Auto-Optimized Audiences: 3 Things You Need To Know

Agencies and brands can spend hundreds of billions of dollars advertising on platforms without ever defining their audience. They no longer have to select demographics such as gender, education, income, or marital status or descriptors like interests and behaviors.

All of the major platforms today make it possible to place a buy without ever specifically defining your audience. Instead, they all auto-optimize to your top-performing customers. 

Google has Optimized Audience in Google Ads for display, discovery and video campaigns. Meta has Lookalike Audiences, Custom Audiences (people who have shown interest in the product) and Special Ad Audience (people who have exhibited similar online behaviors as customers). And TikTok has Automatic Audience in Ads Manager. Instead of defining your audience you just click "auto-optimize."

With digital accounting for 63% of global advertising spend, per Statista, and these platforms representing more than 70% all digital ad spend, according to Ebiquity, nearly all of which can occur in the absence of a defined audience -- we’re looking at a complete paradigm shift in the media-planning process. 



Auto-optimized audiences is a good thing

Performance is better when we let the platforms auto-optimize. 

These AI-driven auto tools have made it much easier to buy media -- completely upending the media-planning and buying process. They are also starting to dramatically change the creative process.

The first step in any media-planning process was formerly defining your target audience. Agencies spent months clearly defining who the current customer was, and determining whether -- and how -- they were better or different than prospects. Lengthy decks examining all the behaviors, attitudes, demographics and media habits of every imaginable audience were prepared as part of the media-planning process.  

Actual media plans were never prepared until the clients confirmed the targets, and all too often media plans had to be revised as new information became available -- for example, if the target skewed a little older than originally thought. 

The consequences for planning the media incorrectly were huge -- or at least we thought they were. After all, we could control what target we bought against, but we couldn’t control if anyone actually saw an ad.

But now, using these AI-driven auto optimization tools that every platform makes easily available and even encourages us to use, media buyers don’t need to buy against any particular audience -- the platform does it for us. 

The time we spent building audiences is no longer necessary 

Kind of… The short answer is: It just doesn’t make sense to waste precious time (or agency fees) trying to anticipate the best audience when the platforms tell us. 

The long answer is: I don’t think agencies and clients have fully absorbed the consequences of our reliance on these AI tools. No one wants to admit that billions of dollars in media can be transacted without knowing the target. In some ways, it makes no sense (I understand that).

We’ve always operated with the assumption that the more we know, the better the outcomes.

But in a world driven by AI, that’s just not the case any longer. 

The media-buying process has already been fundamentally changed with easy-to-use buying platforms that allow for campaigns to be placed globally from one location, targeting the best-performing audiences without ever defining them.  

Creative has also been impacted by AI tools such as Facebook’s new Advantage+ Shopping program, dynamic creative optimization and SEO tools such as Jasper, and Frase that generate blogs, ads, social posts and headlines. 

Creative has also been impacted by AI tools like Facebook’s new Advantage+ Shopping program, dynamic creative optimization and SEO tools (e.g. Jasper,, Frase) that generate blogs, ads, social posts and headlines.

The role of the agency is quickly changing

With the rapid streamlining of the media-planning and buying process, one can only ask what the role of the agency will become.

Will we be whittled down to coming up with big ideas -- and if so, how much will clients pay for that service?

Or can we reimagine the role for human insight and creativity in a world where AI automates what used to be months-long processes in a manner of seconds?

1 comment about "Auto-Optimized Audiences: 3 Things You Need To Know".
Check to receive email when comments are posted.
  1. Jason Burke from All Stage, September 3, 2022 at 12:17 p.m.

    Any concern with any intentional bias built into the AI from the NASDAQ-housed platforms looking at the holistic picture of revenue?

Next story loading loading..