Companies ask me all the time if I can help them reach their target audience.
Usually, companies and business owners think they know who their target audience is, based on the key assumptions that were made before they launched their product.
The logic makes sense —
“We built and launched this product for a particular use case with a particular user in mind. We know this user’s interests, behaviors, and personality types."
As a result, most companies are pretty confused when we tell them there’s no such thing as a target audience, and that they’re asking the wrong questions. It goes against everything that the MBA programs and marketing playbooks preach.
To grow, companies put together marketing material and invest a lot of advertising dollars to show it to the people they think will use it.
They’re searching for product-audience fit but they hardly ever stop to think about audience-creative fit.
Before launching Marpipe, I ran a performance marketing agency where the majority of what we did was attempt to find the perfect audience to target for companies. If one audience didn’t perform well, we moved on to test another, slightly different, one.
The worst part is, most audiences didn’t work, and the ones that did only lasted for a few days. We kept going through this frustrating and expensive cycle, with seemingly no end in sight.
Over time, we noticed something:
Changing the creative while holding the audience constant had a significantly higher impact on performance than changing the audience while holding the creative constant.
This is confirmed by data from Nielsen, wherein creative is by far the single largest influence on Return on Ad Spend (ROAS).
Even small, nuanced changes to visual variables, such as color and object positioning, tend to have large impacts on performance.
Beyond that, we found that every individual component of the creative has a streamlined and measurable influence on overall performance.
Even further, we found that, when you combine the most effective visual variables in a single ad, you can reliably and predictably achieve performance breakthroughs.
Once you understand exactly what an audience prefers to see, selling to them becomes much more efficient.
If this insight is true and reproducible, then audience targeting begins to play a very secondary role.
My team and I have been demonstrating this to be true and reproducible. Here’s an example: After running a multivariate test with hundreds of creative variants, we discovered that men who work in finance have a large preference for seeing videos in black-and-white.
We’ve discovered this to be true across multiple channels — we don’t know why it’s true, but we know it to be true in almost every instance.
When you change the product/brand being advertised, the ads in black-and-white still have a large positive-outlier impact on performance. This piece of creative data is valuable to any brand that wants to sell to men in finance, regardless of what they’re selling.
That’s when we decided to build a tool to batch-render software to hundreds, even thousands, of creative variants for brands with a simple one to two hour internal setup time. As a result, it’s now possible for us to test hundreds of creative variants in a few days and discover what people like to see the most.
When done effectively, this methodology creates a new approach to marketing problems and opens a new category in the space of optimization. But whether you automate this process or not, one thing is for sure:
You can sell almost anything to anyone with the right creative. It’s rarely about the audience, but always about the message.