What should a weekly management report tell you?

A weekly management report should tell leadership four things quickly:

1. What changed?
2. Why does it matter?
3. What decision or action is required?
4. Who owns the next step, and by when?

If the report contains forty numbers, twelve pages of activity and no clear decision, it may be accurate without being useful.

The purpose is not to describe everything the organisation did. It is to give the management team a reliable, shared view of performance, exceptions and commitments while there is still time to act.

The executive answer

A useful weekly management report normally contains:

• a short executive summary;
• a small set of agreed measures compared with target and the prior period;
• explanations for material changes;
• important customer, operational, financial, people or risk exceptions;
• decisions management must make;
• actions with one owner and a date;
• the source, definition and freshness of important figures; and
• unresolved gaps or disagreements that should not be hidden.

The report should be designed backward from recurring management decisions. Do not begin by asking every department which numbers it would like to include.

Decide what the report is for

“Keep management informed” is too broad. A report is more useful when its purpose is tied to the operating rhythm of the business.

For example:

Every Monday, the leadership team uses one report to identify material changes in sales, cash collection, order fulfilment and stock, decide which exceptions need intervention and assign the next action.

That purpose tells the preparer what belongs and what does not. Detailed campaign activity, every support ticket and every purchase request may matter elsewhere, but they do not all belong in the same weekly decision document.

Ask:

• Who will read the report?
• Which decisions recur weekly?
• Which problems become expensive when noticed a month late?
• What information must leaders share before committing resources or making promises?
• Which matters require a weekly view, and which are better reviewed monthly or quarterly?

Use one page as a discipline, not a rigid law

A one-page report forces prioritisation. Supporting detail can sit behind links or appendices for the people who need it.

Some organisations genuinely require more space. The principle is that the first view should tell a leader what is important without requiring a guided tour from the report's author.

The first page should work even when the finance lead, sales director or operations manager is unavailable to explain it.

A practical weekly management report structure

1. Executive summary

Use three to five lines:

• the most important positive change;
• the most important concern;
• a significant cause or constraint; and
• the decision or focus required this week.

An unhelpful summary says:

Sales and operations continued to perform, with several activities completed during the week.

A useful summary says:

Confirmed orders finished 8% below the weekly plan. Demand remained close to plan, but three large orders could not be confirmed because two fast-moving items were out of stock. Management needs to decide whether to accept the supplier's expedited-delivery charge before 2 p.m. Tuesday.

The second version connects the result, cause and decision. The figures are illustrative; the structure is what matters.

2. Focused performance scorecard

Choose measures that reflect the objectives and operating drivers management can influence. Include both financial results and non-financial drivers where useful.

ICAEW notes that organisations commonly need a mixture of financial and non-financial measures, with the latter helping management understand the operational levers behind financial outcomes. Read ICAEW's guidance on non-financial measures.

For each measure, show:

• current result;
• target or acceptable range;
• prior comparable period;
• variance;
• trend where relevant;
• short explanation; and
• owner.

These are fictional figures for illustration, not benchmarks. Each business should choose its own measures and definitions.

3. Important changes and explanations

A variance is the beginning of a management question, not the explanation.

For every material change, distinguish:

• fact: what the records show;
• explanation: the best supported reason currently available;
• uncertainty: what has not yet been confirmed; and
• response: what the business is doing next.

For example:

On-time delivery fell from 94% to 91%. Seven late orders were recorded; five were on routes served by the vehicle that was unavailable on Thursday. Operations is confirming whether the remaining two delays share the same cause. A temporary vehicle is booked for next week.

That is more decision-ready than “delivery performance declined due to operational challenges.”

4. Exceptions and risks

The report should bring unusual or consequential matters forward rather than forcing leaders to infer them from a dashboard.

Include matters such as:

• a major order that cannot be fulfilled;
• completed work that has not been invoiced;
• a sharp change in returns or complaints;
• overdue balances requiring escalation;
• stock below an agreed threshold;
• a supplier failure affecting customer commitments;
• an important vacancy or capacity constraint;
• a control failure or data-quality concern; and
• a commitment likely to miss its date.

State the likely effect, current owner and latest decision date. A risk without a response becomes a recurring paragraph.

5. Decisions required

Separate matters management must decide from information management should merely note.

Use a consistent format:

The recommendation should show its assumptions. If finance has not confirmed the margin, mark that gap instead of presenting the choice as settled.

6. Action log

Every decision or exception should produce a visible next step.

Each action needs:

• one accountable owner;
• a specific outcome;
• a due date;
• status; and
• closure evidence where appropriate.

“Sales and operations to follow up” creates shared concern, not shared accountability. Choose one owner even when several people contribute.

Carry incomplete actions into the next report. Do not erase an overdue commitment by rewriting it as a new one.

Define every important measure

Two branches can report “sales” correctly while one uses invoices and the other uses confirmed orders. The total is not comparable even when the arithmetic is perfect.

Maintain a small measure register containing:

The IFRS Foundation's management-commentary material is intended for external general-purpose reporting rather than a weekly internal report, but two disciplines transfer well: information should be useful to decisions, and company-specific measures need clear definitions and explanations. See the IFRS management-commentary key terms.

Reconcile information before explaining it

When sales, finance and operations report different versions of the same event, do not ask an AI tool to choose the most plausible number.

Identify:

• which system records the authoritative event;
• which date each system uses;
• whether cancelled or returned items are included;
• how customer, product and branch identifiers match;
• when each source was updated; and
• which reconciliation difference remains unresolved.

For example, the CRM may count an opportunity when a customer accepts a quotation. The ERP may count an order after stock is allocated. Finance may count revenue after invoicing according to the organisation's accounting process. These numbers answer different questions.

A combined report should preserve those distinctions rather than averaging them into one comforting figure.

Worked example: the Monday order brief

Consider an illustrative distributor with separate sales, inventory, delivery and accounting systems.

The managing director wants to know:

Which customer commitments this week are most likely to fail, and what must we decide today?

The report uses:

• confirmed orders from the order system;
• available and expected stock from inventory;
• planned delivery dates from dispatch; and
• payment or credit holds from finance.

The result does not need to reproduce every record. It can show:

1. 62 orders are due this week.
2. 49 have stock allocated and no recorded hold.
3. 8 need stock expected before their planned dispatch date.
4. 3 are blocked by a finance hold.
5. 2 have conflicting product codes and need manual checking.

The management section then asks for two decisions: whether to expedite one shipment and whether the finance director will review the three held orders before customer updates are sent.

The report should link to the source records. A person confirms the explanation and decisions. AI may help prepare a draft narrative once the records are reconciled, but it should not invent the reason a stock balance changed or quietly resolve a disagreement between systems.

Dashboard, report or alert?

These formats serve different jobs.

A useful design often combines them: an alert for a time-sensitive exception, a dashboard for investigation and a concise report for the weekly management conversation.

Where AI can help—and where it should not

AI can assist with:

• preparing a draft summary from approved measures;
• identifying material changes using agreed thresholds;
• grouping recurring explanations or issues;
• turning meeting decisions into an action list;
• rewriting technical commentary in plain language; and
• delivering the approved brief through email or another familiar channel.

It should not be trusted to:

• invent a cause when the data only shows correlation;
• change a measure definition silently;
• choose between conflicting source records without an agreed rule;
• hide missing information behind fluent prose;
• make an unauthorised financial, customer or employment decision; or
• replace the accountable owner who approves the report.

If AI prepares the narrative, require source links, timestamps and visible gaps. Keep the figures and reconciliations in deterministic reporting logic where practical.

Improve the report over four weeks

Week 1: agree on the questions

List the decisions and exceptions management actually discusses. Remove measures nobody uses.

Week 2: define and reconcile

Document definitions, sources, owners and cut-offs. Investigate disagreements before automating the report.

Week 3: run the one-page version

Prepare it manually if necessary. Observe which lines management challenges, ignores or turns into action.

Week 4: automate the stable preparation

Only after the report proves useful should the business automate collection, comparison, draft commentary or delivery. Preserve review for explanations and decisions.

Measure the report itself:

• preparation and checking time;
• late or missing source submissions;
• corrections after distribution;
• decisions reached during the review;
• overdue actions; and
• measures removed because nobody used them.

Make the next decision easier

A management report earns its place when it changes attention or action. It should show performance without disguising uncertainty, explain important movements without inventing certainty and leave every decision with a responsible person.

Start with one recurring management question. Agree on the measures and sources. Produce a short report manually. Improve it through real management use. Then automate the stable preparation—not the judgement that makes the report worth reading.

For a simple individual workflow, see how to create a weekly report from a spreadsheet with Gemini. For multi-system reporting, review how AI integration changes management information.

Research and helpful links

• ICAEW guidance on financial and non-financial measures
• IFRS Foundation management-commentary supporting material
• UK Cabinet Office collection on improving management information
• UK Government Analysis Function guidance on management information

Research checked on 20 September 2026. This is a guide to internal management reporting, not accounting, audit, tax or external financial-reporting advice.