Choose metrics that help you make a decision.

Start with the decision, not the dashboard

Marketing reports can contain many numbers without answering the question that matters. Before selecting metrics, identify the decision the report should support. Are you evaluating the relevance of traffic, the quality of enquiries or whether a page helps visitors complete a task?

Different questions call for different evidence. A metric should have a defined meaning, an identified source and enough context to be interpreted. A reporting format is useful only when its readers understand what the information can and cannot tell them.

Keep activity, behavior and outcomes distinct

Activity metrics describe what was delivered. Behavioral measures help show what people did. Business outcomes describe the value of the resulting activity. These categories can be connected, but they should not be treated as equivalent.

A campaign may generate enquiries that do not fit the offer. Reviewing enquiry quality with the business can therefore be as important as reviewing the number reported by the advertising platform.

Agree definitions and limitations

A “conversion” can mean different things in different systems. Document the event being counted, any exclusions and the reporting period. If several tools report the same journey differently, first understand their definitions rather than assuming one must be correct.

• Identify the source and owner of each metric.
• Document changes to tracking, page design or campaign settings.
• Explain missing data and known measurement limitations.
• Keep the reporting period and comparison method consistent where possible.

Turn observations into a next step

A useful review separates observation, interpretation and action. An observation might describe a change in traffic quality. An interpretation proposes a possible reason. The next step identifies how to investigate that reason or what to adjust.

Avoid treating a change in a reported metric as proof that a single action caused it. Audience mix, timing, market conditions and tracking changes can all affect the picture. When evidence is limited, state what remains uncertain and choose a proportionate next step.