How to Research Pricing and Competitors

The most confident wrong answer you will ever read

Ask any AI assistant what a freelance bookkeeper charges in Leeds and you will get a number. It will arrive fast, in a tidy range, with a sentence about regional variation and maybe a bullet list of factors. It will look exactly like research.

It is not research. It is a very good impression of research.

The AI is not looking anything up. It is producing the kind of text that usually follows that kind of question, and numbers are just more text. It has no idea what your competitor down the road charged last Tuesday, and it will not tell you that unless you make it.

This lesson is mostly about that problem, because pricing is the one area where a confident wrong answer costs you real money for years.

What AI is genuinely good at here

It is excellent at structuring the research. It cannot do the research, but it can tell you what to look for, in what order, and what questions your pricing needs to answer. That is worth having, because most people price by feel and a vague memory of what someone said in a Facebook group.

I run [what you do] in [where]. I am reviewing my prices for [the service].
Do not estimate any prices, rates or market figures. If you do not have a source
I have given you, say so instead of producing a number.
Give me:

A checklist of what I should go and find out, most useful first.
Where each item can actually be found (public price lists, directories, trade bodies, asking customers).
Five questions my pricing needs to answer.
An empty table for me to fill in as I gather the real figures.

Now feed it the real numbers

Once you have gathered actual prices, AI becomes properly useful again. Paste in six competitor price lists you copied from their websites and it will find the pattern faster than you will, because you are too close to it.

Below are prices I copied from six competitor websites this week, plus my own.
Every figure here is real and checked. Do not add any figures of your own.
Tell me:

Where I sit in this group, and what that says about how I look to a customer.
What the higher-priced ones include that I do not.
Two ways I could restructure my packages without dropping my price.
Anything in the data that surprises you, and what I should check next.
THE PRICES:
[paste what you gathered, with the source next to each]

The conversation worth having

Where AI earns its keep in pricing is not the number, it is the argument. Most sole traders undercharge because they have never had anyone push back on their reasoning, and asking a friend is awkward.

Here is my price and how I arrived at it: [explain it, including your costs and
how long the job takes].
Play three roles, one at a time.

A sceptical customer: what makes this feel expensive?
A rival charging thirty percent less: how would they undercut me?
A blunt business mentor: what am I failing to charge for?
Be specific and do not be reassuring. Do not invent market data. Work only from
what I have told you.

Run that once and you will usually find at least one thing you have been doing free for years. A bakery discovers it has never charged for delivery. A photographer realises the four hours of editing after every shoot has quietly become unpaid labour.

A quick sanity habit

When an AI hands you any figure, ask it one follow up: where did that come from. If it names a specific, checkable source, go and check it. If it says something like based on general industry knowledge, you have your answer. That is the machine politely telling you it made the number up, and Gemini, Claude and ChatGPT will all do it.