What to Do With Creative Testing Results
Today's question is about determining what works from top performing ads to inform the next batch of creative. Creative testing used to isolate single elements, but now ads generate thousands of combinations through text options, placements, and enhancements. Jon explains why finding one winning combination isn't the goal anymore, how to use the creative testing tool to identify themes instead of specific winners, and why success comes from many winning combinations in aggregate.
Want your question to be answered on a future episode? Go to JonLoomer.com/Question and record your question today.
Question
Hey Jon. This is Chelsey Abbott and I have a Pubcast Question about testing creative. I'm hoping you can provide some insight on the best ways to determine what parts of the copy and content work from our top performing ads in a campaign, and then how we can use that knowledge when creating the next batch of copy and content. Thank you.
Answer
This is a really good question, Chelsey, especially considering how much ad creation and creative testing have changed the past few years.
Back in the day, creative testing was pretty straightforward. We’d create an ad with one primary text, headline, and image or video. And then we might test it against a similar ad with one major difference. This way we could isolate performance changes related to your primary text, headline, or creative.
But ad creation is way different now.
When we create ads, we provide up to five primary text and headline options. We customize creative by placement. And we may accept enhancements like Advantage+ Creative, AI-generated images, backgrounds, or even videos now, AI-generated text options, and Related Media from other ads.
The result is that a single ad could result in thousands and thousands of copy and creative combinations. This is by design.
In this new era of creative diversification, our goal isn’t to find a single winning primary text, headline, creative, or combination. Our goal should be to create many winning combinations that can inspire different people to respond to our ads. And we will know we’re successful when we’re getting the performance we want in aggregate.
But your question is valid because it’s still helpful to know what’s working and what’s not.
Use the creative testing tool in your active ad set to test two to five ads at a time. Continue to create your ads as you normally would by generating lots of copy and creative options.
But when you create those five primary text variations, make sure there are similarities by style or customer persona or angle. Because when you have five options for an ad, those variations will water down your hope to find a single winner.
So when you run your test, you’re not looking for necessarily the best primary text option, even if you use the breakdown options by text. But you can get a sense of the themes that worked best.
This isn’t so you can turn off the other ads or run ads that only look and sound like that one going forward. It’s to provide a general starting point for the types of language you’ve seen work when you create that next batch of ads.
The main thing I’d say is I wouldn’t obsess over finding winning creative like we may have in the past. That’s just not how things work now since an effective ad set will rely on many winning combinations.
But you can absolutely learn from your testing results. Instead of isolating specific text and creative, learn from the general themes and approaches that seemed to work best and apply that to your next ads.
Thanks for your question, Chelsey!