OpinionStory

Don't split test the first page

Mark Builds Brands

Facebook penalizes split testing on landing pages after ad approval because it views post-approval changes as potential cloaking or non-compliance. The speaker demonstrates this through a failed listicle test that only succeeded when retested using new post IDs instead of modifying the original approved landing page.

Summary

The speaker explains why split testing should not be conducted on the first landing page after Facebook ad approval. Facebook continuously scans landing pages, and the platform has strong negative reactions to changes made after an ad has been approved. Facebook interprets post-approval changes in two problematic ways: either as cloaking attempts (trying to circumvent policies by getting approval then changing to non-compliant content), or as a softer rejection of the modified content. The speaker illustrates this with a personal case study involving a listicle that underwent a 50/50 split test with copy changes (different headline, sub-headline, and lead). The control variant declined in performance while the variation performed even worse, dropping by approximately 50%. After pausing the poor-performing test, the speaker reconsidered the variation's potential and retested it using the correct methodology: maintaining all the same winning creatives but using new post IDs in either a new ad set, new ads, or entirely new campaign. This approach yielded dramatically different results—the listicle transformed from an unprofitable page to a top performer that not only beat the control but replaced it entirely, accumulating over $400,000 in spend over 60 days and becoming one of the fastest-scaling funnels in the speaker's advertising career.

Key Insights

  • Facebook views post-approval landing page changes as either cloaking attempts or soft rejections, which can harm ad performance
  • A listicle split test with copy changes resulted in the control declining and the variation performing 50% worse than expected
  • The speaker retested the poor-performing variation using new post IDs rather than modifying the original approved page
  • Using new post IDs converted the listicle from an unprofitable page to a top-performing variant that accumulated over $400,000 in spend within 60 days
  • The correct split testing approach after approval involves creating new post IDs across new ad sets, ads, or campaigns while keeping all other winning elements constant

Topics

Facebook ad approval and post-approval changesSplit testing methodology and best practicesCloaking and platform compliance policiesLanding page optimizationAd account scaling strategy

Transcript

[0:00] Don't do split testing on the first page. The reason you don't split test on the first page is because Facebook is always scanning your landers. And the number one thing that Facebook hates the most is people changing [ __ ] after it's already been approved. So, your ad has been approved, and then you go change the lander after that. Facebook sometimes can view this in one of two ways. The bad way is they view this as cloaking. They think that you're trying to circumvent their policies by getting an ad approved and then changing it to something non-compliant. The second thing is more of like a soft rejection. The best way that I found this…

Full transcript available for MurmurCast members

Sign Up to Access

More from Mark Builds Brands

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.