You got some answers through the A/B test, implemented them in your ads, and saw positive results. Now what? How do you recognize the right moment to put your creatives on another test?
Usually, when ad fatigue hits and results are missing marketers address this issue with fresh creatives, but the new question arises - what is the right way to use these creatives, should you insert them in the ongoing campaign, or create a completely new one losing the learning existing campaign had?
There’s a way to prevent this from happening: building a creative testing framework.
The idea behind building a framework is integrating standard creative tests into your ongoing operations. This approach allows for consistently testing both new and old creatives, ensuring continuous optimization of your campaigns.
In this article, we’ll explain the logic behind testing frameworks, and give you examples of how it can be done.
Why should you build a testing framework?
In a nutshell, a creative testing framework aims to evaluate the effectiveness of new creatives and, over time, measure their performance against established creatives.
The framework provides structure to the testing process, helping you make informed decisions about what creatives should be scaled up or reworked.
As with any strategy you’re implementing, the essential part is establishing a cost-effective testing system tailored to your specific business needs & capabilities. This means determining how tests will be designed, implemented, and analyzed, considering factors such as test reliability and budget constraints.
How to build a testing framework?
Let’s take a step back and answer a question how often should you test your creatives in the first place?
We asked this question Gulce Ozmen, an expert in this field, and she emphasized that is crucial to identify your internal capabilities and establish a testing cadence that’s sustainable for your team.
The structure and complexity of your testing framework will vary depending on the volume of creatives your business produces.
For example, a brand that generates ten new creatives weekly will require a different framework than one producing only two or three creatives a month. More frequent testing will demand a more robust framework with clear processes for managing data and insights, while less frequent creative production might call for a more streamlined approach.
Regardless of the scale, the goal remains the same: establish a system where testing is an integral, ongoing part of creative development.
The question of testing new vs. established creatives on Meta
One common challenge we see our customers face is the consistent outperformance of established creatives in tests on Meta.
This is due to the learning advantage older creatives have over new ones. Established creatives have accumulated valuable historical data that inform optimization algorithms, giving them a performance edge.
When testing new creatives against established ones, it’s crucial to consider the impact of ad auctions. New creatives, with no prior learning data, may struggle initially in comparison to seasoned creatives. This doesn’t mean new creatives are ineffective but highlights the importance of factoring in the auction dynamics during testing.
eCommerce brand example of creative testing framework
We’ll use Meta’s example of a creative testing framework, adjusted for the eCommerce brand. Note that the purpose of phase 2 is to balance the learning history between new creatives and established ones. Excluding this phase can result in a positive bias towards the control creative in phase 3.
Phase 1: Create New Ads
- What you do: Come up with fresh ideas for new ads based on what’s trending, or what’s coming up in your eCommerce calendar (e.g., Black Friday or new product drop).
- Action: Develop 3–5 new creatives, like a product showcase video, influencer testimonial, or carousel highlighting your best sellers.
Phase 2: Build Early Learnings
- What you do: Let the new ads run in a fresh campaign for about 7 days. Facebook’s algorithm will automatically rank the ads based on early performance.
- Action: Watch which creatives start gaining more traction. This is just to get them some early visibility so they can compete with your existing top ads later.
Phase 3: Test New vs. Old Ads
- What you do: Run an A/B test to see if your new ads can outperform your best-performing old ones.
- Action: Set up a 7-day A/B test where new ads go up against your top-performing ads from previous campaigns. Look at key metrics like purchase rates and ROAS.
Phase 4: Scale the Winners
- What you do: Take the winning creative and add it to your main campaigns.
- Action: Add the best-performing ad to your live campaigns. Keep an eye on performance and tweak as needed. Don’t overwhelm your ad set with too many creatives.
How does Hunch help in building a creative testing framework?
Let’s start with the most obvious answer: creative testing is invaluable for your ad performance, but it brings with itself an enormous creative workload. So, scaling different creative variations using Hunch’s AI tools is something that will save you time, letting you focus on strategy.
But what if you need help there? We can assist in this part of the process.
If you’d like to take a step further and execute incremental testing, we can connect with your Meta rep and help you execute it right.
But let’s turn words into action: see the results of the Conversion Lift Study we did for Academy Sports, which drove an incremental ROAS lift of 2.3x.
