The Reports You Used to Build by Hand Every Month

The Reports You Used to Build by Hand Every Month

Part five settled who sees what: permission attaches to the role rather than the person, goes down to the field, and exporting data is its own logged permission. That leaves the matching question — what is there to see?

And this is where most management time goes in a company running on spreadsheets: the monthly report. Nobody builds it for pleasure. They build it because it is the only available way to turn a set of scattered files into a single picture that can be looked at and judged. The problem is not the effort spent. It is that all of that effort buys an old picture.

A report that is ready after the month is not a report

When the monthly report takes two days of somebody's work, it has stopped being a report in any useful sense and become a collection exercise. The distinction is not semantic: a report is an instrument for making a decision, while a collection exercise is an administrative activity that ends in a file.

Look at the timing alone. The month ends, the gathering starts, and the numbers appear on the fifth or tenth of the following month. The decision those numbers were meant to serve — holding a payment, chasing a debtor who has slipped, adjusting a price — expired before the number pointing to it arrived. You are not managing with the report; you are reviewing what has already happened and can no longer be changed.

Worse, a late number loses its power to correct anything. When you discover on the tenth that a line item rose unusually, forty days have passed, and tracing the cause is hard because whoever carried out the transaction no longer remembers the detail and the document explaining it may sit in another file or in a chat. The monthly meeting turns into an explanation of the past instead of a decision in the present.

Where the days actually go

Watch how a monthly report is built by hand and you find that analysis takes only the last hour of it. The rest goes into five preparatory steps:

• Collecting files from their owners. A message to the accountant, another to the sales lead, a third to the warehouse, then chasing whoever has not replied. The last file always arrives late, and the whole exercise waits for it.
• Standardising formats. Dates written in more than one style, the same counterparty spelled three ways, numbers stored as text so they refuse to add up. These are precisely the faults part three of this series covered, except here they return every month instead of being fixed once.
• Reconciling numbers that disagree. The sales figure in accounting does not match the sales figure held by the sales team, so the hunt for the difference begins. It usually ends in a manual adjustment inside one cell rather than a fix to the cause.
• Rebuilding the tables and charts. Copy and paste into last month's template, adjust the ranges, repair a chart that broke because the row count changed. Work repeated in full every month with nothing accumulating.
• Checking it before sending. A final read to make sure no figure dropped out and the totals balance, because everyone knows a manual error is possible and finding it after the report went out is worse.

None of those five is intellectual work. All of it is moving, cleaning and matching — exactly the kind of work addressed by automating internal processes: repeated steps governed by fixed rules, performed by a person because nobody ever built another path for them. Handing those steps to a system does not only save hours; it removes the possibility of the manual error that the final check existed to catch.

Inside a system the report becomes a view of now

When data is recorded in one place at the moment it happens, the meaning of a report changes at the root. It is no longer a file you build but a view you open. Four practical differences follow:

• One source for the number. The sales figure is not calculated twice in two places; it is read from where it was recorded. That removes the question of which copy is correct — the question this series opened with in part one.
• The report is built once and refreshes itself. You define its shape, fields and conditions the first time only, and it shows today's data every time it is opened. The effort is spent once instead of being paid monthly.
• You can go from a number down to its detail. When a figure looks odd, you open it to see the transactions behind it rather than sending a message asking where it came from and waiting a day for an answer. That capability alone shortens most of the argument in meetings.
• The report reaches whoever needs it without a request. A branch manager opens their own view in the morning instead of waiting for somebody at head office to be free enough to build a file and send it.

The first thing you notice is not a nicer screen — it is that the monthly meeting changes subject. Instead of spending half of it agreeing that the numbers are right, you start from numbers nobody disputes and discuss what to do about them. We covered choosing the indicators you actually run operations on in more detail in the manager's dashboard.

The Origami view

When we build a system for a company that has been running on spreadsheets, we start the reporting from the decision rather than the screen. We ask which decisions the company actually makes, who makes them and when, then build the report that serves each one. The reason is that a report assembled from the list of available fields produces a crowded screen full of correct numbers that nobody acts on.

We also make sure every number on a report has a clear path down to the transactions behind it, because trust in a report is not built by explanation but by the ability to verify. That is what we apply when building custom systems across our services: a report its owner cannot check for themselves sends them back to the old file to confirm it, and at that point you have built both instead of retiring one.

Which reports deserve automating first

Do not start by automating everything you build today, because part of it does not deserve building at all. A practical order of priority:

• Whatever repeats at the same frequency. A weekly or monthly report built through identical steps every time is the best candidate, because its rules are already stable and explicit.
• Whatever is assembled from two or more sources. The more files a report draws on, the more time goes into reconciling them and the higher the chance of error. The gain here is larger than automating a report with a single source.
• Whatever more than one person asks for. If three people request the same figure at different times, you are paying to build it three times, and the three versions can disagree.
• Whatever drives a decision with a deadline. Purchasing, collections, scheduling. The value of these reports lies in arriving before the deadline rather than after it, and that is exactly what manual assembly costs you.
• Whatever is tied to a statutory obligation with a date. Returns, invoicing requirements and documents that must be produced on request carry deadlines and rules published by the competent authority, so delay in assembling them is not merely an internal efficiency matter.

Then comes the discipline question, the most useful one in this whole area: a report that has not changed a decision in the last six months does not deserve to be built at all. Take the list of reports your company issues today and write next to each the last decision it led to and when. You will usually find several are produced out of habit, sent, and never opened. Cancelling those is an immediate gain that needs no system and no budget, and it is the one part of this article you can act on this week. The same logic applies to any other technology investment, and we covered it from the measurement angle in how to measure the real return on AI.

Coming in the final part

What remains is to come down from planning to what actually happens after the move. Part seven, the last in this series, covers the first thirty days: the natural resistance, the quiet return to the old file, the data whose gaps are discovered after migration rather than before, the training delivered once and never repeated, and a thirty-day plan for getting through that stretch.

Sources

• Zakat, Tax and Customs Authority — return deadlines, e-invoicing requirements, and retaining documents and producing them on request.
• Saudi Organization for Chartered and Professional Accountants — the accounting standards endorsed in the Kingdom and the framework for preparing financial statements.
• Ministry of Commerce — provisions governing commercial books and records.
• General Authority for Statistics — the official data businesses use to benchmark their performance against sector indicators.
• Small and Medium Enterprises General Authority (Monsha'at) — programmes for raising enterprise efficiency and enabling management and digital transformation tools.