Key performance indicators are the handful of numbers you would act on. Everything else is context. Most dashboards get that backwards, which is why so many are opened once and never again.
The test key performance indicators have to pass
One question decides whether something belongs. If this number moved, would we do anything differently?
Where the answer is no, it is context rather than a key indicator. Impressions usually fail that test. Followers usually fail it. Meanwhile cost per qualified enquiry almost always passes, because a change in it forces a decision about budget.
How many key performance indicators to keep
Three to five. Beyond that nobody reads them, and a report nobody reads is not a report.
The temptation is to include everything in case somebody asks. Resist it, and keep the rest in an appendix. A short front page with the numbers that matter, followed by detail for whoever wants it, gets acted on far more often than a wall of charts.
Choosing them for your own business
Start from the decision you make most often, then work backwards to the number informing it.
- Should we spend more or less? You need cost per result.
- Is the pipeline healthy? You need enquiries and their quality.
- Are we growing? You need new customers, separated from repeat ones.
- Is this profitable? You need contribution, not revenue.
Different businesses land on different sets, and key performance indicators copied from an article about somebody else's company are where most bad reporting starts.
The popular key performance indicators that mislead
Several appear on almost every marketing report and quietly cause bad decisions.
Impressions measure opportunity, not attention. Click through rate can rise while quality falls. Cost per lead falls when you attract people who never buy. Website traffic can grow while enquiries do not. Return on ad spend uses revenue rather than profit, so a healthy figure can still lose money.
Each of them sits somewhere on the chain that runs from impression to profit, and reading them in that order is most of what keeps them honest.
None of those is useless. Each is context that becomes dangerous the moment it gets promoted to a headline.
Leading and lagging key performance indicators
Some numbers tell you what happened and some tell you what is about to.
Revenue is lagging: honest, and too late to act on. Enquiries, quotes issued and conversations started are leading, since they move first and predict what arrives later. Consequently a report containing only lagging numbers describes history, and a report containing only leading ones is optimistic.
Keep at least one of each. A set of key performance indicators built only from what has already happened leaves you reacting, and one built only from early signals leaves you surprised.
Setting key performance indicators a target can support
A target without a basis is a wish, and everybody knows it, which is why nobody defends one.
Base it on what you have done before, on what the business needs to work, or on both. "Cost per enquiry below the level where a customer stops being profitable" is a target with a reason. "Twenty percent better" usually is not, unless something specific changed to make it reachable.
Comparing against the right thing
Last month is the most common comparison and one of the worst, because most businesses move with the seasons.
Compare against the same period last year where you have the history. Use your own trend rather than an industry figure somebody quoted without knowing your margins. And where volumes are small, widen the window, since a month with four customers cannot support a confident conclusion about any of your key performance indicators.
Reporting key performance indicators to somebody else
A report is a decision aid, not an archive. Three sentences before any chart.
What changed, why it changed, and what happens next. Numbers appear as evidence for those three answers rather than as the content. Meanwhile a report that lists everything and explains nothing puts the work of interpretation on the reader, who usually declines it.
Say what you plan to do about it, too. Key performance indicators earn their name when somebody changes a decision because of them, and a report ending without a next step rarely causes that.
Reviewing the set itself
Indicators go stale. A number that was decisive during a launch is often irrelevant a year later.
So revisit the set annually and after any real change to the business. Ask the same question of each: would we act on this. Anything failing moves to the appendix, and something else is probably now worth promoting.
Expect one or two of your key performance indicators to survive that review every year. The rest were useful once and are now habit, which is a different thing entirely.
What a small business actually needs
Usually four numbers: enquiries, the proportion that were genuine, cost per customer, and revenue or contribution.
Those fit on one page, take twenty minutes a month to update, and answer nearly every question worth asking. Key performance indicators do not have to be sophisticated. They have to be few, honest and read.
Put them in the same place every month, in the same order. Consistency is what turns a set of numbers into a trend, and a trend is what people actually act on.
Frequently asked questions
What makes a good key performance indicator?
It passes one test: if this number moved, would we do anything differently. Where the answer is no, it is context rather than a key indicator. Impressions and follower counts usually fail that test.
How many should I track?
Three to five on the front page, with everything else in an appendix. Beyond that nobody reads them, and a report nobody reads is not a report. Keep the detail available for whoever asks.
How do I choose mine?
Start from the decision you make most often and work backwards to the number informing it. Spending decisions need cost per result. Pipeline questions need enquiries and their quality. Growth needs new customers separated from repeat ones, and those four cover most sets of key performance indicators.
Which common metrics mislead?
Impressions measure opportunity not attention. Click through rate rises while quality falls. Cost per lead falls when you attract people who never buy. Return on ad spend uses revenue rather than profit.
What is the difference between leading and lagging numbers?
Lagging key performance indicators such as revenue are honest and too late to act on. Leading ones such as enquiries and quotes issued move first and predict what follows. Keep at least one of each in any report you produce.
How should I set targets?
On what you have done before, on what the business needs to work, or both. A target with a reason gets defended. A round percentage improvement with no basis is a wish, and everybody treats it as one.
What should I compare against?
The same period last year where you have the history, and your own trend rather than an industry figure quoted without knowing your margins. Where volumes are small, use a longer window.
What does a small business really need?
Usually four numbers: enquiries, the proportion that were genuine, cost per customer, and revenue or contribution. Those fit on one page, take twenty minutes a month to update, and answer nearly every question worth asking. Key performance indicators do not have to be sophisticated, only few and honest.