Research · Recognition Economy

The game already changed.

The Recognition Economy names the shift from the attention economy, where brands compete to be seen, to one where they compete to be recognized and retrieved by AI systems. This page is the evidence that the shift is real, and that it has not settled into a stable new normal. The audience is growing fast, the leaderboard of engines is still reshuffling, and even Google is rebuilding its own answers around named, cited sources rather than link lists.

Last verified 20 August 2026

In one paragraph

In eight months, ChatGPT's share of B2B AI referral traffic fell from 89.1% to 62.6%, while Claude, Gemini, and Perplexity all gained meaningfully. Half of US adults now use an AI chatbot, most of them for the first time in the last two years. And Google itself has changed how AI Overviews and AI Mode cite sources five separate times in 2026 alone, actively investing in more citation slots inside the answer itself. None of this describes a market that has finished moving. It describes an economy built around being recognized across a widening, shifting set of systems, not ranked in one.

Why The Game Changed

The audience moved first.

Before any brand strategy needed to change, the people brands are trying to reach already had. This is how fast that happened.

49%

of US adults now use an AI chatbot, up from 33% in 2024 - Pew Research Center, June 2026

42%

use an AI chatbot specifically to search for information, not just to chat - Pew Research Center, June 2026

24%

use an AI chatbot every single day, not occasionally - Pew Research Center, June 2026

The Leaderboard Keeps Moving

One engine's dominance fell by a third in eight months.

Independent B2B referral data, triangulating GA4 brand-panel data, SimilarWeb platform-visit data, and SensorTower app-ranking data, tracked how AI-driven referral traffic split across engines between mid-2025 and early 2026. The market that "AI visibility" describes turned out not to be one engine. It is at least four, and the split between them is still moving fast.

EngineMay–Aug 2025Mar–Apr 2026Change
ChatGPT89.1%62.6%−26.5pp
Claude1.4%18.5%+17.1pp
Gemini2.4%10.6%+8.2pp
Perplexity3.1%7.3%+4.2pp
Copilot3.2%4.0%+0.8pp

B2B AI referral traffic share. ChatGPT remains the largest single engine by a wide margin, but "largest" is no longer "only": Claude's share grew nearly 13x over the same eight months. A separate, broader web-wide measurement independently confirms the direction, reporting Gemini overtaking Perplexity as the second-largest AI chatbot referral source. Source: Goodie, 2026 AI Search Traffic Report, published May 2026, updated August 2026; StatCounter, 2026.

Even The Incumbent Is Rebuilding For This

Google keeps adding more room for named sources.

If the shift toward recognition were just an Aingworth framing, Google's own product decisions would not need to follow it. They are. In a single documented update, Google added five separate changes to how AI Overviews and AI Mode cite sources, all pointed the same direction: more named, attributable, explorable citations inside the answer itself.

01

Suggested exploration angles

A new section linking out to unique, in-depth articles on adjacent aspects of the topic, a fresh citation slot that rewards depth over breadth.

02

News subscription highlighting

Links from a reader's own news subscriptions get prioritized visibility; Google notes users were "significantly more likely" to click a link labeled as their subscription.

03

Creator attribution

Citations pulled from social platforms and discussion forums now show the creator's name, handle, and community, not just the site name.

04

More inline links

A higher density of links positioned directly beside the text they support, rather than bundled at the end of the answer.

05

Hover previews

On desktop, hovering an inline link shows the site's name and page title before a reader commits to a click.

Google's own stated goal is to make links "more useful," not simply more numerous, a distinction that matters: a system optimizing for usefulness is still applying a filter, just to a larger set of candidate citations than before. Source: Search Engine Land, "Google updates links within AI Overviews and AI Mode," May 2026.

Concentration Isn't Uniform

Some categories already play a winner-take-most game.

The Recognition Economy does not move at the same speed everywhere. In a study of 126 million US AI prompts across 22 industries, how concentrated visibility already is varied sharply by category.

News & Media

82.9%

of category visibility is held by just the top 3 brands. A near-total concentration, and a category where being recognized already means being one of a handful of names.

Finance

41.4%

of category visibility is held by the top 3 brands, less than half. A far more open field, where Earned Authority is still genuinely winnable.

Source: Semrush, 2026 AI Visibility Index. Only 36 global brands hold top-100 visibility across ChatGPT, Gemini, and Google's AI surfaces every single month, out of the full 126 million prompts analyzed, a reminder that the Recognition Economy rewards early, category-specific measurement over a generic, one-size score.

This page covers the category-level shift specifically. For how the machine decides which sources to trust, see the Machine Gatekeeper page. For which sources actually earn that trust, see the Earned Authority page.

Back to the research hub
Common Questions

Questions people ask about this page.

Is ChatGPT losing relevance?
No. It remains the largest single AI referral source by a wide margin, at 62.6% of measured B2B referral share. What has changed is that it is no longer close to the only relevant source, which it effectively was a year earlier.
Which AI engine is growing fastest?
In B2B referral share, Claude shows the largest proportional gain, from 1.4% to 18.5% over eight months. In broader web-wide measurement, Gemini has overtaken Perplexity to become the second-largest referral source.
Does this mean a single "AI visibility score" is misleading?
A brand can be strongly recognized by one engine and functionally invisible to another, since each system draws on its own trusted sources and citation logic. A single blended score hides that difference rather than revealing it, which is why we measure and build recognition by engine, not as one number.
Is this page part of the same research as the main hub?
Yes, this is one of several framework-specific evidence pages under /research/, each going deeper on one part of the Aingworth thinking. This one covers the Recognition Economy specifically, the category-level shift. The main hub covers the wider evidence base across Clarity, Credibility, and Earned Authority.

A note on verification

All figures on this page are checked against the primary source: Goodie's own published traffic report, StatCounter's own press release, Search Engine Land's own reporting, Pew Research Center's own report, and Semrush's own reported index. If a figure here has been superseded or misquoted, tell us at info@aingworth.com and we will correct it.

Find out where you stand today.

Run the free diagnostic to find which condition of the Triangle is leaking, or run the Recognition Check to see how often AI actually names you.