Modular design: the secret to better ad creative at scale

What is modular design?

Modular design is a design approach in which placeholders within a template hold space for creative elements to live interchangeably. In more casual terms, modular design lets you mix and match creative elements to quickly create as many combinations as possible.

Modular design in advertising

Modular design is a relatively new approach to designing ad creative. In traditional ad design, ads are created as one-offs; the headline and the image only make sense when paired together. But with modular ad design, multiple creative assets make sense as a unit — no matter how they're combined.

An ad starts as a blank shape, typically a square in an image editing software. Creatives add in product shots, brand images, prices, promotions, messages, and more — each addition can produce infinite variations.

For example, the call to action "Buy now" can be replaced with "Buy today", "Shop now", or "Try today" without losing its meaning.

The results of designing in this way add up fast.

Let's say your creative team comes up with three headlines, four images, and two calls to action. To figure out the total number of combinations these elements create, we multiply them together. 3 x 4 x 2 = 24 ad variations. The greater the number of elements (and the greater the number of variations of those elements) the more ads you can build.

To zoom out on this idea a bit further, think about a multivariate matrix.

A matrix is built by altering one creative asset per ad. The top left ad has a centered product and bottom-centered, black, sans-serif text on a white background. The ad immediately below it is identical other than the serif font. The ad below that is identical to the first, but with white text on a black background.

The process continues until every possible combination has been formed. Designing ad creative in this way not only enables creative teams to scale ad production. It also allows advertisers to templatize and systematically test their ads.

Why modular design is the design approach for multivariate ad testing

Modular design is super important when it comes to testing ad creative — specifically multivariate testing ad creative.

Multivariate testing measures the performance and conversion rates of every possible combination of creative elements. Because we can measure how every variable works with every other variable, we can get some seriously meaningful results. Not only are we able to understand which ads people love and hate the most, but also exactly which variables people love or hate the most.

Let's say, in our example above, the fifth ad from the left in the top row turned out to be our winning ad. Let's say it outperformed the ad to the right of it by 350%. This would tell us that our audience prefers copy on the left in the white background and product on the right. And this is something we could take forward and try again in subsequent tests.

This is a real scenario we've seen before with our customers. It's the reason running many ads in a multivariate test is so important: each and every little permutation can impact performance in a big way.

Modular design best practices

Modular design requires a totally different process from the traditional approach to concepting ad creative. So it can take some practice to master. Here are a few tips for getting started.

• Categorize everything. Organizing your assets into categories can help you determine which ones can be swapped in for one another easily. It's also helpful for analyzing testing results — seeing, for instance, that "product shots with models" outperformed "product shots without models".
• Say the same thing multiple ways. Come up with one headline for your ad design, then see how many other ways you can say it. For example, "Now in all-new colors" can be easily swapped with "Hot new colors for fall" and "Find your new favorite shade." Each can be tested to find out which version performs best.
• Don't let your assumptions get in your way. Bias is a huge reason ads fail so often. People's preferences or opinions are often used to make important design decisions. But only multivariate testing can tell you which creative elements work and which don't.