Brands spend big money running ads on Facebook, Youtube, and TikTok. They all say out loud that creative testing is of prime importance. Yet it’s strange to me that many teams I’ve spoken to don’t have a structured way of testing.
When I ask if they use a framework or a matrix to get to their amazing results, the answer is flatly no.
It’s true - ads are a blend of art and science. The art is the part that we feel emotionally and sensorily when we:
What about the science? Sure we measure the KPIs and results in platform reporting. Things like thumb-stop rate, CTR, conversion rate. But are the test cells set up in a way that allows us to confidently say “yes, for our brand..a video with the product demo works best, especially when paired with a testimonial”. Are the components of the ad cleanly defined and different enough to give us insights?
An ad starts with a good idea - but if you’re not using a framework to leverage that idea to get 4-5 different variations of an ad there’s low-hanging fruit that you can still pick.
Something paid search advertisers use in ad copy testing is an ad matrix. Matrices are used to permutate components together to generate a bunch of finished ads. Admittedly, sometimes the permutations result in weird AI-like results where the headlines don’t really pair well with the descriptions - that’s where the marketer has to apply a human touch to permutate things that fit with each other.
Let’s take a look at how we might apply this to facebook ads. I’ll list out a few dimensions that we can use as components in an ad.
Let’s run through an example of how we might pick a couple of dimensions to generate a matrix of ad variations.
We’ll pretend we’re Supergoop and we’re trying to market their skincare/sunscreen product.
We can position their Glowscreen SPF40 2 ways.
We end up with 12 different permutations for each positioning that we could create.
I know what you’re thinking. You’re saying “yeah yeah in an ideal ad we have all of these dimensions covered”. That’s true. I’d say you’re doing a great job already if you’re doing that. But sometimes maybe the 15 second ad feels way too short to stuff all of these messages in. Or we’re curious as to what opening message will retain viewership past the 3-second mark the most. In those scenarios, it’s helpful to have clarity of thought on which components to move to where in an ad, or to include altogether.
This matrix is also important to layout before you even go for a shoot day. Getting each of these components represented robustly in the shot list for your video production team ensures that you have more lego pieces to play with later when you’re chopping things together for testing.
Let’s finish our thought experiment and queue up 3 ads for Supergoop:
We run them in facebook or tiktok and see how they perform. Added bonus if we use the ad set a/b split test functionality for cleaner results.
Our learnings are:
As the brand owner, this pays off because this specific placement of components and their ordering allows you to pit specific variables (ie. A, B) within each component against each other and enables you to say something like - we know that lifestyle UGC works best to sell and it seems like our current audience is more interested in the hyaluronic acid + B5 than the wearability with makeup.
This is especially important early on when you don’t really have any data just yet on what works. The more robust your matrix is, the larger the net you cast for finding out what works.
When you identify that winning characteristic - you can double, triple down on it to mine that well over and over again. Generally a high performing ad nails the opening hook, the benefit, the demo, and the social proof. It’s why UGC ads perform so well - they often organically flow through this formula.
It’s also highly possible that the sunscreen ad starts to reach a unique segment of the population that was not interested in the moisturizer value prop - unlocking more scale and volume that the moisturizer ads could not.
Initially, this may seem like a heavier load for a design team - but it’s possible that a more structured and defined framework and an intent to shoot things for usability in evergreen campaigns can result in shot lists that are consistent (reducing need to reinvent the wheel each time). Shoots are expensive to arrange and the more you can batch into a shoot the more you have to work with. If we include the right mix of shots in that batch, we come out with more optionality.
It can feel like there’s only one way to sell or message your product. It’s important to have a beginner’s mind and fresh eyes to unlock other ways the product can be valuable for a user. This framework helps facilitate that.