Readiness · Guide

Is your business ready for AI? A plain readiness checklist

The short answer

A small business is ready for AI when five things are true. There is one job that repeats and works badly. Someone can say yes without a committee. The inputs to that job are reasonably stable. You can name what leaving it alone costs. And you want to own the result, not rent it. You do not need a technical team, a data project or a budget line labelled AI. Readiness is mostly about having a clear problem, and most businesses have one.

This guide gives you the five conditions, a way to score them, and the first move for each one that is not yet true. It takes about ten minutes.

Why "readiness" is usually the wrong question

Many readiness checklists ask about data platforms, governance committees and technical skills. Those matter for large organisations. For an owner operated business, they are a distraction, and they make readiness feel like a project in itself.

The question that matters is narrower: is there one job here where AI could help, and are the conditions around it right? If yes, you are ready to try it small. If not, the checklist tells you what to fix first, and that is usually a day or two of work, not a transformation.

The five conditions

1. One person can say yes

The test: can one person decide to try something and sign it off, without a long approval process?

AI projects stall when everyone has an opinion and nobody has the authority. In a small business this is usually the owner or a principal. If the answer is "it depends who you ask", the first job is to name the person.

First move: write one sentence naming who decides, and what they are allowed to decide without asking.

2. A repeated job works badly

The test: is there a task that happens at least weekly, follows a pattern you could explain, and costs you time or errors?

Good candidates are the Sunday night report, the quote built from scratch each time, the enquiries that arrive by three different routes, and the data copied between two systems. Poor candidates are one off tasks, tasks that depend on a relationship, and tasks you could not describe to a new employee.

First move: list your five most repeated jobs and pick the one you dislike most.

3. The inputs are reasonably stable

The test: does the job start from similar information each time, and is that information somewhere you can reach?

AI is good at working through similar material at speed. It is poor at guessing what is not written down. If the key facts live only in someone's head, or in a pile of unlabelled files, the first step is to write them down.

First move: write the job on one page, as if for a new starter, including the awkward cases.

4. You can name the cost of leaving it alone

The test: can you say, in hours, errors or lost enquiries, what the problem costs you each month?

You do not need exact numbers. You need a rough one. "Our office manager spends about ten hours a week on invoice entry" is enough. Without a cost, you cannot tell whether a fix is worth it, and you cannot tell afterwards whether it worked.

First move: estimate the weekly hours or the missed enquiries, and write the number down.

5. You want to own the result

The test: do you want something you keep and control, or are you happy to rent access to someone else's system?

Both are legitimate. If you want to own it, your first questions to any builder are about code, accounts and handover, and our guide to ownership covers them. If you are happy to rent, check what happens to your data and what leaving costs.

First move: decide which you want before you talk to anyone.

Score yourself

Give yourself one point for each condition that is true today.

• 5 points: you are ready to start a small pilot on that one job.
• 3 or 4 points: you are close. Fix the gaps first, which is usually quick, then start.
• 1 or 2 points: do the groundwork. Name the decision maker, pick one job and write it down.
• 0 points: talk to someone before you spend anything. You may not need AI yet, and that is a fine answer.

Our two minute readiness check on the Implementing AI page gives a more detailed picture across ten questions.

What readiness is not

• It is not about tools. Choosing a tool before choosing a job is how businesses end up with an impressive demo and nothing useful.
• It is not about size. A five person business with a clear problem is more ready than a large one with a vague wish.
• It is not about budget. The first useful project is usually small and cheap to try.
• It is not about hype. Being behind is not a reason. A specific cost is.

Responsible use, in plain words

The Australian Government has published voluntary guidance to help organisations adopt AI sensibly. The Guidance for AI Adoption, published by the National AI Centre in October 2025, sets out six essential practices. In plain words they ask you to:

1. Decide who is accountable for each AI use.
2. Understand who it affects and plan for that.
3. Measure and manage the risks.
4. Share the essential information about how you use it.
5. Test it and keep watching how it performs.
6. Keep a human in control.

For a small business, this does not mean paperwork. It means a named owner for each use, a check on the output, a way to switch it off and a plain note of what it does. The full guidance is linked in the references below.

Customer information

Before you put customer details into any AI tool, find out what the tool does with them. Some tools keep what you send, some use it to improve their product unless you switch that off, and terms differ between free and business plans.

Privacy law may also apply to you. The Office of the Australian Information Commissioner says a small business is one with an annual turnover of $3 million or less, and that some small businesses must comply with the Privacy Act regardless of turnover, for example health service providers and businesses that trade in personal information. Rules in this area can change, so check the current position with someone qualified. This is general information, not legal advice.

A safe starting rule: keep personal details of customers out of any tool until you know its terms, and use the least information that does the job.

Your first thirty days

• Week 1: name the decision maker, pick the one job, and write it on one page.
• Week 2: estimate the cost of the job today, and decide what a good result looks like, with two real examples.
• Week 3: run a small trial beside your normal process, with someone checking every output.
• Week 4: compare. Keep, change or stop, and write down what you learned.

Common reasons it stalls

• Nobody owns the outcome, and the project becomes one person's experiment.
• The job was never written down, so the tool guessed.
• Success was never defined, so nobody can say whether it worked.
• The first project was too big.
• The result lives on someone else's account and cannot be moved.

Frequently asked questions

What is AI readiness?

It is how well your business is set up to get real value from AI. For a small business it comes down to five practical conditions: a decision maker, a repeated job that works badly, stable inputs, a known cost and a wish to own the result.

How do I know if my small business is ready for AI?

Score yourself against the five conditions in this guide. Five yes answers mean you are ready to try one job small. Fewer means there is groundwork to do, and the checklist tells you what.

Do I need a technical team to use AI?

No. You need a clear problem and someone who can decide. A technical person helps with building and looking after the result, but the first step is about the job, not the technology.

Do I need clean data first?

You need the information for one job to be reachable and reasonably consistent. You do not need to clean everything. Start with the inputs to a single process.

How long does a first project take?

A well chosen first job can be tried in a few weeks. The time goes into writing the job down, trying it beside your normal process and checking the results, not into the technology itself.

What should I automate first?

A job that repeats weekly, follows a pattern, has a mistake that is small and easy to catch, and annoys you. Our list of ten processes worth automating first is a good place to look.

Is it safe to use AI with customer data?

It depends on the tool and its terms, and on your obligations under privacy law. Until you know both, keep personal details out of the tool and share the least information that does the job.