Five Questions to Ask About Your Management Reporting

Most finance leaders have a general sense of whether their management reporting is good. Very few have tested that sense against anything specific.

The general sense is usually built from proxies. The pack goes out every month. The board rarely complains. The numbers tie to the accounts. These are reasonable signals, and they are also compatible with a reporting process that is slow, fragile, dependent on one person, and quietly misleading about where the business makes its money.

Here are five questions that cut past the proxies. Each one has a clear good answer and a clear bad answer, and each bad answer has a specific, nameable cost. You can run the whole assessment in ten minutes, alone, without offending anyone. The only requirement is honesty about what would actually happen — not what should happen — if each question were put to the test this afternoon.

1. Can you trace any number in the pack to the transactions behind it — today, without the person who built the spreadsheet?

Pick a material figure from last month’s pack. Employee costs, ₹92,00,000, say. Now imagine the person who assembled the pack is on leave, unreachable, and you need to show a board member exactly which transactions make up that number.

A good answer: anyone on the team can follow the figure down through its layers — line to cost centre, cost centre to transactions — and the levels reconcile at every step. The path exists independently of the person who built it.

A bad answer: the path runs through one person’s spreadsheet, and without that person the number can be believed but not demonstrated. This is the drill-down test, and most packs fail it.

The cost of the bad answer is that every number in your pack is a claim rather than a fact. That distinction is invisible in a normal month and decisive in an abnormal one — a due diligence process, an audit query, a board member who asks “can you break that down?” and expects an answer before the meeting ends. Teams without traceability pay in reconstruction: re-deriving, under pressure, work they have already done once.

2. Does the pack arrive while the month’s decisions are still open, or after them?

When does your management pack actually land — not the target date, the real one? And, more pointedly: by the time it lands, which of the decisions it might have informed have already been taken?

A good answer: the pack arrives early enough in the following month that pricing calls, hiring approvals, spending decisions, and course corrections can still respond to it. Management reads it as an input, not a record.

A bad answer: the pack arrives in the third or fourth week, by which point the month it describes is a closed chapter and the current month is half spent. Everyone reads it, nods, and files it. The reasons the pack is late are usually structural rather than personal — the same reasons month-end close takes two weeks in the first place.

The cost of the bad answer compounds quietly. A report that arrives after the decisions is not a decision tool; it is history. The business still makes the decisions — it simply makes them on instinct, and the finance function’s most expensive output arrives too late to improve them. You are paying full price for reporting and receiving archival value.

3. If two people rebuilt the pack independently, would they get the same numbers?

Suppose two competent people, given the same source data and no communication with each other, each produced last month’s pack from scratch. Would the segment P&Ls match? Would the allocated costs land in the same places?

A good answer: yes, because the methodology is written down. The allocation rules, the account mappings, the treatment of one-offs and inter-company items — all documented, versioned, and applied the same way every month. The pack is the output of a method, not a person.

A bad answer: the numbers would differ, because the real methodology lives in someone’s head and in the muscle memory of a particular spreadsheet. The allocation percentages were set two years ago for reasons nobody quite remembers. Judgement calls are made fresh each month and made slightly differently each time.

The cost is twofold. First, key-person risk: the pack’s correctness depends on the continued employment and good health of one individual. Second, silent drift: when rules are unwritten, they change without anyone deciding to change them, and this month’s segment margin is no longer comparable with last year’s — but the pack gives no hint of that.

4. Do you know which products, units, or customers actually make money — or only that the company does?

The company-level P&L says the business earned ₹1.1 crore of EBITDA last quarter. Can you say, with allocated shared costs included, which product lines, business units, or major customers generated that profit — and which consumed it?

A good answer: the pack contains segment-level P&Ls where shared costs — rent, leadership, technology, central functions — are apportioned on a consistent, documented basis, so that each segment’s profitability is a fully loaded number rather than a contribution margin dressed up as one.

A bad answer: there is one P&L for the whole company, and beliefs about which parts of the business are profitable rest on gross margin intuition and anecdote. Or there is a segment view, but shared costs sit in an unallocated lump at the bottom, which means the segment numbers flatter everything.

The cost of the bad answer is misallocated effort at company scale. Businesses in this position routinely subsidise a loss-making product line for years because its revenue is visible and its fully loaded cost is not. Pricing, sales focus, investment, and discontinuation decisions all depend on knowing where the money is actually made. A company-level P&L cannot tell you. It is the aggregation of the story, not the story.

5. When a number is estimated or missing, does the pack say so — or does it read as certain?

Every pack contains soft numbers. An accrual estimated because the invoice has not arrived. A revenue figure booked before the reconciliation is complete. A provision that is a judgement dressed in digits. The question is whether the pack admits this.

A good answer: estimates are marked as estimates. Numbers that will be trued up next month say so. Where data is missing, the gap is visible rather than papered over. The reader can tell the firm figures from the provisional ones — the pack is an honest number, uncertainty included.

A bad answer: everything is printed to the last rupee with equal confidence, and the distinction between measured and guessed exists only in the preparer’s memory. The pack reads as certain because uniform precision looks like rigour.

The cost is a specific kind of failure: the restatement nobody was warned about. When a confident number moves next month, management does not conclude that an estimate was refined — they conclude that finance got it wrong. Each unexplained revision spends trust, and trust is the entire product of a reporting function. A pack that discloses its soft spots almost never surprises anyone; a pack that hides them eventually surprises everyone.

Scoring, honestly

Count your yeses — genuine yeses, the kind that would survive being tested this afternoon rather than the kind that mean “mostly” or “in principle”.

Five is rare. If you can honestly claim all five, your reporting process is better than the overwhelming majority, and your remaining work is refinement rather than repair.

Three or four is workable. The business is getting real value from its reporting, with identifiable gaps that are worth naming and closing in order of what they cost — usually starting with timeliness or traceability, because those two undermine everything else.

Fewer than three means decisions are running on trust. The pack exists, the meetings happen, but the numbers cannot be traced, arrive too late, depend on one person, hide the segment picture, or overstate their own certainty — and management is compensating with instinct and faith in the finance team. That faith may currently be justified. It is still a poor substitute for a process that can prove itself.

None of these five questions requires new software to answer, and several of the fixes are matters of discipline: write the rules down, mark the estimates, publish sooner with footnotes rather than later with polish. But the pattern across the five is hard to miss. The gaps are not analytical gaps. They are assembly and methodology gaps, and they are closable.

Datavrn is management-reporting software for finance teams, virtual CFO practices and enterprise finance functions. Available now: for one standalone Entity, take a Trial Balance through confirmed groupings and disclosures to an Excel working draft with live links or a clean finalised workbook with fixed values. Datavrn prepares the statements; it does not file or certify them. The wider management pack is in active development. Start free with your work email — no invitation needed.