A small business can easily end up with more numbers than it can use. The accounts produce reports, the website has its own dashboard, the till or booking system has another, and the CRM has several more. Few of them get looked at regularly, and fewer still lead to a decision.
A weekly scorecard fixes that by being deliberately small: eight to twelve numbers, each with an owner, a definition and a target, reviewed at the same time every week. This article explains how we choose those numbers and how to run the review so it takes half an hour rather than half a morning.
Start with what you need to know each week
Before you pick a single metric, write down the questions you want the scorecard to answer every Monday. For most owners they look something like this:
- Is enough new work coming in?
- Are we delivering it on time and to standard?
- Are customers staying and coming back?
- Is cash coming in as expected?
- Is the team stretched, or does it have spare capacity?
Each question will lead to one, two or three numbers. If a candidate KPI doesn't answer one of your questions, leave it off, however easy it is to collect.
Leading and lagging measures
A lagging measure tells you what has already happened: revenue last month, gross margin, customers lost. These matter, but by the time they move it is usually too late to change the result for that period.
A leading measure tells you what is likely to happen next: enquiries received, quotes sent, proposals awaiting a reply, hours booked for the next two weeks. You can act on these during the week.
A useful weekly scorecard has both, tilted towards leading measures. If most of your numbers come from the monthly accounts, your weekly review will spend its time looking backwards.
There is a simple test for every candidate. Ask what someone would do differently on Tuesday if the number came in below target. If the honest answer is "nothing", it belongs in a monthly report, or it shouldn't be tracked at all.
One owner, one definition, one target
An owner for every number
Each KPI needs one named person who reports it and explains it. A team or a department can't own a number in this sense, because when it slips everyone assumes someone else is looking into it. The owner doesn't have to control the number completely, but they are the one who turns up with a reason and a next step when it goes off track.
A definition everyone agrees on
Write a one-line definition for each number, including where the data comes from and when it is counted. "Enquiries" might mean form submissions, phone calls, emails or all of them. "On time" might be measured against the date you promised or the date the customer asked for. If two people use different definitions, they will end up arguing about the number instead of the business.
A target and a threshold
Give each number a target and, where it helps, a threshold below which it needs discussion. Base the first targets on your own recent history rather than on what you would like to see, and raise them once you trust the data. A simple colour code works well: green for on or above target, amber for between target and threshold, red for below the threshold.
A sample scorecard
Here is a scorecard for a hypothetical home cleaning company with around fifteen staff, serving households and small offices. All of the figures are made up for illustration. They are not benchmarks, and they shouldn't be used as targets for your business.
| KPI | Type | Owner | Definition | Target | This week |
|---|---|---|---|---|---|
| New enquiries | Leading | Office manager | Form, phone and email enquiries logged in the CRM, Monday to Sunday | 30 | 26 |
| Enquiries waiting for a quote | Leading | Office manager | Enquiries older than two working days with no quote sent | 0 | 3 |
| Quotes accepted | Leading | Owner | Quotes accepted by the customer this week | 10 | 11 |
| Booked hours, next 14 days | Leading | Scheduler | Cleaning hours confirmed in the schedule for the next two weeks | 900 | 870 |
| Available hours, next 14 days | Leading | Scheduler | Contracted staff hours minus planned leave for the next two weeks | 950 | 960 |
| Late starts | Lagging | Operations lead | Jobs started more than 30 minutes after the booked time | 3 or fewer | 2 |
| Re-clean requests | Lagging | Operations lead | Jobs a customer asked us to redo | 2 or fewer | 3 |
| Regular customers lost | Lagging | Owner | Recurring customers who ended their contract this week | 1 or fewer | 0 |
| Overdue invoices | Lagging | Bookkeeper | Value of unpaid invoices more than 30 days past due | Under 2,000 | 2,400 |
| Revenue invoiced | Lagging | Bookkeeper | Total invoiced for work completed this week, before tax | 18,000 | 17,600 |
Reading this takes a couple of minutes. Enquiries are a little below target and three are waiting for a quote, which is worth fixing now before it shows up in next month's bookings. Booked hours sit below available hours, so there is spare capacity, which fits with the enquiry shortfall. Re-cleans are over the limit, so the operations lead should bring an explanation. Overdue invoices need chasing. Everything else can be noted and left alone.
Running the weekly review
- Hold it on the same day and at the same time every week. Half an hour is enough for a scorecard of this size.
- Each owner updates their numbers before the meeting, not during it. A shared spreadsheet with one column per week does the job.
- Start with last week's action list. Check what was done before you look at the new numbers.
- Discuss only the reds and ambers. Greens get a glance.
- For each red, the owner gives a one-sentence reason and proposes one action with a date. If the reason isn't known yet, the action is to find out.
- Once a month, look at the trend over the last four or five weeks instead of single weeks. One bad week can be noise. Three in a row is a pattern.
Keeping the scorecard honest
- Review the list every quarter. Drop any KPI that hasn't led to a decision in three months, and only add a new one when you remove another.
- Don't change definitions quietly. If a definition changes, note the date so you don't end up comparing two different things.
- Watch for numbers that improve on paper. If a number gets better while the business clearly doesn't, the definition may be rewarding the wrong behaviour. Late starts could fall simply because jobs are booked with wider arrival windows, for example.
- Keep data entry light. If updating the scorecard takes an owner more than fifteen minutes a week, simplify the number or pull it straight from the system it lives in.
A KPI scorecard with definitions, owners and a review routine is one of the documents we write at Noovotic.
Examples in this article use hypothetical businesses and illustrative numbers. It is general information, not advice for your situation.
KPI Scorecard, $59 one-time
Ten to twelve numbers that show how your business is doing, clearly defined, with targets and a ready-to-use weekly sheet.






