The Handful of Metrics a Small Operation Should Track Weekly
The workbook grew, the decisions did not get better
Most small operations end up with the same reporting artifact: a monthly workbook with a dozen tabs, each added for a reason nobody remembers. A bank asked for a ratio once. A bad quarter spawned three new charts. Updating it eats half a day, the numbers arrive two or three weeks after the month ends, and the person running the business reads exactly one tab before going back to work.
The reporting habit I trust most comes from about ten years running finance and operations in higher education with a budget north of $30 million, and it is the opposite of that workbook: one page, five numbers, updated weekly in about fifteen minutes. The monthly package answers questions someone asked last year. The weekly page answers the questions you will face on Tuesday: can we afford this, and is next month being built right now.
This is general educational information, not formal accounting, tax, or legal advice. I work non-attest, alongside your CPA, not in place of them.
Why five numbers, and why weekly
Weekly beats monthly for one reason: most operating decisions in a small business are cash decisions with a two-to-six week fuse. Take the job or pass. Chase the invoice or wait. Make the hire or hold. By the time a month-end report is finished, half of those moments have already passed.
Five numbers are enough because the questions repeat. Here they are, with where each comes from and what skipping it costs.
Cash on hand. This answers the bluntest question in business: can we cover payroll and the next two weeks without doing anything clever. Log into every account, add the balances, subtract anything already committed, like checks written but not cleared. Two minutes of work. Skip it and you learn cash is tight from a bounced payment or a payroll scare, when every option left is expensive: a rushed credit line, a stalled vendor, a personal deposit to plug the gap.
Cash expected in the next 30 days. This is what is realistically arriving, invoice by invoice, not what revenue was booked. Pull it from the accounts receivable aging plus known recurring deposits, and discount it honestly: a customer who averages 45 days to pay does not belong in a 30-day window, whatever the invoice terms say. Miss it and you get the classic failure of profitable companies, a healthy income statement next to an empty account, because spending kept pace with booked revenue still 60 days from being money.
Cash going out in the next 30 days. The mirror image: what is due in the same window. Payroll dates, rent, loan payments, tax deposits, the vendor bills in accounts payable, and the recurring charges you can read off any bank statement. The recurring list barely changes, so build it once and edit it weekly. The failure mode here is ambush by the infrequent items: the annual insurance premium, the quarterly tax deposit, the software renewal that quietly went up 18 percent. On a weekly list those appear weeks in advance. Without one, they appear the morning they hit.
Runway. Cash on hand divided by your average monthly net outflow over the last three months, expressed in months. It answers the question underneath all the others: if nothing changes, how long until zero. If you run cash positive, the formal answer is indefinite, so watch the trend in net inflow instead. The cost of not knowing your runway is that month-scale decisions get made by feel. Two owners with identical bank balances should make opposite calls on a hire if one has nine months of runway and the other has nine weeks. Skip the number and you cannot tell which owner you are.
One revenue driver, specific to your business. Not the revenue line. Revenue is a lagging result, and by the time it moves, the cause is two or three months old. Pick the single upstream number that predicts your revenue: proposals sent for a services firm, jobs booked for a contractor, covers served for a restaurant. It should be the earliest thing in your pipeline that you can still influence this week. A hand tally from the CRM, the booking calendar, or the point of sale is fine; consistency matters more than precision. The failure mode is finding out in October that revenue fell, then digging through September's report for a cause that lives in July, when proposals dropped from six a week to two and nobody was counting. Lagging numbers report the crash. Leading numbers report the skid.
What stays off the weekly page
The admission test for a weekly metric is one question: if this number moved, what would we do differently on Monday? If nobody in the room can answer, the metric does not earn a weekly slot.
That test clears out the vanity tier first. Follower counts, impressions, and website traffic go, unless traffic is your revenue driver with a known conversion rate behind it. Cumulative anything goes next, because a number that can only rise cannot warn you. Gross margin to two decimal places goes too, at least weekly, since the bookkeeping behind it posts monthly and you would be reading noise. And the full profit and loss, the budget variance, and the balance sheet were never weekly artifacts. They belong in the month-end package, reviewed once, carefully. The weekly page has one job: the five dials you steer with.
The mechanics that make it stick
One page, and I mean one. Five rows, one column per week, so last week's number sits beside this week's and the direction is visible at a glance. A spreadsheet is fine. Twelve weeks in, you will have a trend history most reporting subscriptions would struggle to beat.
Same slot every week. Monday at 8:30 before the other work starts, or Friday at 4:00 as the week closes, but fixed. Scorecards die from rescheduling, not from difficulty.
Fifteen minutes, enforced. If pulling the five numbers takes an hour, do not skip the week, simplify the sources. A saved report in the accounting file and a recurring-payments list that gets edited instead of rebuilt usually solve it.
One named owner: a person, not a role. The owner pulls the numbers, writes them down, and flags anything that moved more than expected. In a two-person shop that owner is probably you, and that is fine. The point is that the page cannot silently stop existing.
Start ugly this week
Do not design the perfect scorecard first. Open a blank sheet, pull the five numbers by hand, approximate and unformatted, and date the column. Do it again next week. By week four the sheet will have caught something, an invoice aging past 45 days, a renewal creeping upward, a pipeline going quiet, and it will have paid for itself.
And if pulling five numbers takes half a day because they live in six systems that do not talk to each other, that is a finding worth acting on by itself. Mapping where the numbers live and how they should flow is where a Financial Operations Assessment starts, and the one-page scorecard is usually the first thing built from it.