Regarding FB Product Testing: My Methodology in the ASC Era

1 Product: Prepare at least 5 videos + 5 images for the first wave

Step 1:

Series ①: 1-1-10

Series ②: 1-1-Flexible (10 assets)

Step 2: Expansion and Scaling:

1. Copy all non-spending ads from Series ① 1-1-10 into a new 1-1-N series. Set a bid cap for this series with a generous threshold. If its performance outperforms Series ①, adjust the cost-per-conversion bid upward by $5-$20 based on results. Fine-tune until the daily budget is fully spent. Testing shows this new 1-1-N series often uncovers profitable new creatives. This method was discovered by accident this year—cloned ad sets often outperform the original Series ①.

2. For high-spending, high-performing ads in Series ①, test interest audiences and lookalike audiences using the original post. Develop these audiences yourself. This directly controls the variables: the creative is fixed, leaving only the audience to test.

3. If flexible ads in Series ② perform well, identify the top-performing, highest-spending ad by reviewing Facebook posts with comments (focus on those with most likes/comments). Again, use the original post to test interest audiences, lookalike audiences, etc.

Summary: This approach comprehensively covers core factors like creative testing, audience testing (interest keywords, lookalikes), original post performance, and bid caps—making it quite thorough. However, this approach assumes you already have creative assets ready in the first step. What if you lack strong assets in most cases? Then prepare another round: 5 videos + 5 images. The number of rounds depends on how much you value this product.

If you're testing mass-market products for bestsellers, you might select one product, test 10 assets in one round. If no assets perform well, move directly to the next product and repeat the process. If still no success, proceed to the next one.

For brand sites where product selection is fixed, it's an endless loop until quality assets emerge. Once found, proceed with scaling and expansion.

This approach prioritizes testing assets over audience interest due to their critical importance.

Feel free to critique or discuss if you disagree.

This method of filtering creative assets is indeed effective, but I have a question. What is the budget for your ad group to test these 10 creative assets using the 1-1-N method? Do you not set any interest keywords for the ad group? I usually set one key interest keyword and then let these 10 ad groups target the audience within that single interest keyword.

Elliot Price said:

This method of filtering creative assets is indeed effective, but I have a question. What is the budget for your ad group to test these 10 creative assets using the 1-1-N method? Do you not set any interest keywords for the ad group? I usually set one key interest keyword and then let these 10 ad groups target the audience within that single interest keyword.

Click to expand...

Budget: $30–$100. Take the winning ad ID from the 1-1-N setup to create a new 1-1-1 ad for testing interest keywords (or call it 1-n-1, where N represents N groups. The variable between groups is the interest keyword, while all ads use the same original ad post).

In this approach, prepare more videos and images for testing to find profitable creatives. Test different audiences and adjust bids based on performance. Focus on assets first before audience interest. Repeat process for different products until successful assets are found.

Hope it works good and for small budget too

This strategy is focused on prioritizing creativity over audience, leveraging original posts, and scaling with controlled bidding, making it ideal for finding and replicating winning ads. However, caution is needed to avoid discarding potential creatives due to insufficient testing or inappropriate formatting. For mass-market products, you can quickly discard those without seeing results; for established brands, this is a continuous cycle until quality creative emerges—the deciding factor in the overall outcome.

Your article has great potential, but sometimes cost optimization requires Meta to have simpler algorithms.

Looks solid, I’d just add rotating a few fresh creatives each week keeps ASC from fatiguing too fast.