The eight columns you should fill in every Friday

Open a spreadsheet. Name the first column “Week ending” and fill in the next twelve Fridays. You now have the frame for the only management report a small business actually needs, and the rest of this article is what goes in the other columns and why.

We are starting with the mechanics on purpose, because almost every article on this subject stops at a list of concepts and leaves you to invent the file yourself, which is the point at which everyone gives up.

Why weekly, and why so few numbers

Monthly reporting has two problems. It is slow, so a bad trend runs for five or six weeks before anybody sees it. And it is long, so nobody reads it properly, including the person who asked for it.

Jack Stack built an entire management system on the opposite principle at Springfield Remanufacturing, the near-bankrupt International Harvester division his employees bought out, and wrote it up in The Great Game of Business (Doubleday, 1992). The two ideas worth stealing from it are the critical number, meaning the one figure that decides whether this year works, and the weekly huddle, where the numbers get said out loud and forecast forward rather than filed. Stack’s version is elaborate. The small version is a sheet and twenty minutes on a Friday.

The seven columns

Name them exactly. Ambiguity is where reporting arguments come from.

Cash available. The bank balance minus what is already committed: wages, VAT, supplier invoices due inside 30 days. One number, and it should be the first thing on the page.
New qualified leads. Not enquiries. The ones that cleared your bar. Write the bar down in a sentence a new hire could apply without asking you.
Quotes sent, and their total value. Your pipeline as a figure instead of a feeling.
Deals closed, count and value. Against a weekly target, so a bad week is visible in seven days rather than thirty.
Close rate. Closed divided by qualified leads, as a rolling four-week figure. Offer problems and sales problems surface here before anywhere else.
Cost per acquired customer. All acquisition spend divided by customers won, also rolling four weeks, because one week of it is noise.
Capacity committed. How full your team or diary is for the next three weeks, as a percentage. Selling beyond what you can deliver costs more in lost retention than the extra sale brings in.

Most businesses need five of those plus one or two of their own: occupancy, average order value, overdue invoices, scrap rate. Add an eighth column called “What changed” and write one sentence in it every week. In six months that column will be the most useful thing in the file, and it is the one nobody ever creates.

The rules that make it survive

One source of truth per number. If leads live in three places, pick one and make the others follow it.
Same day, same person, same twenty minutes. Consistency beats precision. A roughly right figure entered every Friday is worth more than an exact figure entered whenever someone remembers.
Read direction, not level. Four weeks of trend tells you something. One week tells you almost nothing.
Do not buy a dashboard tool yet. Buy one when you know which numbers you care about and the spreadsheet has become genuinely annoying. That usually takes a year.

Now, the reason cash goes first

The JPMorgan Chase Institute built a dataset from 470 million transactions across 597,000 American small businesses and published it as “Cash is King: Flows, Balances, and Buffer Days” in September 2016. It asked how long each business could survive on the cash in its account if the money stopped coming in.

The median answer was 27 days, on a median daily balance of 12,100 dollars, with wide variation between industries and within them.

Twenty-seven days is less than one payment cycle. It means the median small business is one slow-paying client, one broken machine or one quiet fortnight from a problem that no amount of revenue growth will solve in time, and it means that a figure most owners check by glancing at their banking app on the way to something else is the figure their survival actually runs on. That is why it sits in column two and not column six.

It is also why revenue is a poor headline number. Revenue with no view of margin, cash or capacity has walked a great many busy, growing businesses into insolvency, and they were all able to quote last month’s turnover from memory on the way in.

Decide the responses before you need them

A number you only observe is decoration. Write the trigger next to each one.

Qualified leads down two weeks running: look at the channel before you look at the sales team. Close rate falling while lead volume rises: your lead quality changed and the fix is upstream, not in the pitch. Cost per acquired customer rising while close rate holds: the auction got more expensive and the decision is whether your margin can absorb it, which you can only answer if you have done the repeat-customer arithmetic. Capacity above about 90 per cent: stop selling and fix delivery.

Writing those down in advance is not bureaucracy. It is the only way to avoid making the decision in week three of a bad month, which is when owners cut the marketing budget to zero and discover two months later what pipeline lag means.

What to leave off

Impressions. Reach. Followers. Sessions in isolation. Time on page. Not worthless, but not owner-level, and putting them on a weekly page trains you to watch things you cannot act on. The test is simple: if a figure cannot change a decision you would make this month, it does not belong in the file. Most of what agencies report monthly fails that test, including some of what we have reported ourselves over the years, which is part of why our optimisation work now starts by asking what the owner already looks at.

So: eight columns, twelve rows, twenty minutes on Friday. Put this week in now, even if three of the numbers are estimates and one is a guess you will correct next week. The file is useless today and it is useless next Friday. Around week five it starts telling you things your accountant will not know until March.