Webflow Research Finds Most Companies Are Unprepared for the AI Discovery Era

Study of 2,000 company websites reveals the median brand appears in just 16% of AI-generated answers, earns citations only 6% of the time, and is inaccurately described roughly one-third of the time – largely due to fixable fundamentals most brands already control

As AI assistants rapidly become a primary way people discover brands, products and services, new research from Webflow suggests most companies are failing to show up, or show up accurately, when customers turn to AI for recommendations. The research also pinpoints why, finding most shortfalls trace to fundamentals that brands control.

Webflow, the agentic web marketing platform for modern marketing teams, analyzed more than 2,000 U.S. company websites, spread across industries and company sizes using publicly available data, to measure how effectively brands are represented across leading answer engines, including ChatGPT, Gemini, Perplexity, and Claude. To understand both AI visibility and the main drivers behind it, Webflow scored every site on its AEO Maturity Model, a benchmark built on the four areas that drive AI visibility: technical readiness, content, authority, and measurement. Each site is rated from Level 1 to Level 5.

Among the report’s key findings:

• The median company appears in just 16% of AI-generated answers relevant to its brand.
• The median company receives a citation link only 6% of the time, limiting referral traffic, and therefore revenue, from AI platforms.
• When companies are mentioned, AI gets the facts wrong roughly one-third of the time, with almost no message pullthrough for brands – reducing accuracy and hurting brand fidelity.
• The average company scored just 2 out of 5 in Webflow’s AEO Maturity Model, highlighting significant room for improvement across technical readiness, content, authority and measurement.
• The root causes for low visibility are fixable fundamentals: 62% of sites have broken internal links, 60% are missing basic SEO metadata, and 54% haven’t refreshed even a tenth of their content in the last six months.

Also Read: AiThority Interview with Matej Bukovinski, Chief Technology Officer at Nutrient

“For nearly three decades, companies built websites primarily for people,” said Guy Yalif, Chief Evangelist at Webflow who led the research project and developed the AEO Maturity Model. “Today, websites have two audiences, humans and AI systems. Our research shows most companies are still optimizing for yesterday’s internet, when the brands that win tomorrow will be those that engage people and make their content easily understood by AI.”

The research found that technical fundamentals remain the biggest barrier to AI visibility. Nearly two-thirds of companies have broken internal links or missing SEO metadata, while more than half rarely refresh content or directly answer prospect questions on their website.

At the same time, larger brands already enjoy stronger third party authority signals, including credible third-party mentions and expert-authored content, driving a lot of their AEO maturity and visibility. The study also found that smaller brands can perform similarly to these larger brands, despite typically lack the same third-party authority signals, through better content and stronger technical setup (things the brand fully controls).

The research also breaks down how groups like the Forbes AI 50, the tech titans, and the top ad spenders stack up. For example, the AI-native leaders in the Forbes AI 50 appeared nearly twice as often as the typical company, while a number of award-winning creative brands, including 2026 Cannes Grand Prix winners, had room to grow.

“We view Answer Engine Optimization (AEO) as an evolution of SEO, and it’s clear that this discipline is much more cross-functional than SEO is,” continued Yalif. “AEO spans brand marketing, engineering, content, communications, product, analytics and more. Organizations that rethink how they build and manage their digital presence to drive brand awareness and traffic from answer engines will get oversized benefits in these early days of this new medium.”

Also Read: ​​AI systems – Interoperable AI systems: Connecting models across platforms