Facebook Advertising - How To Split Test Interests At The Ad Set Level in Facebook's Power Editor

TL;DR

Split-test Facebook interests at the ad-set level by duplicating the ad set, isolating one audience segment, excluding overlapping interests, and comparing customer-acquisition costs. Miles Becker demonstrates the method in Facebook’s Power Editor using a Tony Robbins interest audience of about 2.5 million people, while warning that duplication also duplicates the daily budget. Read on for the exact setup, budget precautions, and testing criteria.

Transcript

a Miles Becker here Miles Becker calm and this is another video about split testing in Facebook specifically we're going to look at how to split test some of the demographic psychographic and interests inside of the like at the ad set level excuse me inside of the power editor in the Facebook pay-per-click advertising system so at this point in the... Read More

Key Insights

• Split testing in Facebook ads allows marketers to determine which segments of their audience are most effective, helping to optimize advertising budgets and efforts.
• The process involves testing demographic, psychographic, and interest segments to find the lowest cost per customer acquisition.
• Duplicating ad sets in Facebook's Power Editor can help compare different audience segments without overlap, ensuring clear data results.
• It's crucial to test one variable at a time, such as interests or gender, to accurately determine what impacts ad performance.
• Adjusting daily budgets for split tests is important to maintain control over advertising spend while gathering useful data.
• Excluding overlapping audiences in ad sets ensures that each segment is tested independently, providing more reliable results.
• Running split tests for at least two weeks and gathering 500 to 1,000 actions ensures statistically reliable data for decision-making.
• Advanced split testing techniques can help scale successful ad campaigns by identifying and investing in the most interested audience segments.

Questions & Answers

Q: How do you split-test interests at the ad-set level in Facebook’s Power Editor?

Select the existing ad set, click Edit, and duplicate it within the same campaign. In the duplicate, retain the interest being tested, remove the other interests, and exclude those removed audiences to prevent overlap. Then compare the ad sets using cost per customer acquisition.

Q: Why should you duplicate the Facebook ad set for an interest split test?

Duplicating the ad set creates a separate audience test while preserving the campaign and its advertisement. The transcript notes that the advertisement is duplicated automatically, allowing the same working message to be tested across distinct audience segments.

Q: How do you prevent audience overlap between the ad sets?

Use the exclude section to exclude people who match the interests assigned to the other segment. Miles Becker emphasizes this because audiences can overlap, and without exclusions the ad sets could compete with each other.

Q: What interest audience does Miles Becker use in the example?

He isolates the Tony Robbins interest segment, which he describes as an audience of about 2.5 million people. He compares that segment with the other interests to see which group purchases at a lower advertising cost.

Q: What metric should determine which interest segment performs best?

Compare the cost per customer acquisition for each segment. This is the amount spent on advertising to generate one customer, and the goal is to identify the segment that produces customers at the lowest cost.

Q: What happens to the daily budget when an ad set is duplicated?

The daily budget is duplicated along with the ad set, so the campaign can initially spend twice as much. Adjust the budgets so the two ad sets combined equal the amount you intended to spend.

Q: Which targeting variables can be split-tested at the ad-set level?

The tutorial discusses demographic, psychographic, gender, and interest-based segments. For gender, separate ad sets could target men and women, while the demonstrated test focuses on interests.

Q: How long should the split test run, and how much data should it collect?

The existing guidance recommends running split tests for at least two weeks. It also recommends collecting 500 to 1,000 actions or conversions before using the results for decisions.

Summary & Key Takeaways

• This video by Miles Becker focuses on advanced split testing techniques for Facebook ad campaigns using the Power Editor. It emphasizes the importance of testing demographic and psychographic segments to optimize advertising efforts and costs.

• The tutorial demonstrates how to duplicate ad sets to test different audience segments, such as interests and gender, without overlap. It stresses the importance of testing one variable at a time to gather accurate data.

• Miles advises running split tests for two weeks and achieving 500 to 1,000 actions for statistical reliability. Adjusting daily budgets is crucial to control spending while optimizing ad performance through data-driven insights.