The AI Readiness Checklist is a free, printable 30-question self-assessment that tells a business owner whether their operations are ready for AI automation, and what to fix first if they are not. Six areas, ten minutes, a scored verdict. There is no question about models, prompts or tools, because readiness lives in your operations, not in your software shortlist.
We built it as the first pass of our own AI audit, the diagnostic we run before recommending anything, packaged so you can run it without us.
Who the checklist is for
Owners and operations managers of:
- Professional service firms (law, accounting, consulting)
- Service businesses (agencies, real estate, field services)
- Manufacturers
- Construction companies
- E-commerce stores
No technical background needed. If you can describe how a quote gets sent in your company, you can score every question.
It is not for a business that has already automated most of its back office and wants to compare vendors. That business needs a vendor comparison, not a readiness check.
What's inside the checklist
Thirty yes/no questions in six sections, five each. Tick a box only if it is true today, not after the cleanup you keep planning.
- Repetitive-work inventory. Where the hours actually go: the tasks repeated every week, the person spending five or more hours a week moving information between systems, the report compiled by hand, the enquiry that waits a day for a reply. More yeses here means more opportunity, not more failure.
- Process documentation. Whether your most repetitive task could be written down for a new hire, whether two people do it the same way, whether the rules behind routine decisions are consistent, whether the exceptions are rare enough to list.
- Data readiness. Whether customer information lives in one system, whether your documents follow consistent formats, whether you could export your customer history yourself, whether you know which system holds the true version of each kind of record, and whether you know which records hold personal data and who is allowed to see it.
- Tools and systems. Whether your main software is a mainstream product, whether you can get data out of it without paying someone, whether the team already works in shared digital tools, who owns the software list, who fixes what breaks.
- Team and ownership. Whether one person could own an automation project, whether the team has adopted a new tool in the last two years without a revolt, whether the people doing the repetitive work would hand it over, whether you can give a project two to three hours a week, and whether someone on the team already uses an AI assistant for work.
- Budget and payback. Whether you can put a number on an hour of your team's time, estimate the repetitive hours per week, fund a project that pays for itself within twelve months, price what one mistake in your most repetitive process costs, and pull last year's numbers to prove a change worked.
After the six sections comes the scoring page: a strip where you write each section's count, the total, and the three verdicts. Then a page that turns two of your own answers into what repetitive work costs your business in a year. Then a page of fixes, one block per section, so a weak score comes with a week's work rather than a shrug. The bonus page closes it with five tests you can run on your own numbers.
How to use it
- Set aside ten minutes. Print it, or work through it on screen.
- Go section by section. Be honest; a flattering score helps nobody.
- Count your yeses per section, then in total.
- Read the verdict for your tier, and the reading of your section strip under it.
- Turn to the fixes page for your lowest section and start there.
It works better with two people. Print a copy each and compare. Where two of you disagree on a question, you have found an undocumented process, which is a finding in itself.
What your score means
- 23 to 30: ready to build. You have the two things that matter, real repetitive work and the foundations to automate it. The question is not whether automation pays off. It's which process pays off most. Pick one. One project made to stick beats five started at once.
- 13 to 22: ready for a first project. Most businesses land here. Clear opportunity, plus a few gaps that will slow you down. Pick one narrow, well-documented workflow, prove the payback on it, and fix the gaps the checklist exposed as you go, section by section.
- 0 to 12: foundations first. Don't buy AI tools yet. Money spent on automation now mostly buys frustration. The good news: the checklist just handed you the to-do list. Customer data into one system. Top three processes written down. Then take it again. Most businesses move up a full tier in one quarter.
Why readiness comes before tools
Most AI projects fail because they start with the tool, not the problem. We see the same three failures again and again:
- A chatbot gets deployed, answers basic questions, adds no real value, and gets ignored within weeks. The subscription keeps charging.
- A process gets automated that saves 30 minutes a month while the real bottleneck, the one eating hours every day, stays untouched.
- A monthly subscription gets bought that nobody uses, because it was never built around how the team actually works.
MIT put a number on it in 2025: 95% of AI pilots show no measurable return. The checklist is the cheapest way we know to stay out of that number. Every question maps to a reason projects fail: nobody could write the process down, the data was in four places, nobody owned it, nobody had a payback figure.
What happens after the checklist
If your score says you're ready and you want a second pair of eyes, the last page points you to our free AI audit: a 20-minute discovery call, then a 90-minute audit call, then your AI Opportunity Map as a live report page within a week. Here is what a finished one looks like, on a fictional business.
If your score says you're not ready, the checklist tells you exactly what to fix first, and there is nothing to buy. Take it again in a quarter.
Format and versions
- 8-page PDF, A4, print-ready, 139 KB
- Version 1.2, September 2026
- Delivered by email once you confirm your address. When a new version ships, you get the new file, free.