AI Readiness Checklist
A practical way to choose the first AI workflow worth testing before buying tools, hiring help, or automating the wrong work.
Built for: AI-curious owners, operators, and solo builders who have experimented a little and want the first useful project to be grounded.
Public guide
A low-friction readiness pass that turns one workflow into a first AI project brief.
This page gives you the working version: sequence, checklist, and official resources. The full kit adds prompts, a deeper worksheet, and implementation notes for your inbox.
Keep reading for the public guide, or send the kit when you want the worksheet and prompt pack.
- Pick one workflow that is ready enough to test
- Separate green-light AI work from cleanup-first or high-risk work
- Write a first AI project brief with owner, inputs, review points, and success metric
- Know when to keep a workflow manual or get qualified review before acting
Run the guide
Work through it in order.
Start with one workflow you can see
AI works best when the input, decision, and output are already visible. Start with a task that repeats often enough to matter and has enough source material to review.
- List recurring tasks across sales, service, admin, marketing, operations, and follow-up.
- Circle anything that happens every week or creates repeated friction for customers or staff.
- Write the source of truth for each task: inbox, form, CRM, spreadsheet, document, call notes, or policy.
- Keep tasks where the output is reviewable: draft, summary, checklist, quote prep, status update, report, or task list.
Use the green, yellow, red test
A simple visual signal is usually enough for the first pass. Green means ready to test. Yellow means cleanup first. Red means keep it manual or get expert review.
- Green: the task repeats, source material exists, the expected output is clear, and a human can review it before it matters.
- Yellow: the task is useful, but source material is partial, the handoff is fuzzy, or the first version needs narrow pilot rules.
- Red: the task involves unclear policy, regulated advice, payroll, hiring, legal, medical, financial, tax, insurance, or irreversible customer impact.
- Only move green and carefully scoped yellow work into a first AI project. Red work needs manual review, better source material, or a qualified advisor.
Draft the first AI project brief
The brief is the bridge between curiosity and implementation. If the brief is vague, the AI workflow will be vague too.
- Name the workflow, owner, business reason, expected input, expected output, and where the output goes next.
- Define the review point before any customer-facing message, quote, policy decision, or public claim leaves the business.
- Write one success metric for 30 days: response time, hours saved, missed follow-ups reduced, draft quality, or review speed.
- Write the stop condition: what result, error, complaint, or uncertainty means the workflow goes back to manual.
Keep the first pilot small
The first project should prove usefulness without asking the whole business to change at once. Keep it narrow enough to inspect every output.
- Start with an internal or reviewed workflow before public-facing automation.
- Use real examples, but remove sensitive customer data unless your tools and policies support that use.
- Review outputs weekly for accuracy, missing context, customer risk, and whether the workflow is saving real time.
- Expand only after the human review path, fallback path, and evidence trail are working.
Final pass
Before you call it done
- Workflow owner named
- Reliable source material identified
- Expected output written in plain language
- Green, yellow, or red readiness signal assigned
- Human review point defined before consequential use
- Fallback path documented
- 30-day success metric written
- Weekly stop-or-continue review scheduled
Useful resources
Current links to verify the details.
- NIST AI Risk Management FrameworkOfficial risk-management framework behind the govern, map, measure, and manage approach.
- SBA: AI for small businessSmall-business guidance on starting small, reviewing outputs, and avoiding sensitive data exposure.
- FTC artificial intelligence guidanceCurrent consumer-protection guidance and enforcement posture for AI claims and use.
- OECD AI PrinciplesHigh-level principles for transparency, robustness, safety, accountability, and human-centered AI.
- OpenAI privacy and data controlsUseful reference when deciding what data should or should not go into a general AI tool.
Why this guide exists
Every guide is pulled from a live client engagement. If it is in here, we have run it, measured it, and watched it hold up in the field.
Prefer to walk through it live?
Book a working call. Thirty minutes, mapped to your situation.