[Founder's Note] Context Is Becoming The New Targeting for Ads

[Founder's Note] Context Is Becoming The New Targeting for Ads

OpenAI Ads is now available in Korea.

But what stood out to me from the very first screen was something quite unusual.

It was not the ad placement itself that caught my attention. It was the change in how targeting seems to work.

The Familiar Targeting Options Were Not There

The ad targeting we are used to has always been fairly clear: age, gender, location, keywords, interests, and so on.

But when I looked at the current Ad Group setup for OpenAI Ads, those familiar options were not really there. Based on the screen I saw, “Context hints” appeared to be almost the only meaningful targeting option.

At first, that may look simple. But it may actually be one of the clearest forms of targeting.

A More Granular Signal Than Latest Interest

For years, we have experienced the way Meta ads suddenly show us related products after we visit a category site. Shoes we just looked at. Furniture we just searched for. Travel destinations we just compared.

That approach worked because a person’s “latest interest” is one of the strongest signals behind conversion.

But in an LLM environment, context can capture that latest interest at a much more granular level.

What Situation Is This Person In Right Now?

For example, what matters more than a profile like “a woman in her 30s who works full time” may be the specific problem that person is trying to solve right now.

• Someone is getting married soon and suddenly has skin trouble.
• Someone is comparing air purifiers before moving into a new home.
• Someone is looking for a baby gift from a brand that does not feel too common.
• Someone has a business trip tomorrow and needs a shirt that does not wrinkle easily.

Traditional ads were closer to asking, “Who is this person?” The context behind OpenAI Ads feels closer to asking, “What situation is this person in right now?”

What I Have Seen While Building AEKO

As some of you know, I have spent the past few months building AEKO, an AEO solution specifically for ecommerce brands.

And what I have seen through that process is clear. Even when the question is similar, AI answers change significantly depending on whether the prompt includes context.

“Recommend a good skincare brand” and “Someone has sudden skin trouble before their wedding and is looking for fast calming care” lead AI to think of different brands, product categories, and messages.

So the most important question for brands may no longer be simply, “Who is our customer?” It may become, “What are the customer moments where our brand should appear?”

Where AEKO Is Focused

This is the point AEKO is focused on.

We are building AEKO so ecommerce brands can easily and quickly explore thousands of contexts, track when and where their brand is mentioned inside AI platforms, and eventually design for new advertising environments such as OpenAI Ads.

The context your brand wants to own. For example, “the moment someone is worried about sudden skin trouble before their wedding.” For one person, that may be a moment that happens only once in their life.

Going forward, brands will need to understand, define, and manage these moments.

The Balance Between Precision and Scale

OpenAI Ads appears to allow up to 5,000 Ad Groups per account. In other words, brands may be able to separate and test thousands of highly specific contexts.

Of course, as with all advertising, hyper-precise targeting is not always the answer. If you narrow too much, you lose scale. If you go too broad, the message becomes diluted.

In the end, the question will be how to find the right balance between precision and scale.

But one thing seems clear.

The advertising market is also moving from the world of keywords and interests into the world of LLMs and vectors. Brands may need to think less about simply “who should see us” and more deeply about “in what context should our brand appear.”