Audience Building as an Art Form

As a marketing agency, we all acknowledge the lift we get with better creative. Graphic artists, content creators and video producers/editors can have a significant impact on the performance of any campaign. Every agency puts a distinct focus on it, and it makes sense. We have people on our team dedicated to that effort specifically and they are all held to a very high standard. But we put as much, if not more focus, on another art form. An equally creative and important piece to the puzzle that most agencies have given up on air simply don’t have the tech, in-house, to effectively leverage. Audience building.

I know, I know, every agency claims they do as well, but I am here to tell you, they do NOT. Clicking boxes for targeting selects on a platform and toggling a switch to turn on AI is NOT audience building. In fact, it’s not very smart. Those moves aren’t moves at all; they’re an acceptance in taking what the platform gods will give you. And that is normally average at best, and complete trash most often. You know, you’ve seen it. For more detail on that, you can read previous articles I’ve published on our site and the blog. There is plenty of content that describes how completely untenable platform audiences truly are and why it is not wise to use them. In this moment, let’s focus on the creative use and application data has in the building of audiences and why we consider it an art form.

It has been my contention that building audiences based on demographics alongside is both weak and intellectually lazy.

The failure to incorporate the word “and” into your data pretends that simply being able to afford a “thing” makes you likely to buy a “thing”. Or that simply having a situation present in your life makes you likely to buy something.

An example I love to use to describe the fallacy in this thinking is the automotive aftermarket, specifically for performance products. Millions of people buy sports cars every single year. Tens of millions already own one. If you make an awesome performance exhaust, most marketers consider targeting Corvette owners (for example), even for an exhaust system designed for Corvettes, as a great way to target and build an audience. Platforms tell you that the act of following certain pages is indicative of a likely buyer. Here is where we part company. We don’t think that is nearly enough. If the goal of the ad spend is to drive sales of a performance exhaust, that audience will absolutely waste the lion’s share of the budget. Period. Why?

Because the data tells us that’s the case. There are a ton of Corvette owners that will NEVER buy performance parts at all, much less a performance exhaust. Here is a list of reasons for reference.

• Originalists will keep the car stock in every way on purpose
• Less experienced buyers either feel the car goes fast enough or is too fast
• Many don’t care about such things
• Many don’t want the sound to change at all
• Many are satisfied with the stock exhaust
• Many don’t see the purchase in hobby context at all

There are many more reasons but it’s easy to see those demographics and simply having something present in life does not, in any way, make someone a likely buyer. That audiences advantage stops at “they own the thing the product was made for”. For us, that’s not nearly good enough. Using that data for an audience build will not only drive ROI down but will negatively impact on a campaign in many other ways as well. Consider these things:

• A/B testing that includes audiences that will never buy misinforms the campaigns on what plays well with people likely to buy
• Vanity metrics get skewed by audiences that have zero interest in a product. Is the reach metric relevant if ½ the audience would never buy the product?
• CTRs are a key metric we use to understand how audiences are responding to offers, messaging and content we are showing them. Again, how accurately does a metric like CTR inform the campaign if a significant part of the audience is completely uninterested?
• Frequency is a major metric we assess. It is one of the most important variable in campaigns that speaks most directly to ROI. Knowing if it takes 11 or 100 impressions to drive a conversion speaks directly to strategy shifts necessary to drive profit. How the hell can we tell where that line is if the audience is comprised of people that wouldn’t convert at 1,000 impressions?

Obviously, this is a very basic example I’m using to make understanding the importance of audiences easy to see but apply this logic to more complicated audience buildouts. With “and” in your audience, you are really shooting in the dark.

Let’s talk about “and”. Baseline demographics are foundational, they just aren’t enough. What we require is the presence of these AND other variables. For example, an audience we would build would use Corvette ownership and likely income as well but would also require the “and. Corvette owner AND illuminating intent. Corvette owner AND actively consuming content about performance exhausts. Corvette owner AND frequenting Corvette themed car shows. These are just a few examples of how we add selects today in a compound approach vs an isolated approach. This one act requires some severely different data capabilities AND we have them!