Insights — AI Workflow & Commercial Automation — 3 min read
Should You Use an AI Chatbot on Your Website? A Commercial Decision Guide
A chatbot is an easy thing to add to a website and a surprisingly easy thing to get wrong — the question worth asking first is commercial, not technical.
In short
An AI chatbot is worth using on a website where enquiry volume is high enough that faster first response and basic qualification add real commercial value, and where there is a clear escalation path to a person. It is usually not worth it for low-traffic or highly consultative sites, where a poor chatbot experience can do more damage than the response-time gain is worth.
Website chatbots are now common enough that adding one feels like a default decision rather than a deliberate one. That is usually the wrong way round: a chatbot should be justified by what it does commercially — captures a lead, answers a genuine FAQ faster, qualifies an enquiry — not installed because competitors have one.
This guide sets out the cases where a chatbot earns its place, the cases where it causes more harm than good, and a simple framework for deciding which applies.
What a chatbot is actually being asked to do
Before choosing a chatbot, it helps to separate the jobs it might be given, because the commercial case is different for each one.
| Job | What it involves | Commercial value |
|---|---|---|
| Answering FAQs | Opening hours, pricing ranges, how a service works | Saves time only if the questions are genuinely repetitive |
| Lead capture | Collecting name, contact details and a brief enquiry outside office hours | Useful if enquiries are currently being missed |
| Qualification | Asking a few questions to route or prioritise an enquiry | Valuable where sales capacity is limited and not every enquiry is a fit |
| Customer service | Resolving simple account or order queries | Depends on how many queries are genuinely simple and repeatable |
| Booking or scheduling | Offering available slots directly | Useful where availability is a frequent bottleneck |
When a chatbot genuinely helps
- Enquiry volume is high enough that a person cannot respond quickly, and speed of first response matters to conversion.
- A consistent set of questions is asked repeatedly and can be answered accurately without nuance.
- There is a clear need for out-of-hours lead capture that currently goes to voicemail or an unmonitored form.
- A simple qualification step would help prioritise which enquiries a salesperson looks at first.
- There is a defined, working escalation path to a person for anything the chatbot cannot resolve confidently.
When a chatbot is the wrong answer
- Enquiry volume is low — a chatbot adds a layer between the visitor and a person without enough volume to justify it.
- The sales process is genuinely consultative and early trust depends on speaking to a knowledgeable person, not a bot.
- There is no realistic escalation path, so visitors who need a person get stuck in a loop.
- The business cannot keep the chatbot's answers accurate as pricing, services or policies change.
- The goal is simply to look modern, with no defined job the chatbot is meant to do.
Why a bad chatbot experience costs more than no chatbot
A chatbot that gives a confidently wrong answer, loops a frustrated visitor, or has no way to reach a person, is often worse than a simple contact form, because it actively damages the first impression rather than leaving it neutral. This is the single biggest reason chatbot projects should start narrow — covering a small number of well-defined questions — rather than attempting to answer everything from day one.
A simple decision framework
| Question | If yes | If no |
|---|---|---|
| Do we get enough enquiries that faster response matters? | Chatbot may help | Probably not worth it yet |
| Are the same questions asked repeatedly? | Good candidate for FAQ automation | Consider a simpler contact route |
| Can we keep answers accurate as things change? | Proceed | Fix the maintenance plan first |
| Is there a working route to a person? | Proceed | Build this before anything else |
Commercial ROI considerations
The honest way to judge ROI is against the specific job the chatbot is given — captured leads that would otherwise be missed, or enquiries answered without tying up a person's time — rather than a general sense that it should help. Evans does not promise guaranteed savings or conversion uplifts from a chatbot; the realistic value depends on current enquiry volume and how much of it is genuinely repeatable.
Human approval and escalation
Any chatbot handling lead qualification or customer queries should have a clearly defined escalation point, and nothing it says should commit the business to pricing, terms or promises without review. Where customer data is collected, this should be handled in line with UK GDPR requirements, and specialist advice may be needed for data handling questions.
Not sure which processes are worth automating?
The free automation diagnostic asks a few questions about how work moves through your business and points to the processes most likely to be worth automating. No email required.
Related services