Worth Reading: The Case for Just-Enough Targeting

They are still voices in the wilderness, but a few people are starting to question the digital ad industry’s obsessiveness when it comes to ever more granular targeting. In Digiday recently, Simulmedia CEO and behavioral targeting pioneer Dave Morgan estimated 90 percent of targeting isn’t worth it. Upstream Group CEO Doug Weaver, writing on his company blog The Drift, picks up on the theme with the argument that the online industry has grown up with a problem: it has long been in the service of direct marketers, crafting systems that cater to them, rather than to the big brands.

For years we’ve been driven by the principle that more targeting is always better. It’s just not. It’s just always smaller, more complex, harder to predict and ultimately less scalable. No matter how thinly we slice the bean, there’s always someone standing by with a sharper blade, always a few pennies more for an even leaner slice. Morgan is right: We’ve painted ourselves into a marvelously complex corner and only the truth can set us free — and only if we accept it.
Weaver suggests a return to basics. He’s not against all targeting, but in favor of “just enough” that lets the Web talk to brands on their terms — and maybe siphons some money out of TV budgets. Read the full post here.

More in Media

In Graphic Detail: How AI search is changing publisher visibility

AI platforms like ChatGPT and Google AI Mode are driving more search activity. Some publishers are gaining visibility — but not traffic.

AI royalties for small and midsize publishers: collective licensing’s next big play

Don’t credit OpenAI’s ChatGPT, credit corporate LLMs – enterprise RAG is what’s creating royalty revenue for publishers.

Graphic of a dollar sign-shaped key unlocking a lock, symbolizing the key to unlocking successful performance marketing through the seven stages of development

The Economist licenses its content to enterprise clients’ private LLMs

The Economist is among those to start licensing its content this way – having opened its API to corporate clients with their own data ring-fenced LLMs in August.