Human Coalition

How Meta Ad Campaign Optimization Goals Affect the Cost of Acquisition

Experiment ID: #201362

Human Coalition

Experiment Summary

Timeframe: 11/10/2025 - 12/15/2025

To better understand how Meta campaign optimization goals impact both giving behavior and campaign efficiency, Human Coalition tested two different optimization strategies. The control group optimized for Purchases—Meta’s default goal of maximizing the number of tracked donation conversions. The treatment group optimized for maximum Return on Ad Spend (ROAS)—Meta’s strategy that seeks to maximize the total dollar value returned for every ad dollar spent.

The team hypothesized that ROAS (Return on Ad Spend) optimization might reduce the number of conversions but increase the average gift size enough to improve overall ROI. This trade-off between conversion volume and gift value is central to the donor motivation question: Are there optimization strategies that attract higher-value donors, even at lower volume?

Both groups used the same ad creatives, daily budgets, and targeting parameters, isolating only the optimization goal as the variable under test. This test aimed to surface insights on how optimization settings shape both donor behavior and bottom-line campaign performance.

Research Question

We believe that optimizing Meta ads for Maximum ROAS for a mixed acquisition and retargeting audience will achieve higher ROI than optimizing for Purchases, because focusing the algorithm on revenue-per-dollar spent may prioritize higher-value donors, even if it reduces overall conversion volume.

Design

C: Purchase Optimized Campaign

T1: ROAS Optimized Campaign

Results

	Treatment Name	Conv. Rate	Relative Difference	Confidence
C:	Purchase Optimized Campaign	9.3%
T1:	ROAS Optimized Campaign	9.1%	-2.5%	12.0%

This experiment has a required sample size of 119,678 in order to be valid. Unfortunately, the required sample size was not met and a level of confidence above 95% was not met so the experiment results are not valid.

Key Learnings

Optimizing for Maximum ROAS did not outperform Purchase optimization in this campaign. While the average gift in the ROAS-optimized group was higher ($49.27 vs. $38.24), the overall performance suffered due to a sharp decline in donation volume, with the treatment group generating only 39 gifts compared to 175 in the control. As a result, the gross cost per donor more than doubled in the treatment group ($213.70 vs. $86.17), and overall ROI decreased. The treatment group saw a 2.5% decrease in Donations for All Traffic, which was not significant at a 12.0% level of confidence. Therefore, the hypothesis that focusing on revenue per dollar spent would attract higher-value donors, potentially improving overall ROI, was not supported by the data. This suggests that Meta’s ROAS optimization may struggle to scale donation volume for nonprofit campaigns, especially when the audience includes a large proportion of new or cold prospects. While ROAS optimization likely prioritized users predicted to give larger gifts, the reduced reach or conversion frequency significantly undercut overall efficiency. We learned that gift value alone cannot offset a steep drop in conversions—at least in this audience and campaign context. Optimizing for total donor volume may remain the more effective strategy when the primary goal is acquisition, particularly if follow-up engagement can increase lifetime value over time. These results prompt further questions: Could ROAS optimization be more effective with warmer audiences or higher average gift campaigns? Are there hybrid approaches (e.g., campaign-level ROAS goals with ad set-level Purchase optimization) that could better balance value and volume?

Experiment Documented by NextAfter

Question about experiment #201362

If you have any questions about this experiment or would like additional details not discussed above, please feel free to contact them directly.