Most AI readiness assessments sell you a maturity score and a roadmap. This one is a self-check you run yourself in ten minutes, and it measures the five things that actually decide whether an AI workflow will work in your business: whether the task repeats, whether your information is written down, whether someone checks output, whether your team has permission, and whether you can tell if it worked. Score each section out of 8. Anything under 24 means the problem is not the technology.
Key takeaways
- Readiness turns on your inputs, not your tools. Every business already has the same tools.
- The single biggest predictor of success is whether the source material exists in writing.
- A score under 24 out of 40 means fix the inputs before buying anything.
- Businesses score lowest on measurement, which is why so many AI projects cannot prove anything.
- The assessment costs nothing and takes ten minutes. Paid audits should start from a stronger position than this.
Section 1: task repetition
Score 0 for no, 1 for partly, 2 for yes.
| Question | Score |
|---|---|
| Does at least one writing or admin task repeat every week | |
| Does that task follow roughly the same shape each time | |
| Could you name the three tasks that consume the most staff hours | |
| Does more than one person do the same task differently |
Four questions, 8 points. A business that cannot name its repeated tasks has a process problem, not an AI problem.
Section 2: source material
| Question | Score |
|---|---|
| Do your prices, services or specifications exist in a written document | |
| Are your standard inclusions and exclusions written down anywhere | |
| Could you hand a new staff member the material they need to do the task | |
| Is that material current rather than two years old |
This section predicts success better than any other. An AI assistant with your material produces a usable draft. The same assistant without it produces confident generalities, which is the entire reason people conclude AI does not work for their business.
Section 3: review and approval
| Question | Score |
|---|---|
| Does a person currently check this work before it leaves the business | |
| Is it clear who that person is | |
| Could that person spot a wrong figure or a wrong scope | |
| Would anything go badly wrong if a draft went out unchecked |
Score the last question 2 if the answer is yes, something would go wrong. Knowing where the risk sits is readiness, not a weakness.
Section 4: permission and access
| Question | Score |
|---|---|
| Is anyone in the business allowed to use AI tools on work material | |
| Do you have a paid account instead of a shared free login | |
| Is it clear what data must never go into a general tool | |
| Does the team know who decides that |
A business where AI use is informal and unspoken is a business where working workflows never get shared, because nobody is sure they are allowed.
Section 5: measurement
| Question | Score |
|---|---|
| Do you know roughly how long the task takes today | |
| Do you know how often it happens | |
| Could you tell in a month whether it got faster | |
| Do you know what an error currently costs you |
Most businesses score lowest here, and it is why so many AI projects end with nobody able to say whether they worked.
Your score
| Score | Band | What to do |
|---|---|---|
| 32 to 40 | Ready | Build a workflow this month. You have the inputs, the checks and the baseline |
| 24 to 31 | Nearly ready | Fix the weakest section first. Usually source material or measurement |
| 16 to 23 | Inputs missing | Write down prices, inclusions and standard answers before spending on AI |
| Under 16 | Process first | The gap is process documentation. AI would only produce faster inconsistency |
The uncomfortable finding in the bottom two bands: the work that makes you ready is the same work that makes you easier to hire into, easier to hand over and easier to sell. None of it is wasted.
What to do with a low section score
Low on repetition. You may be a genuinely bespoke business. Look at internal admin rather than customer-facing work.
Low on source material. Spend two hours writing down prices, inclusions, exclusions and your ten most common answers. This single step moves more businesses from failure to success than any tool change.
Low on review. Name the checker before you build anything. An unowned output is how wrong information reaches a customer.
Low on permission. Write three lines: who can use AI, on what material, and what never goes in. Here is how to set team rules.
Low on measurement. Time the task once, this week, and count how often it runs. Here is the measurement method.
Then build one workflow
Readiness is only worth measuring if something gets built. The workshop is a full day at Avani Mooloolaba Beach Hotel on the Sunshine Coast, A$495 excluding GST, capped at 15 people. Bring your score, your weakest section and one repeated task with a safe example, and you leave with the workflow built and tested against your own material.
Private onsite training runs at your workplace across South East Queensland for roughly 4 to 20 people. See the workshop, or request a team quote.
Frequently asked questions
What is an AI readiness assessment?
A structured check of whether your business can actually operate AI workflows: repeated tasks, written source material, a review step, permission and access, and a measurement baseline. Vendor versions often produce a maturity score. This version produces a score and a specific next action.
How much does an AI readiness assessment cost?
This one is free and takes ten minutes. Paid audits from Australian providers are usually the entry point to a proposal, and several local providers run the first session at no cost. A paid assessment should tell you more than this page does. Compare local providers.
What score do you need before starting?
24 out of 40 is the practical floor. Above 32 you should be building this month. Below 24, fixing the inputs is faster and cheaper than any tool.
Why does source material matter so much?
Because an AI assistant answers from what it is given. Supply your prices, rules and format and it drafts your work. Supply nothing and it writes the average of the internet, which is exactly what people mean when they say the output was useless to them.
Is AI readiness different from digital readiness?
Yes. You can be poor at software and strong at AI readiness if your processes are written down and someone checks the work. The reverse is also common: modern systems, nothing documented.
Do you need clean data for AI?
For workflow drafting you need accurate documents, not a data warehouse. Data quality matters for analytics and prediction work, which is a different and more expensive project.
What is the fastest way to raise your score?
Write down prices, inclusions, exclusions and your ten most common customer answers. That usually moves two sections at once.
Should you hire a consultant if you score low?
For process documentation, no. That is internal work nobody can do for you cheaply. Once your inputs exist, training or a build both become viable. Compare training with consulting.
How often should you re-run this?
Every quarter while you are building workflows, because the measurement and permission sections change as usage spreads.
What is the first workflow to build once you are ready?
The repeated task with written inputs and an existing checker. Usually quoting, enquiry replies or a weekly report. Use the task selection guide.