When does a chatbot actually make sense? A simple checklist for small businesses

A chatbot is not a goal in itself. It makes sense when it solves a problem you actually have — and becomes an expensive distraction when it does not. Here is how to work out which category you fall into.

Start with the problem, not the technology

The most common reason a chatbot fails is that it was built because "everyone has one", not because it solved anything. Before you consider a solution, ask one question: which specific problem is it meant to remove?

Good answers are measurable. "We spend too much time on the same questions" is concrete. "We want to look modern" is not.

Signs a chatbot makes sense

A chatbot usually pays off when several of these are true for you:

• You get a lot of repeated questions. Opening hours, delivery status, returns, "how do I do X" — the same questions, day after day.
• The questions have fixed answers. If the answer lives in a document or a routine, a chatbot can fetch it.
• Customers ask outside working hours. A wall of emails every morning suggests people need answers while you sleep.
• The volume is high enough to notice. A handful of enquiries a week rarely justifies a dedicated solution.
• The answers rarely change. Stable facts are easy to maintain; prices that swing daily are not.

If three or more of these ring true, it is worth a closer look.

Signs you should wait

It is just as useful to know when the answer is no — at least for now:

• Every question is unique. If customers never ask the same thing twice, a chatbot has little to lean on.
• The answers require judgement. Cases involving money, legal matters or feelings belong with a human.
• You have not written the knowledge down. If the answers live only in the head of whoever has worked there longest, the chatbot has nothing to learn from.
• The volume is low. If you get three enquiries a day, a good FAQ page is often enough.

None of these means "never". They mean "not yet".

A simple test you can run yourself

You do not need a big project to find the answer. Go through the last hundred customer enquiries and sort them into two piles: those with a fixed answer, and those that require a judgement call.

If the first pile is large, you have a good candidate for automation. If the second pile is bigger, the money is better spent making your people faster — not replacing them.

This exercise takes an afternoon and often saves you from a misplaced investment.

What a chatbot does not fix

It is tempting to think a chatbot fixes poor service. It does not. If the problem is unclear answers, slow follow-up or a website nobody can navigate, a chatbot merely moves the problem — it does not solve it.

A chatbot amplifies what you already have. If your routines are good, they get faster. If they are messy, the mess just becomes more visible.

Start narrow if you go ahead

If you decide a chatbot makes sense, do not build it to answer everything from day one. Pick the five to ten most common questions, let it handle those well, and expand once it has proven itself.

A narrow solution that answers the important things precisely beats a broad one that answers everything roughly. And an always-available route to a human is not optional — it is what makes customers willing to talk to a bot at all.

If you are unsure which category your business falls into, you are welcome to book a quick call and we will look at the numbers together.