Marketing Measurement: Metrics, Formulas and a Practical Plan
Shusaku Yosa
Marketing measurement connects spending and customer behavior to business outcomes so a team can decide what to change. Start by agreeing on the outcome, the cost boundary and the evaluation window. A dashboard is useful only when those definitions support a real decision.
Define the metrics before comparing results
Metric | Formula | Boundary to document |
|---|---|---|
CPA | Advertising spend / target conversions | Name the action: a lead is not necessarily a customer |
CAC | (Acquisition marketing costs + acquisition sales costs) / new customers | Allocate labor and agency costs consistently; account for sales lag |
Conversion rate | Target outcomes / eligible sessions, users or clicks × 100 | Choose one denominator and keep it consistent |
ROAS | Attributed advertising revenue / advertising spend × 100 | Revenue efficiency before product and operating costs |
Marketing ROI | (Attributed gross profit − marketing investment) / marketing investment × 100 | The management definition used in this guide |
Revenue LTV | Average purchase value × purchase frequency × customer lifetime | Use matching time units and label forecasts |
Gross-profit LTV | Revenue LTV × gross margin | Do not compare a revenue-only LTV with a profit-based benchmark |
Other ROI conventions use total costs as the denominator. Google Ads explains a total-cost approach. Print the chosen formula next to the result. A zero conversion count makes CPA undefined; it does not make acquisition free.
A worked example: revenue is not profit
Consider a hypothetical campaign in dollars: $10,000 in ad spend, $30,000 in attributed revenue and $12,000 in cost of goods sold. ROAS is 300%; gross profit is $18,000. With ad spend as the only marketing investment, marketing ROI is ($18,000 − $10,000) / $10,000 = 80%. Add $2,000 in creative and management costs and ROI becomes 50%. These are teaching examples, not market benchmarks.
If the campaign produces 200 inquiries and 20 new customers, advertising CPA is $50. With $6,000 in acquisition sales costs, CAC is ($12,000 + $6,000) / 20 = $900. The different answers reflect different outcomes and cost scopes. Comparing them without those definitions is misleading.
Create a measurement plan by channel
Channel | Leading / business outcome | Sources and checks |
|---|---|---|
Paid media | Qualified inquiry / customer and gross profit | Ad platform, web analytics and CRM; check attribution windows |
Content and SEO | Search visits / inquiry and opportunity | Search Console, analytics and CRM; compare pages over consistent periods |
Recipient clicks / qualified opportunity | Delivery logs and CRM; consider automated opens and unsubscribes | |
Events | Relevant conversations / subsequent opportunities | Attendance records and CRM; separate duplicates and existing deals |
For every metric, record the campaign ID, event name, qualifying condition, exclusions, source, owner, refresh cadence and reporting period. Track a successful form submission rather than a click on its button. Test success, failure and repeat submission, then reconcile the result with actual records.
Move from collection to a decision
- Design: choose the decision, primary outcome and guardrails before launch. Lead volume needs quality checks such as opportunity rate.
- Validate: align time zones, currencies, tax treatment and reporting grain. Keep missing values distinct from zero.
- Analyze: break averages into channel, customer segment and acquisition cohort. Avoid large changes based on a handful of conversions.
- Act: record the observation, explanation to test, change, owner and review date. Keep the number of simultaneous changes manageable.
Separate attribution from incremental impact
Platform reports allocate credit under their own rules. Adding conversions across platforms can count the same sale more than once. Use an agreed order or CRM record for business totals and platform reporting for within-platform optimization.
To estimate what the campaign caused, compare against a credible counterfactual, using randomized tests where feasible. Record seasonality, promotions and sales activity. A before-and-after increase alone cannot establish causal impact.
Use a short decision log
Observation | Check | Action |
|---|---|---|
CPA falls but win rate falls too | Lead mix and qualification | Delay expansion and compare qualified opportunities |
Visits rise but tracked outcomes remain zero | Actual submissions and event configuration | Fix measurement and separate affected periods |
Costs run ahead of budget | Committed costs and sales-cycle maturity | Update the forecast and approve any reallocation |
Connect measurement results to owners, costs and follow-up work. Xtrategy features explains the project and budget-management capabilities available for that workflow. Maintain tracking definitions in the measurement systems that supply your data.