Describe the work as it happens
Start with the trigger: an inquiry arrives, a supplier record changes, or a colleague takes over an open item. Follow the activity through the people and tools involved. Include the informal steps that are easy to leave out, such as a phone conversation or a note saved for later.
The review asks where someone has to stop and reconstruct context. That point often reveals a more useful starting scope than a long list of desired features.
Agree what needs to become clearer
A problem statement should describe an observable difficulty. For example, a coordinator cannot identify unassigned requests without opening several separate files. This is more useful than asking for a “smarter dashboard” without defining the decision it supports.
We identify who will use the proposed change, what information they need, and how they will recognize a useful result. The review should also acknowledge any disagreement about the current way of working.
Choose a proportionate next step
The next step may be a field definition, an improved record view, a prototype, or a small application. Sometimes a clearer convention in an existing tool is enough. A review should make the available options understandable rather than assume a custom build from the outset.
Dependencies can affect the decision. Existing software access, data quality, available reviewers, and the person responsible for ongoing use should be identified before the delivery scope is committed.
Leave with a usable written brief
A review brief records the problem, the intended users, the proposed boundary, and the open questions. It should separate agreed facts from assumptions that still need checking. Any example information used during the discussion should be identified as invented or redacted.
The brief becomes the basis for a further proposal if the project proceeds. It is not a commitment to a particular price, schedule, or technical platform. Those details depend on the scope and are agreed directly.