The 10 KPIs worth watching weekly (and the 40 that aren't)

Every dashboard starts the same way. Someone asks what we should track, everyone contributes a metric, and you end up with fifty tiles nobody reads. Within a month it is decoration.

The failure is not the tooling. It is that a metric was never asked to justify itself. A number earns a place on a weekly review only if it passes one test: if it moved, would you do something differently this week? If the honest answer is no, it belongs on a monthly report or nowhere.

The four that keep you solvent

Cash problems are the only ones that end a business quickly, so they get the weekly slot regardless of what else is going on:

• Cash on hand.The balance, today, across all accounts. Not a ratio, not a trend — the number.
• Projected low point. The lowest balance in the next thirteen weeks, and the week it happens. This is the single most useful number a small business can look at, because it turns a vague worry into a date.
• A/R overdue. What you are owed past terms, not total receivables. Total A/R includes invoices that are simply young; overdue is the part that needs a phone call.
• A/P due in 14 days. What has to go out shortly. Paired with the first two, this is the entire short-term liquidity picture in three glances.

The three that tell you if you are growing

Weekly revenue on its own is noise — it swings on which day an invoice was cut. What you want is the leading edge, the thing that becomes revenue later:

• Bookings or new orders this week. Work sold, whenever it gets invoiced. This moves before revenue does, which is the whole point.
• Pipeline or backlog. Committed work not yet delivered. A shrinking backlog is the earliest reliable warning a service business gets.
• The one driver of your revenue engine. Utilization if you sell capacity, net new customers if you sell subscriptions, conversion rate if you run a funnel. One number, chosen because it is the term you can actually move.

The three that tell you if it is working

• Gross margin. Weekly if you can get it, monthly if you cannot. Margin drift is slow, quiet, and the most common way a growing business becomes a struggling one.
• Labor as a share of gross profit. Your largest controllable cost measured against the thing it is supposed to produce.
• Days sales outstanding. How long customers take to pay. It moves before your cash problems do, which makes it a leading indicator rather than a postmortem.

Give every metric a trigger

Notice the last column. A metric without a threshold is trivia — you look at it, think "huh," and move on. Deciding in advance what level would cause you to act is what converts a dashboard into a control system.

Write the trigger when you add the metric, while you are calm and nothing is on fire. "Act if cash dips below $60,000." "Act if gross margin falls two months running." Then the weekly review becomes a scan for breached thresholds rather than an act of interpretation, which is the only version that survives a busy quarter.

What to do with the other forty

Keep them. They are not worthless — they are just not weekly. Site traffic, customer counts, expense detail, cohort retention: all genuinely useful, all better reviewed monthly when you have enough data for the movement to mean anything.

The point of a short weekly list is not that the other numbers do not matter. It is that attention is the scarcest thing you have, and ten numbers you actually look at beat fifty you have stopped seeing.