Restaurants and venues
The map listing, the booking platform and the website describe your restaurant with opening hours, dishes and diets that don’t always match. When someone asks where to have dinner, the assistant draws on those sources: TrustInsider shows you which version makes it into the answer.
Before alignment
The closing day is written on the website, but the booking platform keeps the old calendar. Anyone asking for a Monday dinner gets your name and finds the shutters down.
A restaurant lives across several listings that nobody updates at the same moment: the map, the booking platform, the guide, the website. Each has its own hours, its own list of dishes, its own photo of the menu, and the assistant doesn’t know which is the most recent. AI claims flags the statement about the opening hours and the source it comes from.
A sample restaurant listing before and after alignment: the wrong Monday hours removed, the closure added and the holiday opening hours published in advance.
Avoid: the answer says the venue is open on Monday, its closing day, because the map and the booking platform show different hours. Do: a single set of hours, identical on every listing, with holiday opening hours published before the holidays arrive.
Allergies and diets
If the menu is silent, the assistant infers coeliac or vegan options from customer reviews. A sentence read that way becomes a promise the kitchen cannot keep.
A claim about a diet is checked against the published menu: it is there or it isn’t, and with what limits. It should be written on the website in precise words, including the shared kitchen, rather than left for someone who ate at your place once to reconstruct. There is a European regulation on food information to consumers, allergens included.
A sample answer to a search for a trattoria with gluten-free dishes: the claim about the diet, the review it comes from and the menu that doesn’t confirm it.
QuestionA trattoria with gluten-free dishes in the centre?
AI assistant · sample answer
You could try the trattoria near the cathedral: it also makes gluten-free first courses for people with coeliac disease, according to a customer review from last year, which speaks well of the service. The venue’s menu doesn’t mention it and it’s best to call ahead.
The coeliac option is presented as a service of the venue.
The source is a customer’s memory, not a page from the restaurant.
The published menu doesn’t state the option and the kitchen doesn’t guarantee it.
After alignment
A high rating on the guide and a low one on the map send two opposite signals. Replying on every platform that is read, not just the favourite one, makes the picture consistent and easier to pick up.
Once hours and menu are aligned, the work moves on to reviews: Actions drafts replies in a consistent tone, without customers’ personal data, and sorts them by platform. At the next cycle you see whether the assistant’s answer still cites the old hours or the review left without a reply. There are European rules on consumer reviews too. In Italy, for restaurants and tourism, the rules of Law 34/2026 on online reviews also apply.
A sample restaurant compared line by line before and after the work on its listings: hours, menu and diets, reviews and what the answer reports.
| Before | After | |
|---|---|---|
| Opening hours | The map and the booking platform show different opening days. | The same hours on every listing, holidays included. |
| Menu and diets | Dietary options are inferred from customer comments. | The website states what is prepared and with what limits. |
| Reviews | Replies only on the owner’s favourite platform. | Replies on every platform that is read, in the same tone. |
| The assistant’s answer | Reports the wrong day and an option never offered. | Takes hours and menu from the venue’s own pages. |
Customers’ questions
The questions in the panel are the ones a customer really asks: where to eat tonight, a place open on Sunday, a menu for someone with an allergy. Each one tests a specific fact about the venue.
For each question the platform compares what the answer states with your listings and shows where to correct the difference. The questions are typical phrasings, chosen by you for your venue and repeated at every cycle.
Three typical questions from a restaurant’s customers: the fact each one tests in the answer and the page where that fact is corrected.
| Customer question | What is checked | Where it is corrected |
|---|---|---|
| Open on Monday evening? | The hours in the answer against those on the listing | On the map listing and on the booking platform |
| Gluten-free menu? | The claim about the diet against the published menu | On the menu page of the venue’s website |
| Room for a group dinner? | The capacity and the booking method stated | On the website and on the booking platform |
The venue’s measurement
One evening gone wrong can fill reviews and conversations in a single day. The platform recognises the spike and starts from those mentions to prepare a response plan, which a person approves in Actions and which closes with a measurement.
The glimpse shows the response plan on a demo brand: the day’s spike with the mentions behind it and the handover to Actions. In day-to-day work, Reviews collects the venue’s ratings over time and AI claims flags the facts in the answer that don’t match your listings; at the next cycle the measurement tells you whether the spike has subsided.
Frequently asked questions
Why the hours come out wrong, whether one review platform is enough, which reviews to start from and what to do when the menu changes.
Because it also reads the map listing and the booking platform, and doesn’t know which is the most recent. The same hours everywhere remove the ambiguity.
No: the assistant combines several sources. Replying where people leave reviews and where the assistant reads makes the picture consistent.
It’s best to start with recent ones and with those the answers pick up: the sources of each answer show which reviews the assistant is actually reading.
You update the menu page and the listings together; at the next cycle AI claims tells you whether the answers still cite the dishes you removed.
Related paths
Each role uses the platform with a different goal. From here you compare the other solutions, the path from brand to verification and the methodology behind the numbers.
How TrustInsider works in six steps: brand, questions, quote, evidence, page-by-page intervention and a check on the same questions.
ReadHow every TrustInsider measure is built: sampling of AI answers, sources and rights, index calculation and declared limits.
ReadHelp with access, credits, collections, integrations and data: guides, FAQs, the site assistant and support requests from your workspace.
Read