For Advertisers
Target the moment, not the cookie
Reach people while they're actively asking for a recommendation, matched to what they actually said.
Conversational Signals
Conversational Signals: The Intent Data Inside AI Conversations
When someone asks an AI assistant for a recommendation, they say what they need, why, and how close they are to deciding. Those are conversational signals — and they're a better foundation for ad targeting than a cookie ever was.
For Publishers
Turn conversations into a monetization signal
Make the intent already inside your AI product valuable to advertisers, without exposing private conversations.
How it Works
Why conversational signals matter now
For two decades, ad targeting has run on indirect evidence: which pages someone visited, what they clicked, what an identifier suggested about them. As third-party cookies fade, that evidence is getting thinner. At the same time, attention is moving into conversational assistants — where people don't leave a trail of clues, they simply say what they want.
Thrad CEO Andrea Tortella put it plainly in a recent Beet.TV interview: "My take is that the future of the cookie is actually context." A request like "best project management tool for a ten-person agency" already contains the category, the buyer, the size, and the stage of the decision. Nothing needs to be inferred.
It also changes what responsible targeting looks like. Because the signal comes from the conversation itself, it can be used within that conversation and clearly labeled — which is the foundation of conversational advertising done right.
Definitions
What is a conversational signal?
A conversational signal is any piece of intent or context expressed inside an AI conversation that can inform which ad, if any, is relevant in that moment: the topic, the stage of the decision, and the specifics a person states in their own words.
| Web signals | Conversational signals | |
|---|---|---|
| What it is | Behavior left behind: pages, clicks, searches | Intent stated directly, in natural language |
| How intent is known | Inferred from patterns and identifiers | Expressed: the person says what they need |
| Decision stage | Usually guessed from the page type | Visible: researching, comparing, or deciding |
| Specifics | Thin: a keyword or a category | Rich: constraints, preferences, timing, trade-offs |
| Dependence on identifiers | High | Low: the context is in the conversation itself |
Read the conversation
Start from what is being asked right now: the topic, the stage, and the stated specifics.
Match with relevance
Place a sponsored recommendation only where it genuinely fits that moment, using contextual, intent-based matching rather than a profile of the person.
Measure on conversational metrics
Judge performance by what happens inside and after the conversation.
Privacy
Privacy and limits
Better signals raise the bar for how they're used. On Thrad, every ad is clearly labeled as sponsored and ads are withheld from sensitive conversations. Signals inform relevance inside the conversation; they are not a license to follow someone around. Read more on how we approach safety in AI advertising.
Common questions
Conversational signals, explained
What are conversational signals?
The intent, context, and stated needs expressed inside an AI conversation — the topic, the decision stage, and the specifics a person gives in their own words — used to decide whether and which sponsored recommendation is relevant in that moment.
How are they different from cookies or first-party data?
Cookies and first-party data describe past behavior. Conversational signals describe a present, stated need. They don't replace first-party data, but they don't depend on identifiers either, which is why they matter as cookies fade.
What makes them richer than web data?
People explain themselves to an assistant in full sentences: what they want, their constraints, and how far along they are. A search query or a page view rarely carries that detail.
Does using conversational signals mean tracking people?
No. Matching is contextual and intent-based, ads are clearly labeled as sponsored, and they're withheld from sensitive conversations.
How do conversational signals improve media buying?
Buyers can place spend against the moment of need instead of an inferred audience, and measure on conversational metrics.
What's next?
As more assistants carry advertising, expect shared standards for how signals are used and disclosed — the same direction the industry is moving on AI transparency, covered in our explainer on IAB's disclosure framework.
Next step
Ready to put conversational signals to work?
Whether you're advertising or publishing, Thrad turns the intent already inside AI conversations into relevant, labeled placements — without relying on cookies or identifiers.