In this article, we'll explain what mistakes to avoid and best practices to follow to create a campaign comparison conversion lift test likely to produce meaningful results.

Choosing the right campaigns

Ensure that the campaigns you choose to test:

• Will be active for the entire time your test is running. If you need to edit a campaign schedule to accommodate a test schedule, do that before you start your test.
• Have only one major difference between them. For example, they might use different objectives or have different target audiences. This helps make your results as meaningful as possible. If your campaigns have multiple major differences, we may still be able to tell you which caused lower cost conversions, but you won't be able to know what caused the winner to have a lower cost. If all you want to do is decide which of two campaigns to put more budget into, testing with multiple major differences may work for you. However, if you're testing to help guide your overall advertising strategy, limit yourself to one major difference per test. After a few tests, you'll have a lot of rich information to draw on as you create new campaigns.
• Haven't been running recently. This can mean creating new campaigns to test, or reactivating old ones. We recommend this because campaigns that have been running for a while may be less effective. If they're almost at the end of their runtime anyway, results of the test may not be very useful. Plus, if you test a new campaign against an old one, they may not have equal potential effectiveness. This decreases the meaningfulness of results.
• Aren't being used in another Test and Learn test. To maintain the scientific rigour required of lift tests, we can't have one campaign active in multiple at the same time.

Setting an effective schedule

Ensure that the schedule you set is:

• At least seven days. However, we recommend running a test for at least a month to ensure that it's powerful enough. Think of seven days as a minimum.
• Aligned with the schedules of your selected campaigns. If you don't want to adjust your campaign's schedules to accommodate your test, adjust your test's schedule to accommodate your campaigns.

Choosing the right conversion events

Before choosing your conversion events, ensure that the sources of the events you want to use are fully implemented and functioning properly. If they aren't, we have troubleshooting guides for each source:

• Facebook pixel
• SDK
• Offline events

Next, think carefully about which events you want to measure. We recommend considering both:

• Which events are most meaningful to your business
• Which events will lead to an adequately powerful test

Ideally, the events that are most meaningful are the same ones that will lead to a powerful test. However, this may not always be the case. For example, if what you care about is purchase conversions on your website, but you're not yet getting many of them, it may be a better idea to test more frequent website conversion events, such as adds-to-basket or page views. While this may not be ideal, it can help you understand what strategy is causing more events that are prerequisites to making a purchase. Once you learn that from your test, you can scale that strategy to drive more people into your overall marketing funnel, which can result in more purchases. Then, when you're getting enough purchases, you can focus your testing on figuring out how to get more purchases more efficiently. While it's tempting to jump straight to that, if you're not getting enough purchases, we won't be able to produce meaningful results.

Learn how to interpret and troubleshoot campaign comparison conversion lift test results with guide.

* Nguồn: Facebook