StudioHawk has warned that some businesses are losing up to 80% of their online traffic as users turn to AI search tools and AI-generated summaries. Founder Harry Sanders said the shift is changing how consumers find brands online.
Tools such as ChatGPT, Claude and Google's Gemini, along with AI overviews at the top of search results, are reducing the traffic that once flowed from conventional search rankings to company websites. Sanders cited research suggesting that an AI overview in Google search results is linked to a 58% lower average clickthrough rate for the top-ranked page.
The warning reflects growing concern among publishers, retailers and service businesses that AI-generated answers are reshaping the economics of online discovery. If users get enough information from a summary at the top of a page or from a chatbot response, they may never click through to the original source.
StudioHawk argues that businesses need to rethink how they present information online to remain visible in AI-driven search. Sanders said relying on standard search engine optimisation practices is no longer enough on its own.
Topical depth
A central part of that adjustment, he said, is building what he describes as topical authority. The idea is that AI systems break broad questions into smaller underlying searches before assembling an answer from multiple sources.
In practice, a user asking for the best accountant for a small business may trigger searches on price, location, specialisation and reviews rather than a single direct query. Businesses that cover a subject area in depth are therefore more likely to appear across those different strands.
"You don't become a trusted source by publishing one good page," Sanders said.
"You become one by covering a topic so thoroughly that you're the obvious place to draw an answer from."
He added that firms should focus on subjects they genuinely understand and answer the full range of customer questions around them, rather than only targeting the most popular search terms.
"Pick the subjects you actually know, and cover every question around them. That depth is what AI reads as expertise," Sanders said.
Original material
Another recommendation is what Sanders calls information gain, or publishing material that adds something not already widely available on the web. He said AI systems are less likely to surface pages that simply repeat existing content.
That places greater emphasis on first-hand experience, proprietary data, specialist opinion and more complete answers. Businesses that bring original material to a page stand a better chance of being cited in AI-generated responses, Sanders said.
"AI doesn't want another version of the same answer. It wants the one that adds something new," he said.
"If your content brings information the rest of the web doesn't have, whether that's original data, real experience, or a genuine opinion, that's what gets you quoted."
He also said page structure matters because AI tools need to extract information cleanly. Clear subheadings, direct answers, lists, tables, summaries and question-led sections can all make content easier for large language models to parse.
"The way to make sure AI can use your content is to make it easy to lift: well-organised H2 and H3 sections, 'how to' questions with clear answers, lists, and tables," Sanders said.
"FAQs, TL;DR summaries at the top of key pages, author pages and AI-friendly product descriptions all give you an edge."
Brand signals
Sanders also pointed to brand recognition as an important factor in AI visibility. He said businesses with consistent mentions across reputable sites and stronger backlink profiles are more likely to be treated as trustworthy sources.
That reflects a broader shift in search, where brand familiarity may influence both direct user behaviour and how automated systems judge a source's reliability. Companies are also being encouraged to monitor traffic from AI tools, track branded search activity and use search analytics to gauge changes in visibility.
"A strong brand, one that's mentioned across credible sites, cited often and linked consistently, is more likely to be chosen by AI as a source," Sanders said.
"If humans know your brand, AI does too."
Common mistakes
Sanders said the biggest problem for many businesses is not a specific technical error, but a failure to respond to changes in user behaviour and search design. He argued that many companies are still relying on conventional SEO content without considering how AI systems read and assemble answers.
"The biggest mistake is using old SEO content alone and ignoring AI structure needs," he said.
Businesses with strong search foundations will find it easier to adjust, while others risk a sharper loss of visibility as AI tools become a more common starting point for product and service research, Sanders said.
"Brands with solid SEO foundations will adapt fastest. Those without will struggle to get seen."
"No need to panic or chase the next AI SEO gimmick. Just work with someone who actually knows SEO and knows how to apply it to AI search," he added.
Sanders said the shift could also create openings for smaller brands if they publish useful, clearly structured material rather than simply competing on size or existing ranking strength.
"It's going to level the playing field," he said.
"AI search gives real visibility to brands that produce clear, useful answers, not just those who rank highest."