Incrementality Testing Roadmap: A 90-Day Guide for Advanced Marketers

Introduction: Incrementality testing is essential for marketers seeking to understand the true impact of their campaigns and drive incremental revenue. This detailed 90-day roadmap provides a step-by-step guide on how to execute and analyze tests at the channel, tactic, and campaign levels. The goal is to uncover actionable insights and implement strategies that optimize ad spend and increase profitability. By the end of this cycle, you’ll know how to set new platform ROAS targets, reallocate budgets, and continuously test for sustained growth.

‍

Day 1-15: Channel-Level Testing for the First Channel

Focus: Start by testing one channel at a time to isolate performance without triggering learning phase disruptions across platforms.

Process:

Select Your First Channel – Begin by choosing a primary channel like Meta Ads or Google Ads. This ensures you’re starting with a key area of your marketing mix.
Holdout Regions – For channel-level testing, you’ll need to exclude holdout regions at the account level. Excluding regions account-wide ensures that all campaigns within that channel are paused in the holdout regions, giving you a clean, accurate measure of the channel's performance across all campaign types.
How to Exclude Locations:some text
- Meta Ads: Open your Meta Ads Manager and go to your Advertising settings. Navigate to the "Audience Controls" section and toggle on "My business can only advertise in specific locations." This setting allows you to exclude or include regions across your entire account, ensuring all campaigns within Meta are excluded from the selected regions.
- Google Ads: Go to "Tools & Settings," then "Placement Exclusion Lists." Create a new list, add the regions you want to exclude, and apply this list across all campaigns.
Run the Test – After setting up your account-level exclusions, run the test for 15 days. This will give you sufficient time to collect data on the channel's incremental impact. Ensure that the holdout regions remain fully excluded throughout the test period, and that campaigns continue running normally in the test regions.
Test Duration – A 15-day window is standard for channel-level testing, allowing you to gather enough data to reach statistical significance.

Action:

Calculate the Incrementality Factor: Once the test concludes, compare iROAS (incremental ROAS) and platform ROAS. For example, if platform ROAS is 2.0 and iROAS is 3.5, the incrementality factor is 3.5 / 2.0 = 1.75.
Set New Platform ROAS Targets: Use the incrementality factor to set a new platform ROAS target. In this case, divide platform ROAS by the incrementality factor (2.0 / 1.75 = 1.14). This new target ensures that your ad spend is focused on driving true incremental revenue.
Reallocate Budget: If the channel proves highly incremental, consider increasing your budget for this channel. If not, shift your focus to another channel or tactic to improve efficiency.

‍

Day 15-30: Tactic-Level Testing (Tactic 1)

Focus: After channel-level testing, focus on a key tactic within the channel, such as Prospecting or Retargeting, to see which tactics are most incremental.

Process:

Select the First Tactic – Begin by testing Prospecting. This tactic is typically used for new customer acquisition and can be crucial for determining which audiences are more incremental.
Holdout Regions – At the tactic level, you’ll need to exclude regions for this specific tactic. Ensuring that only this tactic is excluded from the holdout regions allows for clean data and isolates its impact without interference from other tactics running simultaneously.
How to Exclude Locations:some text
- Meta Ads: Go to the Ad Set Level, navigate to the "Audience" section, and click "Edit" next to "Locations." Select "Exclude" and enter the holdout regions for your Prospecting campaign.
- Google Ads: In your campaign settings, go to the "Locations" section and add location exclusions for this specific campaign.
Run the Test – After setting exclusions, run the test for 15 days. Ensure that only your Prospecting campaigns are active in the test regions, while they remain completely excluded in holdout regions.
Test Duration – Run this tactic-level test for 15 days unless the tactic is higher in the funnel (e.g., TV ads), which may require a longer test period (28-45 days).

Action:

Calculate Incrementality: Compare iROAS to platform ROAS. For example, if Prospecting’s platform ROAS is 1.5 and iROAS is 3.0, the incrementality factor is 3.0 / 1.5 = 2.0.
Set Tactic-Specific Targets: Based on this incrementality factor, adjust your platform ROAS target (1.5 / 2.0 = 0.75). This allows you to scale Prospecting campaigns while ensuring profitability from incremental revenue.
Reallocate Budget: If Prospecting is highly incremental, shift more budget into this tactic. If not, consider testing another tactic like Retargeting.

‍

Day 30-45: Tactic-Level Testing (Tactic 2)

Focus: Test the next key tactic in the same channel, such as Retargeting, to see if it delivers better incremental results than Prospecting.

Process:

Select the Next Tactic – Move on to Retargeting, which typically focuses on re-engaging users who have already shown interest or taken specific actions.
Holdout Regions – Proper holdout region selection is still essential to avoid data contamination. Ensure regions are excluded for the Retargeting tactic.
How to Exclude Locations:some text
- Meta Ads: Exclude regions at the Ad Set Level within the Retargeting campaign by editing the audience locations.
- Google Ads: Exclude regions for Retargeting at the campaign level.
Run the Test – After setting up your exclusions, run the Retargeting test for 15 days, ensuring it operates only in the test regions.
Test Duration – A 15-day window is sufficient for most Retargeting tactics unless there are delays in conversion attribution, in which case extend to 28 days.

Action:

Calculate the Incrementality Factor: For example, if Retargeting shows a platform ROAS of 5.0 and an iROAS of 3.5, the incrementality factor is 3.5 / 5.0 = 0.70.
Set New ROAS Targets: Adjust your target based on the incrementality factor (5.0 / 0.70 = 7.14). Now you know what your Retargeting campaigns must hit to drive true incremental profit.
Budget Reallocation: Based on results, shift budget to the more incremental tactic. If Retargeting performs better than Prospecting, allocate more budget toward Retargeting for optimal results.

‍

Day 45-60: Campaign-Level Testing (Campaign 1)

Focus: Move on to campaign-level testing, where you can analyze different campaign settings, audiences, ad formats, or attribution models.

Process:

Choose the First Campaign Variable – Start by testing broad vs. narrow audiences or video ads vs. static images. Campaign-level testing allows you to optimize how your creative and targeting work together.
Holdout Regions – At the campaign level, exclude regions where your test campaign will not run. This allows you to isolate the impact of each variable without interference from other campaigns.
How to Exclude Locations:some text
- Meta Ads: Exclude holdout regions at the Ad Set Level for the specific campaign.
- Google Ads: Apply exclusions at the campaign level under "Locations."
Run the Test – After setting exclusions, run the campaign test for 15 days to collect sufficient data on the variable being tested.
Test Duration – A 15-day test window is typical unless you’re testing a high-funnel campaign, which may require more time.

Action:

Calculate Incrementality: For example, if broad audience targeting shows a platform ROAS of 2.0 and iROAS of 3.0, the incrementality factor is 3.0 / 2.0 = 1.5.
Set Campaign-Specific Targets: Adjust the ROAS target accordingly (2.0 / 1.5 = 1.33). Now you know that broad audiences need to hit a 1.33x ROAS to be incremental.
Optimize Strategy: If video ads show higher incrementality than static images, increase your investment in video creatives. Similarly, shift budget toward the most incremental audience types.

‍

Day 60-75: Campaign-Level Testing (Campaign 2)

Focus: Test a new campaign variable, such as attribution models or other settings, to further optimize your strategy.

Process:

Select the Next Variable – For example, test 1-day click vs. 7-day click attribution windows.
Holdout Regions – Carefully select holdout regions for this specific campaign variable.
How to Exclude Locations:some text
- Meta Ads: Exclude locations at the Ad Set Level for the campaign.
- Google Ads: Use campaign settings to exclude locations.
Run the Test – Run the test for 15 days, making sure to isolate this specific variable for clean results.

Action:

Calculate the Incrementality Factor: For example, if platform ROAS is 6.0 and iROAS is 4.5, the incrementality factor is 4.5 / 6.0 = 0.75.
Set New Targets: Use this factor to adjust your ROAS target (6.0 / 0.75 = 8.0). This allows you to optimize for true profitability based on attribution models.
Adjust Settings: Shift focus to the attribution model or campaign variable that drives the highest incremental results.

‍

Day 75-90: Campaign-Level Testing (Campaign 3)

Focus: Test a final variable or campaign setup to refine your approach across all active campaigns within the same channel.

Process:

Select the Final Campaign Variable – Choose another significant variable to test, such as different campaign structures (e.g., dynamic product ads vs. catalog ads) or audience exclusions (e.g., excluding recent purchasers). This helps ensure you’re optimizing every aspect of your campaigns for incremental performance.
Holdout Regions – For this final test, it’s important to select the most relevant regions for exclusion to get an accurate understanding of the variable’s impact. Ensuring clean data at this stage is essential for wrapping up your 90-day cycle with actionable insights.
How to Exclude Locations:some text
- Meta Ads: At the Ad Set Level, exclude locations by editing the audience section. Ensure all regions you are testing have consistent exclusions.
- Google Ads: Exclude locations at the campaign level under “Locations” in the settings.
Run the Test – Execute the campaign test for the next 15 days. Keep in mind that this is your final campaign-level test in the cycle, so aim to gather insights that can tie back to your broader campaign strategy across the entire channel.
Test Duration – A 15-day test period should be enough to see the impact of this final variable, barring long attribution windows for upper-funnel activities.

Action:

Calculate the Incrementality Factor: For example, if your campaign structure test results in a platform ROAS of 4.0 and an iROAS of 3.0, the incrementality factor would be 3.0 / 4.0 = 0.75.
Set Final ROAS Targets: Divide the platform ROAS by the incrementality factor (4.0 / 0.75 = 5.33) to determine your new target for that campaign structure. Now you can understand what your true performance goal should be for that specific setup.
Refine Campaign Strategy: Based on this test’s outcome, finalize adjustments to your campaign structures, audience exclusions, or any other tested variable. This ensures that you are optimizing your campaigns with the most precise and impactful data possible.

‍

Summary: Adopting a Culture of Ongoing Incrementality Testing

The 90-day roadmap outlined above serves as a complete cycle of channel, tactic, and campaign-level incrementality testing. But incrementality is not a one-time exercise. To maximize long-term profitability and keep your marketing strategies aligned with changing customer behavior, platform algorithms, and industry trends, it’s crucial to adopt a culture of ongoing testing.

Here’s how you can incorporate this process into your marketing strategy:

Repeat the 90-Day Cycle
After completing the first round of testing, start the process over. Return to channel-level testing with any adjustments made from the previous cycle’s insights. By continuously testing and refining channels, tactics, and campaigns every 90 days, you will stay ahead of changes in ad platform behavior and market shifts.
Regularly Reassess Holdout Regions
With each new cycle, re-evaluate your holdout regions to ensure they are still optimal. Changes in market conditions or customer behavior may require adjusting which regions are held out to get the best results.
Update Platform ROAS Targets Regularly
As you uncover new incrementality factors in each cycle, update your platform ROAS targets. This will ensure that your ad budget is always optimized for true profitability, even as market conditions and platform algorithms evolve.
Integrate Incrementality Testing into Strategy Decisions
Make incrementality testing part of every major strategy discussion within your marketing team. Whether planning a new campaign, expanding into a new channel, or adjusting your media mix, use the results from your tests to guide decisions and ensure that your ad spend is truly driving incremental returns.
Incorporate Ongoing Collaboration with Stella
Stella’s platform plays a key role in making sure your incrementality testing is both effective and efficient. With our tools, you can continually analyze and optimize your holdout regions, monitor your tests in real time, and make data-driven decisions based on clear, statistically significant insights. Partnering with Stella ensures that each cycle of testing is conducted properly, yielding the most actionable results.

‍

Why Ongoing Incrementality Testing Matters

Marketing is not static—platforms change, audience behaviors shift, and your competition evolves. Without ongoing testing, you risk falling behind and relying on outdated assumptions about what drives real growth. By consistently repeating this 90-day cycle, you’ll:

Stay Ahead of Learning Phases: Alternating between channels and tests helps you avoid prolonged learning phases in your ad platforms, which can hinder campaign performance.
Maximize Incremental Revenue: Continuously refining and optimizing your campaigns ensures that you are always driving the highest possible returns from your marketing spend.
Make Data-Driven Decisions: Incrementality testing provides clear, actionable data. Instead of guessing which campaigns are performing, you’ll know exactly where to allocate your budget for maximum impact.
Adapt to Market Changes: By regularly testing, you can stay flexible and quickly adapt your strategies to changes in the market, audience behaviors, or platform updates.

‍

Next Steps

By following this roadmap, you’ve set the foundation for a marketing program built on incrementality testing and continuous optimization. Remember, the key to success lies in making this process a core part of your strategy—not just a one-off exercise. Incrementality testing will not only improve your campaigns but also empower your team to make smarter, more profitable decisions at every stage.

Ready to start your next cycle? Get in touch with Stella to ensure that your next round of tests is set up for success. Together, we can help you implement a seamless and ongoing culture of data-driven marketing.

‍