When running paid marketing campaigns, it’s critical to determine whether your ads truly drive incremental conversions or simply capture customers who would have converted anyway. Incrementality testing helps marketers measure the real impact of paid campaigns by comparing the behavior of exposed and unexposed audiences.
In this guide, we’ll explore two of the most effective methods for incrementality testing: Geo Lift Testing and Public Service Announcement (PSA) Testing.
What is Incrementality Testing?
Incrementality testing isolates the effect of a marketing campaign by measuring how many conversions were directly influenced by advertising. Unlike standard attribution models, which assign credit based on user touchpoints, incrementality testing removes bias by introducing a control group.
Why Incrementality Testing Matters
- Helps determine if paid media is actually driving additional conversions.
- Identifies wasteful ad spend on users who would convert organically.
- Provides unbiased insights beyond last-touch or multi-touch attribution.
- Enables better budget allocation for future campaigns.
Method 1: Geo Lift Testing
How Geo Lift Testing Works
Geo Lift testing measures campaign effectiveness by splitting geographic regions into test and control groups:
- Test Group: Areas where paid ads are actively running.
- Control Group: Areas where ads are paused or excluded.
- Compare conversion rates between groups to measure incremental lift.
Steps to Run a Geo Lift Test
- Select Comparable Geo Regions
- Use cities or states with similar demographics, seasonality, and purchasing behavior.
- Example: Running ads in California (test) vs. Texas (control).
- Pause Ads in Control Locations
- Ensure no paid ads are shown in the control group.
- Run the Campaign for 4-6 Weeks
- Track conversions, revenue, and key performance indicators (KPIs).
- Measure Incrementality Using Lift Calculation
Use the formula:Lift (%) = [(Test Group Conversion Rate - Control Group Conversion Rate) / Control Group Conversion Rate] * 100 - Analyze Results & Optimize Budget Allocation
- If the test group shows significant lift, the campaign is driving new conversions.
- If lift is low, reallocate budget to higher-performing channels.
Tools for Geo Lift Testing
- Facebook Geo Lift Study (Built-in tool for measuring incrementality)
- Google Ads Experiments (Geo-based A/B testing)
- Uber’s CausalImpact Package (For advanced statistical modeling)
Method 2: PSA Testing
How PSA Testing Works
PSA testing involves replacing paid ads with non-commercial messages (e.g., charity announcements or public service ads) to create a true control group.
Steps to Run a PSA Test
- Divide Your Audience Into Two Groups
- Test Group: Receives your paid campaign ads.
- Control Group: Receives PSA ads unrelated to your product.
- Ensure Similar Exposure Levels
- Both groups should have equal ad impressions and frequency.
- Measure Differences in Conversions
- Compare conversion rates between PSA and paid ad groups.
- Calculate Incremental Lift
- If the paid ad group outperforms the PSA group, the campaign has true incremental impact.
Why PSA Testing Works
- Controls for brand awareness effects that might influence organic conversions.
- Removes confounding variables that can distort paid campaign attribution.
Tools for PSA Testing
- Google Ads Experiments (Split audience A/B testing)
- Facebook Brand Lift Studies
- Custom A/B testing scripts in Looker Studio & BigQuery
Interpreting Incrementality Test Results
Key Metrics to Measure
| Metric | Definition |
|---|---|
| Incremental Lift % | Percentage increase in conversions due to ads. |
| Cost per Incremental Conversion (CPI) | Ad spend divided by incremental conversions. |
| Return on Ad Spend (ROAS) | Revenue generated from ad spend. |
What to Do Based on Test Results
- High Lift: Increase budget for high-performing campaigns.
- Low Lift / Negative Lift: Reduce spend or reallocate to organic efforts.
- Neutral Lift: Ads may not be necessary—focus on alternative acquisition strategies.
Best Practices for Running Incrementality Tests
- Use Statistical Significance – Ensure sample sizes are large enough for reliable results.
- Run Tests for At Least 4 Weeks – Longer test durations improve accuracy.
- Control for Seasonality – Run tests outside major shopping events to avoid external influence.
- Segment by Audience & Channel – Test different segments separately (e.g., new vs. returning customers).
Final Thoughts
Incrementality testing is one of the most powerful ways to measure the true impact of paid campaigns. By leveraging Geo Lift and PSA testing, marketers can make data-driven decisions that improve ad efficiency and ROI.
Next Steps
In upcoming articles, we’ll explore:
- How to Automate Incrementality Testing with BigQuery & Looker Studio
- Advanced AI-Driven Attribution Models for Paid Campaigns
- Blending Incrementality Testing with Multi-Touch Attribution
Stay tuned for more insights on data-driven marketing strategies!