Quick answer
Use the weekly review to make a few decisions from real records: what remains unresolved, what keeps recurring and what may disrupt next week. End with an owner and date for each agreed action. Reviewing every metric available is less useful than resolving the problems that keep returning.
Bring a small, consistent evidence pack
Choose one comparable trading period, such as Monday to Sunday, and gather the records you already have. Start with unresolved operational problems, exceptions from daily tasks, supplier issues, staff cover for the coming week and a brief financial view from your existing accounts or till.
A restaurant owner describing continual firefighting on Reddit illustrates why this review can matter, but a meeting will not remove the work by itself. Someone needs time to carry out the actions afterwards. Keep the evidence pack to material that can change a decision.
- Open problems with their last action and current owner.
- Repeated failures, not just total tasks completed.
- Known deliveries, events or absences that could affect next week.
- Sales and relevant costs from the same dates, with any missing information marked.
- Last week's agreed actions and their actual outcomes.
Use an agenda that reaches decisions
An example 30-minute review can spend five minutes on previous actions, ten on repeated or unresolved issues, ten on the next week's risks and five confirming ownership. The timings are a starting point; a small team may need less, while a serious issue deserves its own discussion.
Do not let a long update consume the whole slot. If an issue needs fact-finding, agree who will investigate and by when. If an action has stayed open for three reviews, ask what blocked it rather than copying the date forward.
Compare rates and context, not isolated totals
Ten complaints may mean different things in a week of 100 transactions and a week of 1,000. A quiet day can also make a task-completion figure look better without any process improvement. Add the relevant denominator and note unusual events such as roadworks, a local festival or missing staff.
Keep cash and profit views separate. British Business Bank explains that a cash forecast tracks payment timing, whereas a profit-and-loss plan measures income and costs on its accounting basis. A strong sales week does not by itself prove you can afford next week's bills.
Further reading: British Business Bank: Reforecasting in uncertain times
Turn observations into one specific change
Consider this original example: the team records three missed customer collection calls, all when the same person is serving alone at lunch. 'Improve communication' is too vague. A useful action is to trial a named backup for collection calls during that period and check the missed-call record the following week.
Limit the number of changes to what the team can actually implement. Give priority to safety and serious service risks, then choose a small operational improvement. If you change five processes at once, it becomes harder to tell which one helped.
Weekly review action Observed problem and evidence: Decision or experiment: Person responsible: Support or budget needed: Due date: What we will check next week:
Keep the review separate from blame
Ask what made the work difficult and what the team tried. An unresolved problem can reflect a missing part, inaccessible information or an impossible workload. Individual employment matters belong in the appropriate confidential process, not in a general operations discussion.
Send the short action list to those who need it and keep sensitive financial or personal details restricted. At the next review, start with what changed. A useful meeting leaves fewer ambiguous obligations, even when the business still has plenty of work to do.
Unpickle provides recurring task records and reported problems that can inform an operations review. Keep financial reporting in your accounting system and turn agreed operational actions into clear jobs. Use task records in your weekly review.