Definition
A PSA (public service announcement) test replaces your ads with neutral ads to estimate incrementality while keeping auction dynamics similar.
Example
Instead of pausing ads, you serve a neutral PSA creative to a holdout group and compare conversions.
How to use it
- Useful when you cannot fully turn off ads without affecting auctions.
- Interpret with care; PSA design and audience leakage can bias results.
- Ensure the PSA creative is neutral and does not change user intent.
Common mistakes
- Using a PSA that attracts clicks and contaminates the control group.
- Running tests too short to cover conversion lag.
Why this matters
This term matters because it affects how you interpret performance and make budget decisions. If you use inconsistent definitions or windows, ROAS/CPA can look "better" while profit gets worse.
Practical checklist
- Write a 1-line definition for "PSA Test" that your team will use consistently.
- Keep the time window consistent (weekly/monthly/quarterly) when comparing trends.
- Segment results (channel/plan/cohort) before drawing big conclusions from blended averages.
- Sanity-check with a related calculator from the same category on MetricKit.
- Read the related guide (e.g., Attribution vs incrementality: what to trust, when, and how to test) for context and common pitfalls.
Where to use this on MetricKit
Calculators
- Break-even ROAS Calculator: Estimate the break-even ROAS based on contribution margin assumptions.
- Target ROAS Calculator: Estimate a target ROAS to cover variable costs plus a desired margin buffer.
- Paid Ads Funnel Calculator: Model CPM -> CTR -> CVR to estimate CPC, CPA, ROAS, and profit per 1,000 impressions (with margin and variable costs).
- ROI Calculator: Calculate Return on Investment (ROI) for a campaign or project.
- Incrementality Lift Calculator: Estimate incremental conversions, incremental ROAS, and incremental profit from a holdout test.
Guides
- Attribution vs incrementality: what to trust, when, and how to test: A practical guide to attribution vs incrementality: common attribution models, window pitfalls, how MER/marginal ROAS fit in, and how to run holdout/geo tests.