The AI Readiness Checklist: 10 Questions Before You Buy Anything
Most failed AI projects were decided before the first tool was purchased. These ten questions tell you whether you're ready — and what to fix first.
5 min read · Mapleforge Labs
Why readiness beats tools
Vendors will happily sell you seats, credits, and platforms today. But the difference between an AI pilot that pays for itself and one that quietly dies is almost never the tool — it's whether the workflow, the data, and the people around it were ready. Run through these ten questions honestly; every 'no' is a finding, not a failure.
The 10 questions
1. Can you name the specific task this AI should do, in one sentence, with a person who owns it? 2. Do you know how long that task takes today, and what an hour of it costs? 3. Is the input data digital, findable, and allowed to be used this way? 4. Does the task have a checkable right answer, or a human who will verify outputs? 5. Have you read your AI vendor's data terms — what is stored, for how long, and is it used for training? 6. Do your existing contracts or policies (NDAs, FERPA, HIPAA, client terms) restrict this data? 7. Who reviews the first 100 outputs before this touches a customer? 8. What does success look like in numbers, and when will you measure it? 9. What is your exit plan if the vendor raises prices or shuts down? 10. Has one skeptic — not the champion — tried to break it?
What to do with your answers
Eight or more yeses: you're ready for a scoped pilot. Four to seven: fix the data and verification questions first — they're cheap to fix now and expensive later. Fewer than four: start with training and one low-risk internal workflow, not a purchase. This checklist is the front door of our Readiness Assessment (from $1,500, with 50% credited toward a follow-on build within 60 days), which turns your answers into a written, prioritized roadmap.
Want this applied to your organization?
Start with the free discovery call — or read about the services these guides come from.