Consumers are increasingly asking ChatGPT, Claude, Perplexity, or Gemini directly, before Google even enters the picture. This is not a future scenario. It is already happening, and it is quietly changing how a purchase decision comes about.

The orientation path is getting shorter

Brands that steer mainly on search intent, aiming to pull traffic to their own site, are seeing that approach deliver less than it used to. Not because consumer demand is disappearing, but because it is landing somewhere else. That is the core of the shift. The question itself does not disappear, only the place where it lands changes. Anyone who wants to follow that shift needs a measure that looks at the market itself. Not just at their own website traffic.

A ladder of specificity

Why does one brand drop out of AI answers more easily than another? The answer lies in how specific the question is.

Ask about a brand by name, and the answer usually comes from that brand itself. Ask for the best options within a category, and a comparison appears. Make the question more specific, and a smaller, specialised publisher can suddenly become the more relevant source. A large retailer can be highly visible for broad category questions, yet lose that visibility to a niche publisher the moment a question becomes more specific.

This creates a clear blind spot. A brand can be losing ground on these more specific, high-intent questions for weeks before that loss ever shows up in its own website analytics.

What “Share of Answer” measures

Closing that blind spot starts with measuring the right thing. Not self-selected test prompts, but actual, measured orientation volume. That means tracking which publishers AI answers put forward as sources within a category, and for what share of that volume a given brand is visible. That percentage is the Share of Answer. A concrete benchmark that can be tracked both before and after any activation.

Where the honesty stops and the hype would normally start. A metric like this invites overclaiming, so it is worth being precise about what it is not:

• It is not a claim that a brand’s own organic findability inside AI answers will improve; that effect is untested and would be measured separately before anyone says otherwise.
• It is not a way to attribute AI-driven traffic one to one to a specific placement; the measurement technology to do that reliably does not exist yet.
• It is not evidence of a causal link between reach and brand awareness.

What is left after ruling those out is still useful. It gives you a measured starting point, and a way to act on it deliberately rather than by instinct.

The solution: AI Publisher Discovery

This is the gap that Daisycon and Trendata built AI Publisher Discovery to close. Trendata’s Demand Intelligence platform continuously tracks how consumers orient themselves, following the questions they ask via search, social, and AI, including ChatGPT, Gemini, Claude, and Perplexity, alongside Google and Bing. Data is refreshed monthly, with visibility up to 48 months back and forecasting up to 12 months ahead.

Daisycon translates those insights into direct commercial action. That means looking not only at the publishers already active within a category. It also means identifying relevant sites that frequently appear in AI answers but are not yet part of the network. These sites can then be approached and activated within a campaign.

Where this works best

The biggest opportunities sit in categories where consumers actively compare their options and look for guidance, typically physical, comparable products. Questions like “what is the best X for Y”, “which X suits me”, or “X versus Y” tend to lead directly to a shortlist of brands and products. That shortlist is the new AI shelf: the question is no longer only whether a brand is visible when someone searches for it, but whether it is included when someone asks AI which options are worth considering.

Where things stand today

The pairing itself is new, but neither side is starting from scratch. Trendata’s forecasting methodology is evidence-based, grounded in research into how online orientation behaviour relates to market and sales development, and is reported to reach over 85 percent accuracy against historical data. That sits on top of Daisycon’s affiliate marketing experience, built up over more than 25 years, and Trendata’s own track record of more than 8 years in AI-driven market intelligence.

Rather than opening this to every advertiser at once, it is launching with a limited group first. We will start with a pilot to measure current visibility and identify where the opportunities actually sit. It also shows what activation delivers in practice. This means the wider rollout can be shaped by results rather than assumptions.

Want to discover, as one of the first advertisers, what role your brand plays in the AI orientation of your category? Leave your details here and be the first to know when we launch.