By Patrick Coffee

For marketers exploring AI search, this year's Black Friday weekend delivered both a moment of clarity and a series of questions.

The websites of 20 large retailers from Best Buy to Etsy averaged 183,000 daily visits from people sent by ChatGPT and other large language models, according to marketing software firm Semrush. That is still tiny compared with traditional search traffic but nearly eight times last year's daily average. In 2023, daily results ranged from zero to the low hundreds for each site.

AI search is on a trajectory to drive revenue. How companies should prepare remains up for debate.

"Every chief marketing officer today is thinking about how they're showing up in ChatGPT," Anil Chakravarthy, president of the digital experience business at Adobe said last month when the software company announced a $1.9 billion deal to buy Semrush.

Will AI search replace traditional search?

Benjamin Houy is firmly in the camp of "don't touch that dial." The entrepreneur earlier this year used his own money to launch Lorelight, a tool designed to optimize the presence of brands in LLM results. It lasted about six months.

Rather than revealing new ways to influence artificial-intelligence algorithms, Lorelight's data simply reinforced the value of time-tested practices for traditional search engine optimization, like appearing in respected publications and producing content that provides consumers with legitimate expertise.

"I just stopped believing in the core premise of the product, which was that it could help companies rank higher on AI," Houy said.

Another school of thought might be called "SEO is dead," as many business headlines have proclaimed, or at least vastly diminished. Believers this year packed funding rounds for a swarm of startups selling answer engine optimization, or AEO, also known as generative engine optimization , which they argue requires different tactics than SEO.

AI search might not immediately supplant SEO, but its trajectory over the next five years is "a one-way street," according to Brian Stempeck, co-founder and chief executive of AI search optimization startup Evertune. When Stempeck recently asked a group of University of Virginia students whether they use Google as their primary search platform, no hands went up, he said. They mainly use AI tools like OpenAI's ChatGPT, Google's Gemini or X's Grok, or TikTok, he said.

A lot of money rides on whose read is right.

The global SEO and AEO services market will grow to $171 billion by 2030 from $81.4 billion last year, partly driven by growing AI optimization efforts, according to a prediction by research firm MarkNtel Advisors.

Evertune and Profound, two of the largest AI search optimization startups, raised $15 million and $35 million, respectively, in August.

Other kinds of businesses are joining the AEO gold rush. Press release distributor GlobeNewswire in September announced a new product, for example, that uses Profound's data to help create and track PR campaigns. Evertune joined with affiliate marketing platform Impact to help create paid content like product guides that are designed to show up in large language model search results.

Or will there be a merger of disciplines?

An initial idea that AI optimization would replace SEO as traditional search withered has evolved into a view that the two disciplines will merge as both kinds of search persist, said James Cadwallader, chief executive officer and co-founder of Profound. SEO firms will likely add AI services, he said, and vice versa.

Part of the solution for companies seeking bigger or better profiles in AI results is to make sure large language models have material to ingest, for example by putting out more news releases, according to Cadwallader. That matters more than in traditional SEO because LLMs generally need more information to answer detailed, conversational questions that build on a given user's history.

"There's no mystery around it," he said. "If you create lots of content, the models will suck it in."

Some marketers, however, worry that the suggestions they are hearing closely resemble content farming, a practice as old as the internet that gives priority to quantity over quality.

"We've seen this approach before," said Jennifer Vianello, chief marketing officer at car sales platform Cars Commerce, which runs Cars.com. "Cheap or easy" approaches, like adding links or key phrases, have been used for years to rank higher in search results, but this strategy doesn't deliver added value to users, she said.

Some tweaks developed by the company's own editorial team, such as bullet points, takeaway summaries and other elements designed to make webpages more easily digestible, have helped make content more appealing to LLMs, but are also continuations of Cars Commerce's established SEO practices rather than new strategies, Vianello said.

Extremely detailed, internally produced product information does seem to help the travel company Expedia Group in AI chats, according to Chief Marketing Officer Jochen Koedijk.

Expedia has appeared more prominently in AI responses about hotels, for example, after adding granular descriptions of amenities like complimentary parking, streaming services and pools at certain locations, Koedijk said .

AI search also values recency even more than traditional search does, research by Profound suggested. Material published or updated within the previous 13 weeks is 50% more likely to appear in chatbot responses, the research said.

The biggest distinction between traditional search and AI is large language models' greater emphasis on user-generated content about brands on platforms like Reddit, Quora and YouTube comment sections, according to Andrew Warden, chief marketing officer at Semrush.

Brands can succeed in AI search by "returning to the roots of who they are, what their value proposition is to their customers, how they're telegraphing that in the market," Warden said. Ensuring that customers understand a brand's value proposition, and that they discuss it in these kinds of forums, is key, he said.

Brand marketing to influence consumers will be more valuable in the long-term than the current swarm of AEO activity, much of which amounts to marketers trying to boost their results on the cheap just as they did in the early days of Google, said Tom Critchlow, executive vice president of audience growth at Raptive, a platform that helps creators and the publishers they work with make money.

Marketers should focus on experimentation and be wary of anyone claiming certain knowledge of future AI search models, said Koedijk, the Expedia executive.

"If anybody tells you they know how that will be in two years, I don't believe them," he said.

Patrick Coffee writes for WSJ Leadership Institute's CMO Today. Reach him at patrick.coffee@wsj.com.

(END) Dow Jones Newswires

December 18, 2025 06:00 ET (11:00 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.