Testing advertising creative has been on the top of my mind as of late. I can’t understate the importance of running good creative as advertising spaces become more crowded. Good creative, as it always has, will be what cuts through the noise on these platforms as they become increasingly crowded.
In saying that, testing your creative should become an important part of your campaign implementation and optimization.
But what’s a good process, and where to begin?
Today I want to discuss a scaleable process by which you can consistently be testing your creative to increase your return on ad spend (ROAS) and gain valuable insights into the mind of your target market.
There is a flow to the process, one that isn’t clunky and one that can gain you truly valuable insights into the mind of your target market. However, to begin, let’s first discuss some important concepts.
What is A/B Testing?
A/B testing has been a concept since the 1920s and is one of the best ways to get unbiased information from consumers. To simplify, A/B testing is the process of running two different pieces of creative side by side to the same target audience for the purpose of insight discovery. At Nexus, I often refer to this as “red shirt, blue shirt”.
Us marketers tend to get deep into the weeds and begin missing the bigger picture of what we’re trying to actually accomplish. That is, to gain valuable insights into what creative elements work best. So I often find it useful to circle back to the idea of “red shirt, blue shirt” to ensure we’re testing the right elements to determine the right outcomes. It becomes very easy to understand what we’re testing for and what we want to learn when you ask the question “what’s the red shirt, blue shirt?”.
Problem Definition
Most clients I speak to often understand the importance of running multiple creatives at the same time. However, they don’t quite get why they’re not gaining the insights they truly desire. For example, can you spot the difference between these two photos?
Mechanically they’re both your average Instagram 1080 x 1080. Thematically they’re both dads reading to their kids. Visually they both have books, exactly one parent and one child. So what’s the red shirt, blue shirt between these photos? What are we testing for? What kind of insights could we glean from this sort of A/B test?
The problem here is that there is no proper definition of what we want to learn, and therefore, no proper direction on creative implementation. Without knowing what we want the output to provide, there is little thought about the input.
So, where to begin?
Define What You Want To Learn First
“Begin at the beginning,” says Lewis Carol in his book Alice in Wonderland, “and go on till you come to the end: then stop.” In most cases, this is solid, if not deeply philosophical advice. However, when it comes to advertising creative it’s best to think of the end first and work our way backward.
Advertising copy and creative was often overanalyzed and processed by many stakeholders in a campaign back in the days of traditional advertising. While the dawn of digital advertising introduced a need for speed it also introduced an opportunity for fast testing on a large scale. If you find yourself debating between two unique selling points (USPs) or two slam-dunk benefits, you now have the ability to run both of them at the same time and let the customers decide for you.
In the case above, we’ve defined that we want the audience to decide which USP works best. This is an example of thinking of the output first and working from there.
So what are some creative components we can analyze?
Deconstructing The Creative Components
There are three general aspects that you should be looking analyze:
Mechanical: Creative elements such as the aspect ratio of the asset, the potential playtime of a video, and average color, and resolution. These are all descriptions of your asset that have one true value.
Visual: The presence of physical objects such as people, animals, or cars, or if the setting is indoors, outdoors, or the use of overlaid text as an image. These are concepts that describe the visual components of an ad.
Thematic: The theme of an ad is the highest-level description of a creative campaign. Creative assets can be designed to evoke a particular emotion from the viewer. They can also be designed with different production choices in mind: smooth or choppy, fast-paced or slow, and methodical.
Now that we have our creative components defined, we can begin the process of testing different elements of the creative and discovering insights about our target customer.
The Creative Testing Process - 4-Easy Steps
Step 1: Define
In the first step, as we’ve discussed above, you define what you want to learn.
Step 2: Make
We then make the creative, deconstructing the creative to its component elements and testing them one at a time.
Step 3: Learn
The most important step here is learning. What KPIs are we looking at to determine insight? Common data we could use to determine better performing creative (depending on your goal) is click-through-rate, total conversions, cost-per-conversion, conversion-rate or return-on-ad-spend (ROAS) just to name a few.
Step 4: Adapt
Now that you’ve gleaned some creative insights, it’s time to adapt that creative. This could be anything from taking the winning theme of two images and incorporating it into a video. Or it could be choosing an image that better evokes that theme. Or you could decide that one theme works better for one age group than another age group and begin to clearly define what emotions each age group wants from your messaging. The point is that you now have a direction to adapt and improve your creative.
How often should you change your advertising creative?
While there is no single answer to this question, platforms like Facebook suggest when your ad reaches a frequency of 2-3, or if an ad has been in-market for more than 2-weeks, then it might be time to freshen it up. However, I often balance reach, frequency, time-in-market, and ultimately results to determine when it’s time to change creative.
With platforms like Google Search and Display, the analysis becomes a little more difficult as in some cases they don’t give you frequency numbers. They also use responsive creative which takes many different creative assets and uses machine learning to deliver the best ad to the end-user.
Championing Your Creative
There is one last point I want to make here and that is something I call the “battle royale” but I’ve often heard it called “championing your creative”.
This is the process I used for periodically changing my creative and ensuring you’re running your best-performing ads while maintaining a flow to the campaign and your processes.
I always begin a campaign with a minimum of two pieces, testing specific creative components, of course. Then every 2-or-so weeks, I keep my best performing creative in-market (the champion) and switch out the creative that is not performing as well (the loser). Then I implement a new piece of creative that either tests a similar creative component or a different creative component (the challenger).
This process will create a flow of analysis within your department, creates a great dialogue with clients, and gets a boost from platform algorithms by keeping creative fresh for the end-user.