FIELD GUIDE 17 / Data & measurement
Marketing metrics: build a ladder from delivery to value
A useful measurement plan forms a chain: was the message delivered, was there an opportunity to notice it, did relevant behaviour change and did that change create durable value? No single metric answers every step.
QUICK ANSWER
The idea in one minute.
A clear framework for reach, attention, action, incrementality and business outcomes. Every metric answers a different question A dashboard needs definitions and decisions Business outcomes require a credible comparison
KEY IDEAS
What you will learn
- Every metric answers a different question
- A dashboard needs definitions and decisions
- Business outcomes require a credible comparison
01
Separate delivery from effect
Impressions, reach and frequency describe distribution. Viewability and completion describe the opportunity to see or hear. Clicks and visits describe interaction. None proves that advertising caused a sale, and a low-interaction format can still build memory.
Label each metric by the question it answers. This prevents a delivery number from becoming a business conclusion merely because it is available quickly. Include the unit, time window, population, data source and known exclusions.
02
Connect intermediate outcomes
Qualified visits, product views, enquiries, trials and add-to-cart events can diagnose movement toward the goal. Define quality before launch: a lead may need to meet eligibility criteria, and a visit may need enough engagement to indicate the page actually loaded and was used.
Brand studies can measure awareness, recall, consideration or perception. Use questions and samples designed for the construct. A favourable answer after exposure is association, not automatically a durable change in behaviour.
03
Reach business value carefully
Revenue is not profit. Add margin, returns, service cost, retention and time to value where they matter. Customer lifetime value is a model that depends on assumptions about future behaviour and discounting; publish those assumptions with the result.
Incremental value compares what happened with a credible estimate of what would have happened without the activity. Randomised holdouts are strongest when feasible. Geographic tests and well-designed observational methods can help when randomisation is unavailable, but require explicit assumptions.
04
Design dashboards for decisions
A dashboard should show objective, primary outcome, leading indicators, guardrails, target range and the decision owner. Trend lines and comparisons are usually more informative than isolated totals. Annotate launches, outages, price changes and measurement breaks.
Remove metrics that do not change a decision. More numbers can reduce understanding when definitions conflict or teams select whichever figure supports a preferred story. A metric dictionary and regular quality review make the dashboard a shared language rather than a scoreboard.
SOURCE DESK
Research and further reading
These links lead to public guidance, open textbooks or freely accessible research. This guide explains the ideas in original language; open the sources to examine context, definitions and limitations.
- Marketing Plan Progress Using Metrics ↗OpenStax introduction to selecting and interpreting marketing metrics.
- Metrics Used to Evaluate Advertising Campaign Effectiveness ↗OpenStax overview of delivery, response and financial indicators.
- Digital Analytics Program ↗US Government programme resources for consistent public-site measurement.
- Causal Inference ↗Open educational material on potential outcomes, experiments and observational comparisons.