AI readiness framework

Start with the workflow, not the model.

Most AI programmes fail before the technology becomes the problem. The workflow is vague, the baseline is missing, the risk owner is unclear, or the team never changes how it works. This framework helps choose a first use case worth piloting.

Get the free worksheet → Read the framework

Educational resource · No paid service or booking

Start here

Use this resource when…

Tool overload

The team is testing AI without a business case

Move from scattered demos to one workflow with a baseline, an owner, known risks, and a measurable improvement target.

Data anxiety

You do not know what can safely leave the business

Classify the data, permissions, review requirements, and failure impact before selecting a model or automation platform.

Adoption gap

The pilot works, but nobody changes behaviour

AI creates value only when it fits the real workflow, earns trust, and has an owner for exceptions, feedback, and continuous improvement.

The framework

The five-part AI readiness test

A useful first pilot is narrow enough to measure, safe enough to reverse, and important enough that the team will actually change behaviour if it works.

Choose a workflow, not a technology

Start with repetitive decisions, document handling, customer conversations, forecasting, or internal search where delay and inconsistency already cost the team.

Output: One bounded workflow

Measure the current baseline

Capture time, volume, error rate, rework, escalation, and customer impact before automation. Without a baseline, every demo looks impressive.

Output: A before-state scorecard

Map risk and human review

Decide what data is sensitive, which outputs require approval, what an unacceptable mistake looks like, and how users can correct the system.

Output: A risk and review matrix

Run a narrow pilot

Use a representative sample, keep the workflow reversible, compare against the baseline, and record exception types instead of hiding them.

Output: Evidence from a controlled pilot

Design adoption and ownership

Name who monitors quality, handles exceptions, updates knowledge, trains users, and decides when the system should be paused or expanded.

Output: An operating model for scale

Quick self-check

Is this use case ready for a pilot?

If you cannot answer these questions, the next step is workflow discovery—not another AI vendor demo.

The AI use case begins with a real workflow and named owner.
The current time, volume, cost, and error baseline are documented.
Sensitive data and access rules are classified before tools are selected.
The team knows which outputs require human review.
The pilot has a representative sample and clear comparison method.
Exceptions and incorrect outputs are captured, not silently corrected.
Users understand how the system changes their work and where it can fail.
Expansion depends on measured value, acceptable risk, and operational ownership.

Free printable worksheet

Download the AI Readiness Scorecard

A printable worksheet for scoring workflow value, baseline clarity, data risk, human review, pilot quality, and adoption ownership.

Useful without a sales call
Immediate PDF download
Occasional educational updates only