Free getting-started workbook

What’s your next move?

The FatFish getting-started workbook

Give this 30 minutes. Leave with one clear priority, a small test and a plan for the next seven days.

From the founder of FatFish: 20 years helping build businesses, with a degree in AI earned 23 years ago.

Bring what you have

An idea you have started exploring, a service you want to sell, an AI-built prototype or a product already in customers’ hands. You do not need to code or know the startup jargon.

Choose your starting point. Complete that exercise. Put your next move on the seven-day plan. Use a notebook, print the pages or type into the editable Word version.

What you leave with

One clear priority
The question most likely to waste your next week or your next £1,000.
A small test
Something you can run this week, with a time and spending cap.
A seven-day plan
Who you will involve, what result means continue, change or stop, and when you will review.

Free workbook

Send it to me

One email. Print it, write on it, or type into the Word version.

Find your starting point

What is holding you back?

Choose the question that best describes today. You can come back to the other sections later.

Where do I start with AI?

What you will make: A useful task to test, with a way to judge the result.

Is this ready for customers?

What you will make: A list of launch blockers and who will fix them.

How do I get people to buy?

What you will make: One clear offer and a first-customer experiment.

How does this become a business?

What you will make: A basic money check and the next commercial decision.

If you do not know who needs this or whether they will pay, start with customers. If you already have people waiting but cannot deliver safely or reliably, start with readiness. If people buy but each sale loses money, start with the business numbers.

Pick the uncertainty most likely to waste your next week or your next £1,000. Work on that first.

What’s inside

Four short exercises. One next move.

01 / Using AI

Start with a task you can judge

Pick something you do repeatedly and understand well. Start with a task whose mistakes you can spot before anyone relies on the output. Run three small trials, keep a person responsible for the final result, and only keep the workflow if quality holds and effort falls.

A prompt you can reuse without invented prices, dates or promises
A timed comparison of your usual process against AI plus human corrections
A written definition of “good enough” before you test

02 / Getting ready

Can a customer actually use it?

Ask someone unfamiliar with your product to complete its main task while you watch. Record what happens. Do not coach them past problems. This is a first check, not a security audit or compliance sign-off.

Core journey, emails and billing, access and ownership
Accessibility, recovery, capacity, support and obligations
Each finding sorted as blocker, next, or unknown, with an owner and date

03 / Finding customers

Sell one useful outcome first

Choose a small group you can reach and a problem they already experience. Write an offer that says who it helps, what changes and what they should do next. Then run a first test with five relevant people.

A one-sentence offer you can say out loud
A message you can send without a campaign
How to read replies, interest without buying, or a first sale

04 / Building the business

Does a sale leave anything over?

Work out what is left after the costs that come with a sale, then how many sales cover your monthly fixed costs. For an app or subscription, compare monthly revenue with hosting, AI usage, fees and support.

Price, delivery cost and cost to win a customer on one page
The number of sales needed to cover fixed costs
The next decision: change the numbers, or test whether you can sell and deliver that volume

See what a good plan looks like

Small tests. Useful answers.

The workbook includes two fictional examples that turn a vague ambition into a test with a time cap and a decision rule. The targets are for those tests. They are not universal benchmarks.

An AI-built scheduling app

Who: Independent tutors managing lessons by text.

Biggest unknown: Will tutors use a separate booking link?

This week’s test: Invite five tutors to use a test booking flow with fictional learner details. Observe without helping.

A bookkeeping setup service

Who: Freelance designers doing their own bookkeeping.

Biggest unknown: Is a £150 fixed-scope setup session worth paying for?

This week’s test: Speak with five relevant contacts; offer the paid session where the problem and scope fit.

A test earns the next step. A completed demo or one sale does not yet establish a sustainable business.

Your seven-day plan

Make your next move

The last working page asks you to complete this before you close the guide. Keep the test small enough to run this week.

What have I started, and who is it for?
What is the biggest unanswered question?
What exactly will I do to test it?
Who will I involve or contact?
What time and spending limit will I set?
What result would mean continue, change or stop?
When will I review the result?

Put the first action in your calendar now. If you cannot name the action, make the test smaller.

£500 Business Builder Review

If you have competing priorities, an unfamiliar technical problem or an offer you cannot quite explain, bring what you have. We will help you decide where to focus. Before the session we review what you have built, your customer, pricing and the information you share. Then a 75-minute working session on your biggest blocker. Afterwards: three priorities, a practical 30-day action plan, and one follow-up clarification email.

See it on the packages page

Get the workbook

Planning takes around 30 minutes. Running your chosen test takes longer. All examples are illustrative, not client results or promises.

Want the rest of the playbooks and the videos? Get the Learning Pack for £299.

Free workbook

Send it to me

One email. Print it, write on it, or type into the Word version.