Measuring something does not automatically improve it. A useful business metric helps you answer a question, notice a change or decide what to investigate. A poorly chosen one can reward activity that looks impressive while missing the purpose of the work.
Before adding another dashboard, ask: what decision would this information help us make?
Start with the business question
“How many enquiries did we receive?” and “Are we attracting suitable customers?” are different questions. The first is a count; the second requires an agreed definition of suitability and information about what happened next.
Choose a small set of measures that illuminate the question. Do not assume that a metric is useful because it is easy to collect or widely used elsewhere.
Define exactly what is counted
Write down the measure, source, period, owner and important exclusions. For a rate, define both the numerator and denominator. Specify how duplicate records, missing information and test activity are handled.
For example, an enquiry-to-booking rate is not meaningful until “enquiry” and “booking” are consistently defined. A clicked booking link is not a completed appointment, and a test submission is not a genuine prospect.
Establish a credible baseline
Check the starting information before setting a target. Does it cover a representative period? Did the collection method change? Are there gaps or unusual events that affect the comparison?
Keep the underlying counts alongside percentages. An apparent increase from one to two outcomes is different in scale and uncertainty from an increase from one hundred to two hundred. Neither comparison, by itself, explains the cause.
Pair activity with outcome and quality
Activity measures can show what happened in a process; outcome measures show a result of interest. Neither tells the whole story. More calls may not mean better conversations, and faster handling may come at the expense of accuracy.
If you use an early indicator to anticipate an outcome, test whether the relationship actually holds in your context. Do not treat an assumed leading indicator as a reliable prediction.
Watch for unintended incentives
Ask what someone might do to improve the number without improving the work. A target for completed tickets, for instance, could encourage premature closure if quality is ignored.
In a hypothetical service team, the review pairs response time with reopened requests and a small quality check. The purpose is to understand service, not rank people on a single number.
Explain how information will be used and collect only what is appropriate. Personal monitoring, employment decisions and sensitive data require suitable safeguards and specialist input; a dashboard is not a substitute for fair management.
Investigate changes before claiming results
If a measure improves after a new process begins, consider other explanations: demand, seasonality, staffing, definitions or missing records. A before-and-after comparison does not automatically show that the change caused the result.
Use the information alongside feedback from the people doing and receiving the work. Record what is known, what remains uncertain and what you will investigate next.
Make reviews lead to decisions
Give each review a purpose. Decide whether to continue, adjust, investigate or stop an activity, and name an owner for the next action. Match the review timing to the decision rather than reporting every number as often as possible.
Our goal-setting guide and decision-making process can help connect measures to practical commitments.
Book a free 15-minute fit call, or apply for two weeks of free coaching. Applications are reviewed manually; acceptance and particular results are not guaranteed.
