In this guide 6 sections
No universal list exists. Most small businesses need a handful of KPIs across sales, collections, operations and service, each with a written formula, a data source, one accountable owner and a review rhythm. Choose them by working backwards from the recurring decisions, such as chasing a payment, reordering stock or rostering a shift, instead of copying another company’s scorecard.
Let Your Business Model Decide, Not a Generic List
A KPI must feed a decision. List the recurring decisions first: whether to chase a customer for payment, whether to reorder an item or whether to change who works Saturday. Work backwards from there.
Borrowed lists rarely fit. A subscription software company and a furniture retailer share almost no recurring decisions, so a list written for the first one misleads the second about what deserves attention.
Take a packaging manufacturer. Its decisions are flexographic press scheduling, raw-material reordering and delivery timing to a few large customers, so its useful KPIs are slitter utilisation, material wastage and on-time dispatch. A butcher would ignore all three.
Group Metrics by the Area of the Business They Serve
Metrics fall into groups. Pick two or three from each group that matters instead of loading one area with a dozen figures and leaving the rest of the business unmeasured.
The list below is a menu. A business with no field staff has no use for a capacity metric, and a services firm with no physical stock has no use for a return rate. Drop every group without a decision.
- Sales: revenue against target, enquiry-to-order conversion, average order value.
- Collections: overdue balance by age bracket, days sales outstanding, invoices paid on time.
- Operations: order fulfilment time, return or error rate, on-time delivery.
- Service: first-response time, unresolved ticket count, repeat complaints.
- Capacity: staff, equipment or vehicle utilisation against available hours.
Give Every KPI a Formula, a Source, an Owner and a Cadence
A metric name alone invites argument. Two people can both say revenue and mean different things, depending on whether returns, taxes or advance payments are included, so every KPI needs a written formula. Precision means two people reach one number.
Alongside the formula, record the system it comes from, the person accountable for spotting a wrong figure and the rhythm at which someone actually reviews it. Annual review is archaeology. Most operational KPIs need a weekly or monthly look, otherwise they cannot change a decision before the moment for that decision has passed.
Keep the formulas in one place. A shared definitions sheet that lists each KPI, its exact calculation and the date the calculation last changed ends the quarterly argument about whose number is correct.
Vanity Metrics and Misleading Comparisons
Vanity metrics only flatter. Total lifetime website visits and cumulative sign-ups are common examples, because they keep climbing no matter how the business is performing right now. A number that can only rise never signals a problem, so it works as decoration, and it takes up the screen space that a genuine warning needs.
Comparisons can mislead too. This month’s revenue against last month looks alarming for a seasonal business unless it is set beside the same month a year earlier. A jump from two orders to four looks dramatic on a chart, yet it represents two extra orders, which is hardly a trend worth a meeting. Add a target, a prior-period comparison and a trend line so no single figure can be read the wrong way by a hurried reader.
Build a Small, Role-Specific Scorecard
Give each role its own short scorecard. Five to eight metrics for the sales lead and a different five to eight for operations beat one long dashboard for the whole business. Every number must be actionable.
Start smaller than feels complete. A scorecard of five metrics checked weekly and acted on beats a twenty-metric dashboard opened once and never mentioned again after the launch meeting. Adding a metric later is easy, whereas persuading people to stop watching one that no longer matters takes a long argument and a lot of goodwill.
Review each scorecard every quarter and retire any metric that has not changed a decision since the last review, because stale metrics train people to ignore the page.
KPI definition dictionary
A template with one row per KPI and columns for name, formula, source system, owner, review rhythm and the specific decision the number informs. The columns matter more than the examples.
Open a blank worksheet to printQuestions owners ask
Are the same KPIs useful for every industry?
Only a few financial basics, such as collections and cash position, travel well across businesses, whereas operational KPIs follow from what each business actually does. A field-service firm tracks technician utilisation. A retailer tracks conversion. Borrowed scorecards produce metrics nobody acts on.
How do we choose a small useful set?
List the decisions you make regularly and keep only the metrics that inform one of them, capping each role’s scorecard at five to eight. If you cannot name the decision a number would change, leave it off, however interesting it looks in a demo or a competitor’s annual report.
Who should own a KPI?
Give it to the person closest to the process that produces the number, since that person can tell when a figure looks wrong and fix the cause. A warehouse supervisor owns stock accuracy.