Why do our numbers always arrive too late to use?

The management accounts for March arrive on the nineteenth of April. By then the founder has already made April's decisions, April is two thirds over, and the numbers describe a month nobody can do anything about.

So they get looked at for four minutes and filed. Which means the company is carrying the entire cost of producing them and receiving almost none of the benefit, and everyone involved slowly stops believing the exercise is worth doing.

The frustrating part is that a slow close is rarely caused by the accounting. It is caused by four or five inputs that arrive late, and by a process where everything waits for everything else.

What a late close actually costs

Worth quantifying, because it is usually treated as an irritation rather than as a cost.

A decision made on numbers that are three weeks stale is made on a picture that is a month out of date, since the numbers describe a month that ended before they were produced. In a company growing 5% a month that is a meaningful drift, and in one where something has just started going wrong it is the difference between catching it in one month and catching it in three.

It also compounds. A late close means a late pack, which means a late review, which means the questions raised in that review get answered during the following month's close, which makes that one late too. Companies in this loop describe it as being permanently behind, and they are.

And it has a second-order effect that matters more than the first. When numbers are always late, people stop asking for them, and the business reverts to being run on instinct with a reporting function attached to it for form's sake.

A number that arrives after the decision has been made is not information. It is a record.

The four things that actually hold it up

1. Missing bills and expense claims

The single largest cause in most small companies, by a wide margin. The books cannot close because eleven invoices have not arrived and four people have not submitted their claims.

It is a behavioural problem rather than a finance one, and the fix is a cut-off date that everybody knows, communicated before the month ends rather than during the chase. Anything that arrives after the cut-off goes into next month, and the current month carries an accrual for it.

That last part is what makes the cut-off work. Without an accrual policy, holding the line means the accounts are wrong; with one, the close stops waiting for paperwork.

2. Reconciliations that start too late

Bank, payment gateway, wallet and card reconciliations are done after month end because that is when the statement is complete. In practice, 90% of the month can be reconciled during the month, leaving only the last few days for afterwards.

A company reconciling weekly closes several days faster than one reconciling monthly, and finds errors while the transaction is still recent enough for somebody to remember it.

3. Everything happening in sequence

The most common structural cause. One person does the reconciliations, then posts payroll, then does accruals, then reviews, then builds the pack. Each step waits for the last, and any delay anywhere pushes the whole chain.

Several of those are genuinely independent. Payroll does not need the bank reconciled. The pack template does not need the final numbers to be built. Revenue recognition does not wait for expense accruals.

4. Waiting for perfect information

A close that waits for the exact figure on every line will always be late, because two or three lines are never available in time: a utility bill, a variable supplier invoice, a commission calculation that depends on someone else's data.

Estimate them, mark them as estimates, and true them up next month. The difference between an estimate and the actual on a small line is almost always smaller than the cost of the two weeks spent waiting for it.

A close that lands on the fifth working day

Illustrative, and deliberately unambitious. Most companies can reach this without changing their accounting firm or their software.

• ·Before month end: chase outstanding bills and claims against a stated cut-off. Reconcile everything up to the 25th. Build the pack template for the month so only the numbers are missing.
• ·Day one: cut off invoicing and claims. Download the final statements.
• ·Day two: finish the reconciliations, which is now only the last few days. Post payroll, including employer contributions and the gratuity accrual.
• ·Day three: accruals and prepayments, including the estimates for anything that has not arrived. Revenue recognised for the month it was earned.
• ·Day four: review the trial balance line by line against last month, and explain anything that moved more than your threshold.
• ·Day five: books closed and locked. Pack populated from closed books.
• ·Day six to eight: commentary written, pack reviewed by someone who did not prepare it, and sent.

Nothing in that sequence is sophisticated. What makes it work is that the preparation happens before the month ends rather than after it, and that the estimates are allowed.

What your accountant needs from you

A large share of a late close is caused on the client side, and it is the cheapest part to fix because it costs nothing but a habit.

• ·Bills and expense claims by a stated cut-off. This is the single biggest cause and the easiest to change.
• ·Questions answered in a batch, quickly. A list of eleven queries that takes three weeks to answer adds three weeks to the close, and most of them take a minute each.
• ·Read access to the bank, rather than statements emailed on request.
• ·Advance warning of anything unusual: a new entity, a large one-off transaction, a change in how something is billed, a customer who has stopped paying.
• ·One person on your side who owns the relationship, so queries do not wait for whoever is free.

A firm receiving clean inputs on a schedule closes faster than one chasing documents, and it is considerably more willing to commit to a date. Asking for a fifth working day close while sending bills on the twelfth is asking for something that cannot be delivered.

Fast and wrong is a different failure

It is worth saying, because the correction to a slow close occasionally overshoots.

A close that hits the fifth working day by skipping the provisions, ignoring cut-off and recognising revenue on invoice is not fast, it is inaccurate. The numbers arrive on time and they describe something that did not happen, which is worse than late numbers, because people act on them.

The test is whether the same close, examined by your auditor at the year end, would produce a large set of adjustments. If it would, the speed came from omission. The reconciliation between the management pack and the audited accounts is how you find out, and it is worth building once a year for exactly this reason.

The version that works is fast because the inputs arrive on time and the work is parallelised, not because steps were skipped.

When the software is genuinely the problem

Rarely, and it does happen, so it is worth knowing what it looks like rather than reaching for it first.

The signal is not that the software is old or unfashionable. It is that a specific step in the close cannot be done inside it and has to happen in a spreadsheet every month: revenue allocated by hand because the system cannot hold the contract terms, or inventory valued outside the system, or multi-entity consolidation done manually.

That manual step is where the time goes and where the errors come from, and no amount of process discipline removes it. A company doing four hours of spreadsheet work every month to produce a number its system should generate has a tooling problem.

What is not a tooling problem: bills arriving late, no cut-off date, reconciliations left until month end, or a pack nobody has designed. Changing software solves none of those and costs a quarter of disruption, which is why it is worth being honest about which situation you are in before anyone starts looking at demos.

The report that nobody reads

There is a second version of this problem, and it is worth separating: the numbers arrive on time and still change nothing.

That is a readability problem rather than a timing one. A pack of twelve tabs with no commentary, or a P&L in statutory format, or a dashboard whose definitions nobody agreed, produces the same outcome as a late close. It gets four minutes and gets filed.

The fixes are different. A slow close is a process problem. An unread pack is a design problem, and it is solved by writing four paragraphs explaining what moved and why, in the language the business uses, addressed to someone who was not in the room.

Two months where the date will not hold

Any close calendar meets two predictable obstacles in an Indian financial year, and planning for them is better than being surprised.

April is the first. The March close is also the year end, so it carries the audit preparation, the final accruals and every adjustment that was deferred through the year. A company that closes on the fifth working day for eleven months will not do so in April, and pretending otherwise just means the date gets broken and then stops being respected.

The second is whichever month carries your heaviest statutory load, which for most companies is the quarter ends when TDS returns fall due alongside everything else. Your accountant is working to a filing deadline that outranks your management reporting, and it should.

The answer is not to abandon the date. It is to publish a calendar with those months marked at a later date, agreed in advance, so the exception is planned rather than being a failure. A close date that is honoured ten months a year and openly relaxed in two is a working discipline. One that slips whenever things get busy is not a date at all.

What to change this month

1. 01Set a close date and tell everyone, including your accountant. Pick one that is achievable rather than aspirational, and move it earlier once it holds.
2. 02Set an expense and invoice cut-off, three working days before month end, and communicate it in advance rather than chasing afterwards.
3. 03Agree an accrual policy for the three or four items that are always late, so the cut-off can be held without making the accounts wrong.
4. 04Move reconciliations to weekly.
5. 05Build the pack template before the month ends, so the last day is populating rather than designing.
6. 06And record how long each close actually took. Two quarters of that number is what turns a good intention into a trend, and it is the thing that makes the improvement visible to everyone involved.

The free month-end close checklist on this site sets out the whole sequence day by day with the reconciliations listed, and the finance calendar covers the statutory dates that sit inside the same window. Neither requires anything of your accountant beyond agreeing to a date.

About Simplify

Simplify is a finance clarity and investment readiness practice working with founders across India, built on six years inside startups. We write about the questions founders bring before a decision, not after it.