Contact us

Talking Too Small-The Risk Of Over-Optimising Audiences3 min read

3 min read

Since as long as advertising has existed, so have advertising budgets. Every marketer has had the imperative to spend their budgets in ways that would have the greatest impact. Marketers have often done this by trying to advertise first to the people most likely to be influenced by an ad.


If you’re in digital marketing, you’re probably doing this with audience targeting. Modern marketers have unprecedented abilities to target very, very specific groups of people. When used wisely, this can drive enormous improvements to a campaign’s effectiveness, whether the goal is awareness lift, brand perception, or hard conversions. Used unwisely, however, over-targeted marketing poses risks to a brand’s long-term health.

Taken to an extreme, the audience most likely to respond – ready to convert immediately, absolutely insensitive to price, and with high LTV – maybe just a few people. If you could target them with just a few dollars, the campaign return would be off the charts!

Unfortunately, for all but a precious few companies, it takes more than a couple die-hard fans to make up a market. Targeting just those precious few may pay big dividends in the short-run, but isn’t a long-term strategy. Highly-segmented digital audiences can be less-extreme versions of the same problem.

The primary problem with talking to just a small group of people is that your pool of customers is very limited. You may get great results at first, but frequency will build quickly and creative fatigue will hit early. Revamping creative may get another group into the funnel, but repeated refreshes will hit diminishing returns.

Even identifying an audience that works now isn’t the same as planning for future growth. A middle-aged demographic may respond at lower rates because they already have your competitors’ products, but will need to buy again in the future. An 18-24 male in March and an 18-24 male shopping for a girlfriend’s Christmas present are likely to have very different purchase behaviors. Failing to adapt your marketing to changing behavior will put a cap on results.

Relying too heavily on audience data may even guide the business into a false sense of market/product fit. I once worked on a large household brand who found their media resonated extremely well with people in the 70+ demographic. The brand management team was very excited and wanted to shift as much money as possible into that demo. I couldn’t understand the excitement or the shift in money. Where was their brand going to be in 10 years?

Finally, over-optimizing audiences can deprive algorithms and agency teams of the data volumes needed to optimize for wider performance. Tiny, highly-specific audiences can drive enough conversions to make up an acceptable volume in some cases, but if each one contributes just a trickle of clicks and fewer on-site actions, it becomes much more difficult to test imagery, messaging, landing page strategy, and other important campaign components.

Imagine what would happen if Coke launched today and, instead of trying to have Coke in reach of anyone who wanted one, tried to get a Coke ad near 24-34-year-old females who like soft drinks on Facebook and had recently visited their website.  Their clickthrough rates would probably be great, but would they be on the path to being a global brand? Probably not.

A brand exists in customers’ minds. To build a brand, you have get in front of those customers. You have to give people a reason to buy. Just getting your product in front of people who are ready to buy it right now can be a quick win, but it’s just a quick win, not a long term strategy.

Your brand may not have the ambition to be the next Coke and you still have to make sure you put budget where it will go the furthest. Just be careful how tightly “furthest” is defined, or you may wind up talking to a very small group of people.