AI Opportunity Checklist for Small & Medium Businesses

1. Describe the work

Choose a recent example. Write down what started the task, who handled it, which systems were involved and what they produced.

Be specific about the frustrating part: searching for background information, retyping details, drafting a document or waiting for somebody to pass the work on.

2. Identify the information

List the information that a correct result depends on. Mark where it lives, who maintains it and who is allowed to use it.

For an introductory discussion, describe the information or use a fictional example. You do not need to send confidential documents or customer records.

3. Define a result you can check

Describe what should be different for the person doing the task. Record the current process so that a trial can be compared with it.

Agree how you will judge the result: completeness, accuracy, remaining corrections and whether a colleague can follow the process. Do not assume a demonstration proves a measured improvement.

4. Write down the awkward cases

Include a missing field, an unusual request, a duplicate and an unavailable system. Decide what should stop the workflow and what should be handed back to a person.

For customer-facing work, identify who approves the proposed result and how they can see the source material.

5. Name the owner and the next step

Choose who can explain the task and who would look after a working version. Consider access, instructions, colleague guidance and reporting a problem.

Bring that outline to an AI Opportunity Call. The aim is to establish whether there is a practical fit and the appropriate next step, rather than complete a technical audit in the call.

Take a copy to your team

Download the plain-text worksheet and fill it in using a suitable example from your work.

Put this into practice

Connected topics

• AI Implementation for Small & Medium Businesses: Where to Start