Does a targeted lookalike audience convert at a higher rate than a broad lookalike audience on Meta?

Hoover Institution

Experiment Summary

Timeframe: 02/07/2025 - 02/12/2025

This experiment was conducted for the Hoover Institution, a public policy think tank dedicated to promoting ideas for a free society. The goal was to evaluate the impact of audience targeting strategies on Meta by testing a broad lookalike audience against a more targeted lookalike audience based on previous donors from economic content offers. The objective was to determine which approach would yield better cost efficiency for donor acquisition.

Research Question

We believed that using a lookalike audience based on past donors from economic content offers would improve cost efficiency in donor acquisition because the audience would be more aligned with historical donor behavior.

Results

This experiment has a required sample size of 11,220 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

Main Takeaway

The broad lookalike audience significantly outperformed the targeted lookalike audience, achieving a lower cost per donor and a lower cost per thousand impressions (CPM). With an 80% confidence level, this suggests that Meta favored the broader audience due to the larger data set, leading to more efficient delivery and better results.  

Performance Metrics and Insights:

Donor Acquisition Efficiency: Cost per donor was 75% lower for the broad audience. CPM (Cost Per Thousand Impressions): The broad lookalike audience had a 22% lower CPM, indicating that it was more cost-efficient to reach potential donors.  

Next Steps

1. Prioritize Broad Targeting: Use broad lookalike audiences when possible, as they provide better cost efficiency and leverage Meta’s optimization.
2. Refine Targeted Strategies: Reserve highly targeted lookalikes for campaigns testing strategies that significantly differ from current donor acquisition efforts.
3. Lower CPMs for Single Offers: Explore ways to reduce CPMs in single-offer campaigns, such as testing multiple ad sets or adjusting budget allocation to optimize for lower CPM levels (e.g., targeting $35 CPM).

This experiment highlights the effectiveness of broad targeting on Meta and underscores the importance of leveraging platform algorithms for optimal delivery. Further testing will refine audience segmentation strategies to balance cost efficiency with audience specificity.  

---