The Monthly Reporting Pack: What Gets Read First, and What to Cut
Most packs at this size are assembled in the order the ledger produces and read by nobody past page two. This is the pack we issue, the order we issue it in, and the pages we removed to get there.
A monthly pack has one reader who has to act on it and several who only have to be reassured by it. Nearly every pack handed to us at the start of an engagement was built for the second group.
The statements are correct. The formatting is careful. The first page is a balance sheet, because that is the order the ledger produces it in and the order whoever built the pack was taught. The owner opens the file on a phone somewhere between two meetings, scrolls, finds nothing that tells her what to do about the next fortnight, and puts it away. A few months later somebody asks why the reporting is not being used, and the honest answer is that it was never arranged to be used. It was arranged to be complete.
So here is the position, before the walkthrough that argues it. Order the pack by the question each page answers, not by the hierarchy of the financial statements. A written page leads. Cash is second. The statements go at the back, where they work as evidence rather than as reading. And then cut roughly a third of what is currently in the file, because at this size the useful work on a reporting pack is almost entirely subtraction.
The company below is a composite of the ones we do this for: about fifty-five people, two branches, project-based revenue, an operating line with a covenant on it, and one outside shareholder from a minority sale a few years back who sits on a board that meets quarterly. The sequence is the one we actually run on the engagements where we issue the pack ourselves.
Who opens it, and for how long
Start here, because the order of the pack is a consequence of this table and not a matter of taste.
| Reader | When the file gets opened | What gets read | What has never once been opened |
|---|---|---|---|
| Owner or CEO | Within an hour of it landing, usually on a phone | The written page, then the cash page, then whichever line on page 3 the written page told her to look at | Account-level detail, the ratio page, the statements at the back |
| Branch or line manager | The same day, if the pack is shared past the owner | Their own column, and the allocation line under it | The balance sheet, in any month, ever |
| Outside shareholder or director | The evening before the meeting, in one sitting | The written page, the trend page, the cash page, and the prior meeting’s follow-ups | The ageing, unless a receivable is already a live issue |
| The bank’s analyst | On receipt, quarterly rather than monthly | The covenant calculation and the balance sheet, which get spread into their own model | The commentary, the operating measures, the branch split |
| The external accountant | Not during the year | Nothing in the pack. They want the close binder | All of it, in a sense |
Nobody in that table reads the pack front to back. Nobody ever has, including the person who built it, who reads it in build order, which is a different order again and the one that quietly becomes the page order if nobody intervenes.
Two consequences follow. The first is that page order is a real decision with a real cost, because attention is spent in the first minute and whatever is on page one gets it. The second is that a pack serving five readers this differently should not be five documents. It should be one document with an order that suits the reader who has to act, and appendices deep enough for the readers who have to verify. The bank is the partial exception, and the covenant work that reader needs is its own build, covered in the lender reporting pack piece. The rule that matters here is that both come off one set of figures. Two packs built independently will disagree in the second quarter, and the credibility never fully comes back.
The order, and why the statements go at the back
| Page | What sits on it | The question it answers |
|---|---|---|
| 1 | One written page: what happened, what it means, the decisions on the table, and what we got wrong last month | What do I need to do about this month? |
| 2 | Cash: the month’s actual movement by receipts and payments, closing bank, headroom against the facility, then the forward thirteen weeks with last month’s forecast scored against what really landed | Are we fine, and if not, which week does it bite? |
| 3 | Result by branch and by line: current month, year to date, prior year same period, contribution shown both before and after allocated overhead | Where did the result come from, and which part of the business made it? |
| 4 | Balance sheet with the working capital block pulled out, plus the debt and facility schedule | What is tied up, and what do we owe? |
| 5 | Operating measures on a trend of at least thirteen months, with definitions printed on the page | Is the shape of the business changing? |
| Appendix A | The statements, with comparatives, in conventional presentation | Can somebody check this? |
| Appendix B | AR and AP ageing by bucket, with concentrations named | Who owes us, and who are we behind with? |
| Appendix C | Definitions, allocation bases, version and effective date | What do these words mean here? |
Three reasons the written page leads, in descending order of how much they matter.
The reader’s decision is forward and every statement is backward. An owner deciding whether to take on a subcontract crew next month is not served by a page describing a period that finished a week ago. She is served by a sentence that says the margin on the branch that would run that crew has fallen for two consecutive months and here is the reason. That sentence is the product. The statements are the substantiation for it.
The pack is not competing with other packs. It is competing with an inbox. A document that requires the reader to construct her own summary from five statements will lose to a document that does not, every time, and the document that wins is usually a text message from a branch manager with a worse number in it.
Writing the first page is a control on the preparer. This is the reason nobody expects and the one I would keep if I could only keep one. A person who has to write four honest paragraphs about the month cannot get there without having understood the month. It surfaces the accrual that was posted because it is always posted, the variance that was labelled timing three months running, the clearing account nobody has looked at. We have caught more real errors in the writing of page one than in the review of page four, and it is not close.
The usual objection to statements at the back is that a reader might stop checking the numbers. In practice a reader who wants to check walks to the back and checks, and a reader who does not want to check was never going to be converted by finding a balance sheet on page one. What the front position actually buys is the appearance of rigour, which is not the same product.
Lock day: the pack starts when the ledger stops
Nothing in the pack gets built from a moving trial balance. The month-end close ends with the period locked and the lock date recorded, and the pack is built from a trial balance exported after the lock and saved into the month’s binder under its own name. The pack cites that export. When someone six weeks later gets a different number out of the ledger, and someone will, the question becomes what changed after the lock, which is answerable in a minute, instead of whose spreadsheet is right, which is answerable never.
Two things happen at the lock and both are ten minutes of work.
The reviewer exports the trial balance, the general ledger detail for the month, and the sub-ledger ageings, and drops all four into the binder. Then the reporting workbook is pointed at that export rather than at a live connection. A pack wired live to the ledger is a pack that silently restates itself, which means the PDF the board has and the workbook you are looking at during the meeting no longer agree.
Lock day plus one, morning: the variance read
The reviewer reads the result by branch and line against the prior month and against the same month last year, and writes down every movement past the threshold agreed at the start of the year. Each one gets a cause traced to a transaction or a decision, in a sentence, with the transaction identified in the working file.
In our composite month, three things moved. The eastern branch lost four points of gross margin. The cause was a project that absorbed subcontract hours nobody had priced, and the sentence names the project. Overhead was up on the prior month because the annual insurance renewal landed in one month rather than being spread, which is a presentation choice we made and which is going to raise the same question every year unless the written page says so once. And the largest customer paid outside its normal behaviour for the first time in about a year, which does not touch the result at all and belongs on the cash page rather than the variance list.
Anything without a cause yet is written down as an open question with a name and a date against it, and it stays on the pack until it closes. A reporting pack that never carries an open question is not being written honestly. Nobody understands every movement by the second business day after a lock.
Lock day plus one, afternoon: page one
The written page has a fixed shape and a hard length limit of one page. If it runs to two, something on it belongs on page 3 instead.
- What happened. Three or four sentences on the month, in the order of what mattered rather than in statement order. Margin, then cash, then anything structural.
- What it means. The part that is a judgement rather than a fact, and labelled as one. If the margin loss on the eastern branch is a pricing problem rather than a one-off project, say so and say what it is based on.
- The decisions on the table. Numbered, each with an owner, each with a date it stops being useful. If there are no decisions this month, write that there are none. Some months there are none, and inventing one to fill the section is how the section stops being read.
- What we got wrong last month. Any variance explained as timing that did not reverse, any forecast week that missed badly, any correction to a prior pack. This section costs the preparer something to write, which is exactly why the reader starts trusting the rest of the page.
Most packs carry the commentary, and most of them carry it at page four, behind the statements. I have built packs that way. It is the wrong place, and the reason is not subtle: the page that tells the reader where to look cannot sit behind the pages she is supposed to look at.
Issue
Same day, always, and the day itself matters less than that it never moves. The pack goes out as a dated PDF with the period in the filename and the same filename structure every month, sent by the person who built it, with a covering note naming the period and listing the decisions from section 3 in the body of the email. That last part is not a flourish. Some readers will act off the covering note and never open the file, and that is a success rather than a failure, because the decisions reached them.
The live workbook goes into the shared drive at the same time, in the same folder as the trial balance export it was built from. Nothing in the pack is a black box the client cannot open after the engagement ends.
What happens on the other end
The owner reads the covering note within the hour and the pack that evening. She replies to two of the three decisions and ignores the third, which is information: a decision nobody wants to make is usually a decision that was framed wrong.
The outside shareholder opens the file the night before the quarterly meeting. He reads the written page, the trend page, and last quarter’s follow-ups. He does not open the appendices, and the one time he does, it will be because something on the written page did not sound right, which is precisely what the appendices are for.
The branch manager reads his own column and disputes the overhead allocation. He does this most months. The allocation basis is in Appendix C with a version and an effective date, which converts a recurring argument into a single conversation about whether the basis should change, held once, with the comparatives restated if it does.
What to cut
This is the part that improves a pack fastest, and it is the part nobody does, because removing a page feels like producing less. Here is what comes out.
| Cut | Why it is in there | What replaces it |
|---|---|---|
| The monthly cash flow statement, prepared by the indirect method | It is in the annual statements, so it feels obligatory monthly | The cash page: actual receipts and payments for the month, closing bank, headroom, and the rolling thirteen weeks |
| Account-level P&L detail, the several hundred row export | Somebody asked for detail once, years ago | A line in the covering note saying the detail is in the binder and will be sent on request |
| Budget variance columns, where the budget was set in one week last autumn and never revisited | The budget exists, so it gets a column | Either a quarterly re-forecast that makes the column mean something, or no column |
| The ratio page: current ratio, quick ratio, return on assets, inventory turns computed monthly | It looks like analysis | The two or three measures somebody is contractually or operationally measured on, on page 5 |
| A chart repeating a table that is already in the pack | Charts look like effort | Whichever of the two the reader actually uses. If the point is a trend, keep the chart and cut the table |
| Departmental P&Ls for departments nobody manages as a unit | The chart of accounts has the dimension, so the report has the page | Reporting that follows how the business is actually run, which usually means fewer splits than the ledger supports |
| An executive summary that summarises the pack | Packs are supposed to have one | The written page, which says what to do rather than what is on the following pages |
The first row is the one that gets an argument, so let me defend it properly. The indirect cash flow statement is a reconciliation of net income to operating cash, built for a reader who needs a statutory presentation and knows how to read one. An owner asking where the cash went cannot use it, and I have watched capable people stare at one and quietly conclude that the finance function is speaking a language at them. Cut it monthly. Keep it at year end, keep it wherever a credit agreement requires the pack to be prepared consistently with the annual statements, and put the month’s actual money movement on page 2 in the words people use for money.
The third row will be unpopular for a different reason. A budget nobody re-forecasts is not a plan, it is a record of what somebody believed in the autumn, and reporting variance against it every month for a year trains the reader to skip variance columns entirely. That habit then costs you the month when the variance genuinely matters. Either re-forecast quarterly or take the column out. Leaving a dead budget in the pack is worse than not budgeting, because it consumes the reader’s trust in the one place you most need it.
The removal test
How much of a pack is never opened is a question people answer with a confident-sounding fraction. Nobody knows. A PDF in an inbox is not instrumented, and asking a reader which pages they read gets you the pages they feel they ought to read.
So test it instead, once, at the start of an engagement. Agree with the owner in advance that for one month a named page is coming out. Not a page anyone is contractually owed, and not the appendices, because the appendices exist to be available rather than to be read. Then issue the pack without it and see who asks. In our composite company, the ratio page and two of the departmental splits came out this way and nobody has mentioned any of them since. The account-level detail came out and one person asked, in month two, for one account, which is now a line in the covering note rather than forty pages in a file.
The version of this that fails is doing it silently and letting the reader discover the gap. Agree it first. The exercise exists to find out what the pack is for, and catching a reader out would poison the answer as well as the relationship.
The review meeting, and the only thing it is for
The meeting happens the day after issue, in a standing slot, and its agenda is not the pack. Reading a pack aloud to people who received it the previous day is the most expensive way to distribute a document ever devised.
The agenda is four items: the decisions from page one, the open questions from the variance read and who is closing them, the cash trough in the forecast window and what has been done about it since it appeared, and the follow-ups from last month with their status. Every item leaves with an owner and a date, and reappears on the next pack until it is closed. A follow-up log that carries three items forward for four months is telling you something true about the company, and it is better to have that in writing than in everyone’s vague sense that things are not moving.
What we do not do
We do not build a second version of the pack for a different audience. Lenders, directors and outside shareholders get the same figures with a different cover, because a company that maintains two sets of management reporting will eventually have to explain why they differ, and there is no good version of that conversation.
We do not add a page because a month went well or remove one because it went badly. The order and the contents are set outside the month and changed deliberately, with a note and restated comparatives, or they are not changed.
We do not put a number on the written page that is not on a later page, and we do not put a judgement on the written page without labelling it as one.
And we do not issue a pack from an unlocked period. If the close is not done, the pack is late, and a pack that is late and right costs a great deal less than a pack that was on time and gets reissued.