AI chatbot examples: six website conversations that turn questions into customers

Most chatbot examples online are screenshots of menus. Choose an option, enter your email, someone will be in touch. They are examples of what not to do. The conversations below are the patterns we see work on business websites, written to show the shape of a good exchange rather than a script to copy.

1. The availability check

A visitor asks whether a service, room or property is available. The chatbot answers yes, adds one helpful detail and asks a single follow-up that reveals intent: Is this for you, or for someone else? Is it for a weekday or a weekend? It then offers a time. Three messages, one booking.

2. The price question

Roughly what does this cost? The weak answer is It depends, please call us. The strong answer gives the range or the factors that set the price, then asks the two questions that would let someone quote properly. The visitor feels answered and the business gets a qualified enquiry.

3. The nervous first-timer

Common in dental, legal, medical and fitness. The visitor is hesitant and the question is really Will this be okay? The chatbot reassures in plain, warm language, explains what the first visit involves and makes booking feel small. Tone matters more than information here.

4. The urgent problem

Our heating has stopped and it is minus two. The chatbot recognises urgency, captures the essentials in two questions, flags the lead as urgent and tells the visitor exactly what happens next. No menus, no forms, no waiting for office hours.

5. The comparison shopper

The visitor is clearly weighing options and asks a pointed question. The chatbot answers honestly, including where your business is not the right fit, and offers something concrete: a sample, a case study, a call with the owner. Honesty here wins the ones worth winning.

6. The returning customer

Someone who has bought before and wants to reorder, rebook or check something. The chatbot recognises the context, skips the introductions and gets to it. Familiar customers should feel familiar.

What they have in common

Each one answers before it asks. Each asks only what a good salesperson would ask. Each ends with a next step rather than a dead end. And each hands the business a lead with the story attached. That is the standard to hold any chatbot to, including ours.