Why the Friday spreadsheet is stale before you open it
Most companies know the Friday spreadsheet. On Friday afternoon someone opens the e-shop, then the invoicing tool, then the CRM, copies figures into one workbook, adds a few formulas, and emails the result. It usually takes an hour or two, in some companies half a day. And by Monday morning the numbers are already out of date.
Time is not the only problem. Every copy-paste is a chance for a mistake: a shifted row, a forgotten filter, a formula someone overwrote. The file then gets forwarded, and two weeks later it exists in three versions — “report_final”, “report_final_v2”, and “report_final_really”. When a manager and a salesperson look at two different revenue totals, nobody knows which one is right.
The outcome is worse than the error itself: people stop trusting the numbers. In the meeting, instead of “what do we do about this?”, the discussion becomes where the figure came from and whether it is current. A report nobody trusts is just one more document someone had to prepare.
Start with questions, not charts
When people start on reporting, most jump straight to appearance: which charts, which colours, which tool. That is the wrong end. First ask which decisions in your company come up regularly, and what you need to know to make them well.
For example: order more stock or not? Contact customers who have not bought in a while? Are unpaid invoices under control, or is it time to start calling? Can we afford another hire? Behind each of these questions stand one or two concrete numbers. If a decision has no number attached to it, it does not deserve a line in the report either.
Write the questions on a single page and note next to each who handles it and how often. The content of the report then comes from that list — not the other way round. You end up with a document people use, not one that just gets filed away.
Five to eight numbers, not fifty
The temptation to put everything measurable into the report is strong — after all, we already have the data. But a twenty-page overview does not get read. People flip through it, find nothing that stops them, and close it.
A good limit is a handful of numbers that fit on one screen or in a short email: weekly revenue, number of new orders or enquiries, overdue invoices, stock levels for the main products. Another company will pick different numbers, but the principle is the same. Each one benefits from a single comparison — against last week or against plan — because a number on its own, without context, says nothing.
If you later find something is missing, you add it. It is far easier to add a number someone keeps asking about than to remove twenty nobody looks at.
Let the data come from what you already have
Most of the data you need already exists in the company. Orders sit in the e-shop, issued invoices in the invoicing tool, contacts and enquiries in the CRM, payments and costs in the accounting system. None of it needs to be introduced from scratch. You just need to stop being the human link between them.
Most of these systems can hand over their data themselves — through an export or a programming interface (API). An automation pulls it at regular intervals, calculates it by agreed rules, and stores it in one place. No human hand is needed, and the typos disappear with it.
One thing matters: every number needs a single source of truth. Revenue comes from invoicing, not from the e-shop and the salespeople's spreadsheet at once. Stock comes from the warehouse system. If two figures differ, you decide once which one counts and write the definition down — for instance, whether revenue includes VAT. Arguments of the “mine comes out differently” kind disappear.
A Monday email, or a live dashboard
Finished numbers have to be delivered, and there are two main routes. The first is a scheduled email: every Monday at seven a.m. the owner or team gets a short overview. Nobody has to open anything or go looking; the number comes to you. It suits things reviewed weekly or monthly, and people who do not want another login.
The second is a live dashboard — a page where numbers are always current and you can click through to detail. It fits where the situation changes during the day: orders waiting at dispatch, enquiries waiting for an answer, the balance on the account. Its drawback is that you have to open it. If you forget, it was pointless.
In practice the two are often combined: a Monday email with the key numbers and a link to the dashboard for anyone who wants to dig deeper. Add a simple alert for when something goes off track — say, an overdue invoice above an agreed amount. A good report should not only inform you but warn you in time.
Where it does not pay off
Automated reporting is not for everyone. If there is little data — a few dozen orders a month and a single system — a spreadsheet you fill in over ten minutes is more sensible than any automation. You would never earn back the time spent setting it up.
The second case is sneakier: numbers nobody acts on. If a report is read for two months without a single reaction, automation will not make it more useful. First work out what you actually need it for — and if the answer is nothing, drop it. Likewise, you cannot automate poorly kept records: if invoices are entered irregularly or customers are duplicated in the CRM, the report will just show the wrong number faster. Order in the source first, automation second.
But if your data sits in two or three systems and an hour or more goes into it every week, it is worth a closer look. At Lucorp we help with this kind of overview — from choosing the numbers to delivering them.