This is such a common mistake 🤦‍♂️

Key Concepts

• Learning Phase: The initial period when Facebook’s algorithm is gathering data on an ad’s performance to optimize delivery.
• Ad Reset: The loss of accumulated data (likes, shares, comments, conversion data) when an ad is significantly edited, forcing it back into the learning phase.
• Data-Driven Decision Making: Relying on performance metrics rather than subjective opinions when managing ad campaigns.
• Parallel Testing: Running multiple versions of an ad simultaneously to compare performance.

The Pitfalls of Editing Live, High-Performing Ads

The core argument presented is a strong recommendation against editing ads that are already performing well, particularly those with significant spend and conversions. The speaker emphasizes that even seemingly minor changes can be detrimental to an ad’s performance by resetting its accumulated data and forcing it back into Facebook’s learning phase.

This “ad reset” means the ad loses all previously gathered signals – likes, shares, comments, and crucially, conversion data. Facebook’s algorithm uses this data to optimize ad delivery to the most receptive audience. By editing, you essentially erase this learning and begin the optimization process anew.

The Learning Phase & Algorithm Impact

The speaker directly links ad editing to the disruption of the learning phase. Facebook’s algorithm requires a period of data collection to identify the optimal audience and delivery strategies for an ad. Interrupting this phase through edits negates the progress already made. A detailed explanation of the learning phase and the Facebook algorithm is available in the speaker’s course (link provided in the transcript).

Parallel Testing as a Solution

Instead of directly editing a successful ad, the speaker advocates for a strategy of parallel testing. This involves duplicating the existing ad and making the desired changes to the duplicate. Both versions then run simultaneously, allowing for a direct comparison of performance.

The speaker states, “I’m happy to duplicate the ad and run it in parallel in hopes that this edited, changed, better ad, right, does better than the original.” However, they add a crucial observation: “Nine times out of 10, it doesn’t.” This highlights the often-counterintuitive reality that the element a marketer believes needs improvement might actually be a key driver of the ad’s success.

Removing Subjectivity & Data Reliance

A central theme is the importance of removing subjective opinions from ad management. The speaker urges viewers to “Remove your subjectivity. Let your data do the talking.” This emphasizes a data-driven approach, where decisions are based on observed performance metrics rather than assumptions or aesthetic preferences. The example of a typo or branding inconsistency is used to illustrate this point – even perceived flaws shouldn’t trigger an edit if the ad is demonstrably effective.

Real-World Application & Case Study (Implied)

While no specific case study is detailed, the entire argument is built on the speaker’s experience managing ad campaigns. The “nine times out of 10” statistic suggests a pattern observed across numerous campaigns, implying a strong correlation between ad editing and performance decline. The scenario presented – a brand requesting an edit on a successful ad – is a common occurrence in agency-client relationships, making the advice highly relevant to digital marketers.

Promotional Information

The speaker concludes by mentioning a Fourth of July promotion offering $400 off their course, which contains a more in-depth explanation of the learning phase and Facebook algorithm. This course is accessible via a link on their profile.

Conclusion

The primary takeaway is a cautionary one: resist the urge to edit live, high-performing ads. The potential disruption to the learning phase and the loss of accumulated data outweigh the perceived benefits of minor improvements. Prioritize parallel testing and data-driven decision-making to optimize ad campaigns effectively.