A five-page fillable workbook to move from 'we should use AI' to a specific workflow, evidence to collect, options to compare, and a decision your team can make.
Choose a real piece of work
Use this workbook with a person who performs the work and a person who can authorize a change. Start with one recurring problem, such as incomplete client documents, supplier invoice exceptions, or orders that must be re-entered across systems. Write the trigger and the finished outcome before discussing software.
The workbook is an original Clairvance planning aid with fillable fields. Save a local copy before entering information. It is also suitable for printing. You can use role titles instead of personal names and general system descriptions instead of sensitive customer records. Completing the file does not send its contents to Clairvance.
Record the process that actually happens
List the main steps, handoffs, and systems. Then describe what happens when the normal path fails: missing information, a disputed amount, a duplicate request, or an unavailable person. Include the hidden work in spreadsheets, inboxes, and conversations. These exceptions often determine whether a proposed change is practical.
State the result in operational language. 'A reviewed order ready to enter the ERP' is easier to test than 'an intelligent order process.' Assign a business owner and a person responsible for gathering representative examples. If ownership is unclear, resolving it is a valid next step.
Collect a baseline you can defend
Record volume, observation period, handling minutes, waiting time, and rework. Note where each value came from: system records, timed observations, or an estimate that still needs checking. Separate active effort from elapsed time. Three days in an inbox is not three days of labor, though the delay may still matter to service.
For a planning estimate, monthly volume multiplied by average handling minutes and divided by 60 gives monthly handling hours. If the sample is small or unusual, record that limitation. Do not turn an estimate into a precise savings claim. Count review and exception effort when comparing the proposed process with the baseline.
Compare value, feasibility, and readiness separately
For each candidate, identify the business outcome, system access, input consistency, review capacity, and evidence gaps. The workbook uses written judgments rather than a universal maturity score. A promising opportunity can require more observation or an existing software change before an AI implementation makes sense.
Compare at least one alternative to AI. That may be eliminating a redundant step, configuring a licensed feature, improving a handoff, or using a supported import. The goal is a useful shortlist and a reasoned choice. Rejecting a weak candidate is a productive outcome.
Leave with a bounded pilot brief
Choose the workflow boundary, the people participating, the systems involved, and the cases the first pilot will cover. Record the acceptance criteria, the person who makes the launch decision, the human review step, and the way to stop or revert the process. Define the evidence that would justify expanding it.
The final page includes a worked arithmetic example using constructed values and a checklist for the next discussion. The example is illustrative, not a client result. Use the completed workbook to align operations, finance, and IT, or ask Clairvance to help review the problem and implementation options.