A free AI readiness assessment: five questions to answer first
You don't need a scorecard to know if your team is ready for AI. You need honest answers to five questions. Here they are.
Why check first
Most AI money is spent before anyone asks whether the work is ready for it. Then the project stalls, and the team decides AI doesn't work for them.
Gartner predicts that through 2026, organizations will abandon 60% of AI projects that aren't supported by AI-ready data (Gartner). Data is one reason. In my experience the bigger one is that nobody wrote the work down.
These are the questions I ask on a first call. You can answer them yourself.
1. Is the work written down?
Could someone who has never done this job follow your instructions and get an acceptable result? If the answer is "it's in Sarah's head", AI can't do it yet.
Ready looks like: a clear trigger, the inputs, the steps in order, the judgement calls, and two or three examples of finished work you were happy with.
2. Can AI reach the information?
The AI needs to read what the person reads: the sheet, the database, the inbox, the call notes. If two sources disagree, someone has to say which one wins.
Ready looks like: the inputs live somewhere the AI can be connected to, and you know which one is right.
3. Who will run it?
A tool only one person understands stops the day they're out. Someone on the team has to own it, and more than one person has to be able to change it.
Ready looks like: a named owner, and a team willing to learn.
4. Do you already have the tools?
Most teams already pay for an AI tool. Buying another rarely fixes an undefined process.
Ready looks like: you know what you pay for today, and you're willing to start there.
5. What happens when it's wrong?
It will be wrong sometimes. One odd email to a customer is enough to get a tool switched off for good.
Ready looks like: a person checks the output before anything reaches a customer, at least at first. In one pipeline I built, the AI drafts every email and sends none. The founder reads and sends each one. More in Drafts, never sends.
What to do with your answers
If you said yes to all five, pick one piece of work that happens every week and start. If you said no to the first question, start there. It's the most useful thing you can do, and the page is worth having even if you never automate anything.
For the long version, with twelve questions and a scoring guide, read The AI Readiness Assessment: 12 Questions to Ask Before You Spend a Dollar. If you'd like a second opinion, tell me about the work and I'll give you my honest read on the first call.
Common questions
What is an AI readiness assessment?
A check of whether your team can actually run AI on a piece of work before you spend on it. It covers the process, the data, the people, the tools and how you'll catch mistakes.
Do we need clean data before starting?
Not all of it. You need the data for the one process you start with to be reachable, and to know which source wins when two disagree.
What's the most common gap?
The work was never written down. Fixing that is hours per process, not weeks.
Should we buy a new AI tool first?
Usually not. Start with what you already pay for. A clear process runs on any of the major tools.