How to Test Facebook Ad Creatives With Less Waste

Transcript

If you're testing Facebook ad creatives, there's a good chance you're doing it wrong. And I don't mean just a little wrong. I mean fundamentally, structurally, completely wrong, wasting hundreds, if not thousands of dollars every single day on your ads. Across the brands we worked with at the Moonlighters, I have spent millions of dollars across th... Read More

Key Insights

• Ad-set budget optimization is inefficient when operators manually guarantee spending for every creative. This approach can direct money away from proven winners and toward new ads whose hit rate is generally lower than that of creatives already operating successfully at scale.
• Facebook attribution does not display every impression, interaction, engagement, or earlier touch involved in a purchase. Because people may encounter several ads before converting, judging ad sets mainly from the reported critical touch can lead operators to overspend or underspend in the wrong places.
• Uneven performance makes forced ad-set allocation costly. In the cited client account, the highest-spending ad set generated 1.1 return on ad spend, while lower-spending ad sets generated 1.5, 1.8, and 2.5, showing that budget rank did not match reported return.
• Meta’s ad-level creative-testing feature guarantees delivery by requiring a selected daily budget and test duration for two to five ads. The speaker considers guaranteed and evenly distributed delivery a disadvantage because it creates additional cost centers instead of prioritizing creatives that Meta identifies as stronger performers.
• Classic unrestricted campaign-budget optimization became less effective after Andromeda because newly introduced creative packs reportedly received hardly any spending. Although the structure previously allowed strong new ads to scale gradually, insufficient initial delivery made it harder for fresh creatives to prove their conversion potential.
• The M4 structure combines campaign-level budget allocation with seven-day minimum budgets on new ad sets. This arrangement is intended to provide enough delivery for testing and conversions while preserving the algorithm’s ability to direct more of the campaign budget toward ads demonstrating better performance.
• Creative organization works best when each pack combines a specific customer avatar with a specific concept. The speaker reports noticing targeting problems without this joint grouping since July 2025, with the timing corresponding to the Meta Andromeda updates discussed in the presentation.
• The prospecting CBO campaign serves as the central campaign for scaling, prospecting, and new customer acquisition. In the demonstrated setup, it uses campaign-budget bidding, website as the conversion location, conversion maximization, a configured data set and conversion event, and standard attribution.

Questions & Answers

Q: How should Facebook ad creatives be tested with less wasted spending?

Facebook ad creatives should be placed in a prospecting campaign-budget campaign, organized into ad sets by both customer avatar and creative concept. Seven-day minimum budgets should be assigned to new testing ad sets so they receive enough initial delivery to generate conversions. The remaining allocation stays under campaign-level control, allowing Meta to increase spending on stronger performers while limiting unnecessary spending on weaker tests.

Q: Why can ad-set budget optimization waste creative-testing money?

Ad-set budget optimization can waste money because it forces a predetermined amount of spending onto new creatives regardless of their relative performance. That money must either be diverted from established winners or added to the account’s total budget. Since new ads generally have a lower hit rate than already scaled ads, manual allocation can place resources in less productive areas and charge a premium for finding winners.

Q: Why does Facebook attribution make manual budget allocation difficult?

Facebook attribution does not reveal every touch, impression, view, engagement, or interaction that contributed to a conversion. It reports what Facebook considers the most critical touch, which the speaker says is often the last touch. A person may first encounter one ad and later click another, so performance can be shared across ad sets in ways that make manual spending decisions incomplete or misleading.

Q: What is wrong with Meta’s ad-level creative-testing feature?

Meta’s ad-level creative-testing feature requires advertisers to select two to five test ads, choose a daily test budget, define the test duration, and select a performance measurement. It then ensures delivery to those ads and distributes spending evenly. The speaker views that guarantee as a disadvantage because money is forced onto every test instead of being concentrated on the ads Meta identifies as stronger performers.

Q: Why is an unrestricted CBO campaign no longer optimal for new creatives?

An unrestricted CBO campaign previously worked by allowing new broad creative packs to start slowly and gain spending when their performance justified expansion. After the Andromeda changes, however, the speaker observed that new ads were receiving hardly any spending. Without sufficient initial delivery, new creatives may not collect enough conversions or performance evidence to demonstrate whether they deserve a larger share of the campaign budget.

Q: What is the difference between the M3 and M4 Facebook ad structures?

The key difference is that M4 introduces minimum ad-set budgets and groups creatives by both customer avatar and concept. M3 allowed packs to operate more freely under campaign-level allocation. M4 still uses a prospecting CBO campaign, but seven-day minimum budgets ensure that new tests receive some delivery while the broader campaign system continues directing additional spending toward better-performing ads.

Q: How should creative packs be grouped in the M4 structure?

Creative packs should be grouped by customer avatar and advertising concept together, rather than being assembled without a clear pairing of audience identity and message. The speaker says that omitting this combined grouping can make targeting less accurate. This pattern was reportedly observed beginning in July 2025, at the same time as the Meta Andromeda changes discussed in the presentation.

Q: How is the prospecting CBO campaign configured in the example?

The example creates a prospecting CBO campaign that combines scaling, prospecting, and new customer acquisition. Campaign budget is selected as the bid strategy, with a demonstration budget of $1,000 per day. At the ad-set level, the conversion location is set to the website, optimization is set to maximize conversions, the data set and conversion event are configured, and standard attribution is selected.

Summary & Key Takeaways

• Common creative-testing methods can allocate money inefficiently. Ad-set budgets and Meta’s ad-level testing feature force spending toward unproven creatives, potentially diverting resources from established winners. A classic campaign-budget structure gives Meta more control, but the speaker reports that new ads began receiving very little spend following the Andromeda updates.

• The proposed M4 prospecting structure combines algorithmic campaign-budget allocation with controlled minimum spending. Seven-day minimum budgets give new ad sets enough delivery to generate conversions and demonstrate potential. Successful creatives can then ramp quickly, while weaker ads receive limited testing rather than an unnecessarily large or evenly distributed share of the total budget.

• Creative packs should be organized by customer avatar and advertising concept together. The speaker says this grouping became especially important after changes observed since July 2025, around the Meta Andromeda updates. The example setup uses a prospecting CBO campaign for scaling and acquisition, with website conversions, standard attribution, and conversion-focused optimization.