration: the reply names a real detail about the restaurant, tied to what the guest said.

“The picanha was perfect, we'll be back!” — ★★★★★

The usual reply

Thank you for your feedback! We hope to see you again soon.

With real details

So glad the 400 g picanha came off the charcoal just the way you like it! We serve it from midday every day here in Vila Madalena — next time, try it with our banana farofa.

You stay in control

You choose whether automatic replies apply to positive reviews, negative ones or both, you can pause at any time, and you can take a published reply down. Everything is logged. There is an hourly limit, and replies pause automatically if a wave of negative reviews arrives.

Each plan has a monthly allowance per location: up to 300 new reviews on Essencial and up to 1,000 on Profissional. As you approach the limit you are told, and nothing is charged without your agreement.

What we do not do

  • We do not promise a position in Google or a mention by AI assistants: nobody controls those answers.
  • We do not do review gating — filtering who is asked for a review — which Google prohibits.
  • We do not reply automatically to sensitive topics.
  • We do not invent information: a reply uses only what you confirmed.

Go further

Want to see this at your restaurant?

The pilot runs for 30 days at one location, and there is no payment when you sign up: we run a demo and confirm the price, scope and start date before any charge.

See the pilot programme

Frequently asked questions

Does Renver publish replies without my approval?

Yes, when the review is safe — that is the point: taking the routine off your hands. Before anything is published, every review and every reply goes through automatic checks. You choose whether automatic replies are on for positive reviews, negative ones or both, you can pause at any time, and you can take a published reply down. Sensitive topics are never answered automatically.

What about negative reviews?

Ordinary complaints — a long wait, a cold dish, the service, the price — get an automatic reply that acknowledges the problem, without promising a refund or admitting fault, and mentions a change only if you confirmed it. If the complaint involves health, allergies, legal threats, safety, discrimination or money, or if an honest answer depends on a fact you have not confirmed, it goes to the manager with an alert.

Can the AI invent facts about my restaurant?

Replies use only facts you confirmed, at most one per reply, and only when it relates to what the review was about. A second agent checks the text against your confirmed facts before publication; if something does not match, the reply is rewritten once and otherwise goes to the manager.