Marketing Campaign Measurement Plan: KPIs, Baselines and ROI
A marketing campaign measurement plan defines the outcome a campaign should change, the metric used to judge it, the comparison needed and the decision that follows. Build the plan before launch. A dashboard can show what happened; it cannot, by itself, establish how much of that result the campaign caused.
This guide provides a practical scorecard for brand and marketing teams in Vietnam. It focuses on metric definitions and reporting decisions. For choosing surveys, comparison groups and research methods, see the separate guide to campaign effectiveness research.
Editorial guide by INTAGE Vietnam. The worked example below is hypothetical, not an INTAGE client result. Header image: AI-generated illustration of a fictional campaign-analysis meeting.
Start with one decision and one primary outcome
Write the decision first: continue a launch, change the creative, improve a landing page or reconsider the media allocation. Then identify the audience, geography and measurement period. A national awareness campaign and a local lead-generation campaign should not be assessed using the same primary outcome.
Choose one primary metric, supporting diagnostic measures and a guardrail. For example, qualified enquiries can be the primary outcome, landing-page completion can diagnose friction, and cost per qualified enquiry can be a guardrail. Agree what counts as qualified with the sales team before collecting results. Clicks and likes should not silently replace that definition during reporting.
A campaign KPI scorecard
| Campaign objective | Primary outcome | Useful supporting evidence | Interpretation limit |
|---|---|---|---|
| Build awareness | Change in unaided awareness or ad recall | Consistent audience survey, reach and frequency | Exposure or recall is not proof of purchase |
| Generate enquiries | Qualified enquiries using agreed criteria | Form completion, contactability and sales acceptance | Form submissions can include duplicates or unsuitable leads |
| Support sales | Additional completed sales or contribution | Transactions, returns, margins and a credible comparison | Attributed revenue is not automatically additional revenue |
| Encourage repeat purchase | Repeat purchase within a defined window | Customer cohorts and comparable purchase opportunities | A short window may miss the normal repurchase cycle |
Keep awareness questions and interview conditions consistent across waves. The brand awareness and recall guide explains why unaided recall should be measured before showing respondents a brand list.
Define the numerator, denominator and cost boundary
Conversion rate = completed target actions ÷ eligible opportunities × 100%. State whether opportunities mean users, sessions, leads or ad clicks. Use the same unit and time window throughout a comparison. Do not divide customer counts by an unrelated impression total and call the result a website conversion rate.
Cost per qualified enquiry = included campaign cost ÷ unique qualified enquiries. Specify whether the cost includes media only or also creative, agency and production work. Remove duplicate records using an agreed rule, and record whether enquiries can actually be contacted.
ROAS = attributed revenue ÷ advertising spend. This is a revenue-to-spend ratio, not a profit measure. Returns, discounts, gross margin and sales that would have occurred anyway can change its business meaning.
Campaign ROI = (incremental contribution before campaign costs − total campaign costs) ÷ total campaign costs × 100%. Define contribution consistently, after the relevant variable costs. If incrementality has not been credibly estimated, label the calculation as a scenario or assumption rather than proven causal ROI.
Choose the baseline before reading the result
A before-and-after comparison is useful context, but seasonality, competitor promotions, changes in distribution, stock availability and price can also affect outcomes. Record those conditions alongside the campaign dates. In Vietnam, avoid treating a Tet trading period and an ordinary month as directly equivalent without adjustment.
When causal measurement matters, consider a randomized holdout or a carefully designed geographic comparison. Google Ads documents Conversion Lift as a controlled comparison for estimating additional conversions; availability and study requirements depend on the account. An experiment still needs sufficient data, a suitable design and uncertainty reporting. A positive point estimate alone is not conclusive.
Worked example: more leads, but what business value?
Suppose a fictional campaign costs VND 120 million in total and records 400 unique enquiries, of which 100 meet the agreed qualification criteria. Cost per enquiry is VND 300,000; cost per qualified enquiry is VND 1.2 million. The second figure answers a different and often more useful question.
If the reporting platform attributes VND 600 million of revenue to VND 80 million of advertising spend, reported ROAS is 7.5. That does not establish profitability. In a separate illustrative scenario, if a credible comparison estimates VND 180 million of incremental contribution before the VND 120 million total campaign cost, ROI is 50%. Without that incremental contribution estimate, the ROI conclusion remains unproven.
Turn the findings into the next test
Separate three statements in the report: what was observed, what the comparison supports, and what remains uncertain. Show the reporting window, definitions, data exclusions, sample limitations and costs beside the headline result. For awareness surveys, report sample sizes and uncertainty; for leads, show the qualification and follow-up lag.
A low completion rate may suggest testing the form or offer. Strong engagement with weak qualified demand may suggest a mismatch in audience or message. Neither pattern alone establishes the explanation: use interviews, creative testing or a planned experiment to investigate it. Keep the next test specific enough to change a decision.
Discuss a measurement brief with INTAGE Vietnam
Share your campaign objective, target audience, timing, available data and the decision the study should inform. INTAGE Vietnam can discuss the appropriate scope through its market research services; methods, feasibility and deliverables should be agreed for the project.
