Websites & front desk

A website chatbot that answers with your own content.

A website chatbot helps your visitors find an answer and move on to the next step.

What you need

Unanswered questions: where a website chatbot begins.

Your information is on the site, but not always easy to find. A dedicated base helps visitors understand your offer, your services and your terms. What remains is to decide what may be said publicly, and what should point to a person.

Le socle de cet usage

Les connaissances à réunir

Pages publiques du site, FAQ, fiches services, conditions et informations pratiques validées.

Comprendre la base de connaissances
In your day-to-day

What a website chatbot changes.

Guide the discovery

Present your services starting from the visitor’s question.

Make information reachable

Gather the useful answers in one place, beyond the site’s menu structure.

Spot the gaps

The questions you receive help you enrich your FAQ and your content.

A concrete example

A website chatbot: from knowledge to answer.

Illustrative example: the real result depends on the sources, the model and the configuration.

Question ou consigne

« Quelle formule convient à une équipe de cinq personnes ? »

Le chatbot retrouve les conditions de vos offres dans la base publique, présente les critères de choix et oriente vers la page de détail ou un conseiller.

Getting started

A website chatbot on a first scope.

01

Prepare a public base

Select only the content meant for visitors.

02

Configure the answers

Set the chatbot’s role, its tone and its limits.

03

Add it to the site

Use the WordPress or HTML integration, then test the key questions.

The framework to plan for. A public chatbot must use a base meant for the public. Provide a route to a human when an answer is missing or the situation is personal. In every case, decide who approves the answers before they go out, how often the sources are refreshed, and what must stay out of scope.
Frequently asked questions

What you want to know about the website chatbot.

Can the chatbot make an answer up?

It draws on the sources you have chosen. The risk falls with a clean corpus and a narrow scope, but the standard answers still need reviewing.

What happens when it does not know?

Better to plan an explicit way out: a pointer to a form, a phone number or a person, rather than an approximate answer.

Are visitors' conversations kept?

How long conversations are kept, and what they are used for, are decided with you and then stated in your privacy policy.

Let us start simply

Let us look at your website chatbot with your own data.

A first conversation to choose the corpus, the users and the scope worth having.

RDV démo

Besoin d'aide ?