This week, Glossy spoke with several experts and brand founders about the emerging field of AI search optimization. The optimal strategies are still emerging, but read on for a look at how brands are strategizing to show up in AI search results.
For years, brands have struggled with search engine optimization. Google’s black box of searchability has led to any number of strategies, from keyword bidding to posting keyword-stuffed articles that take forever to get to the point.
But now, there’s a new search optimization brands need to worry about. Increasingly, people are using LLMs like ChatGPT to search for products. These models behave similarly to Google, in the sense that they crawl the internet for information and return results relevant to a user’s input. But the way they do it is very different. Many of the strategies that work for SEO need be rethought to handle LLM search optimization.
Over the last few weeks, Glossy spoke with several brand workers about how they’re approaching optimizing search for AI models and where they’re finding good information. The field is still very new, so many said they are still figuring out their strategies, but a few common themes stood out.
Hayley Segar, founder of the “Shark Tank”-approved and now patented swimwear brand Onewith, said AI search optimization has been a constant topic among her brand founder group chats and meetups.
“The way I understand it, based on the buzz in the e-commerce community, is that AI search optimization has much more to do with long, conversational phrasing,” Segar said.
As opposed to traditional SEO, in which short, blunt keywords are incentivized, LLMs are more likely to pick up high-value, long-form text like essays and blog posts.
“With Google, for us, it’s just simple keywords like ‘flattering swimsuit,’ that work,” Segar said. “On ChatGPT, it’s way more conversational. We have to make sure the content on our site rephrases the hypothetical question, almost like Jeopardy. People are searching, ‘Where can I find a comfortable swimsuit that fits like underwear?’ So we have blog posts on our site phrased the same way.”
Onewith has been publishing at least one blog post a week since its founding in 2021, which Segar said has given her an advantage when it comes to AI search optimization.
Tom Nowak, CMO of the Midwest-based fashion retailer Evereve, agreed that high volumes of LLM-readable content is key. That’s part of the reason many brands are now launching Substacks, he said. But this text doesn’t necessarily have to come from published articles.
“Evereve produces more than 200 Evereve TV videos each year about new styles, important trends, outfitting advice and styling guidance,” Nowak said. “These videos run on social media, get streamed in stores and get posted on YouTube. When posted on YouTube, each video is automatically transcribed and these transcriptions are LLM-search gold.”
Caitlin Berzok, svp of digital at Therabody, told Glossy that LLMs are changing the way her team thinks about product descriptions and pages.
“The search engine isn’t just Google anymore, it’s everywhere,” she said. “[LLMs] talk like we talk. Whereas Google will pick up on all these meta tags and things, LLMs are looking for descriptions. It’s changing the way our writers have to think.”
Berzok said she’s been encouraging her team to check the sources of LLM search results. Both ChatGPT and Google’s AI Overview will usually show links to where it pulled information. For Berzok and Therabody, that’s often Reddit. According to data from online visibility platform Semrush, Reddit and similar platforms like Facebook or YouTube appear in 68% of all AI search results.
Mac King, the CRO and co-founder of the Shopify development agency Domaine, said it’s imperative that brands start tracking how much traffic is coming from LLMs and familiarize themselves with how their customers are using them. Domaine works with major fashion brands like Timex and Oscar de la Renta.
“If fashion brands want to get serious about AI search, the first thing they should do is start tracking it,” King said. “Set up a dashboard using your Google Analytics, and treat AI search as its own traffic source. And don’t just keep an eye on ChatGPT; brands should look at what’s happening across all the major players, like Perplexity, Claude, Google’s AI mode and any other platforms your audience might be using. You can’t improve what you don’t measure, so the very first move is to start paying attention to the data.”
King also said fashion is a bit of a tougher category, since style is more subjective. But some fashion brands Domaine works with have already seen their AI search traffic increase by 400% this year by using insights from customer focus groups and referral data. King suggested using tools like Google’s Search Console to find the exact terminology and phrasing your customer is using and mirroring it in your web content.
As for where brands can find good information, multiple founders told Glossy they’ve been turning to experts and advisors like Brennan Woodruff of the business incubator Mhub and marketing consultant Nik Sharma for advice, as well as consultancies like Profound for help in understanding the new landscape.
Kari Dowiak, the founder and designer behind the contemporary eyewear brand Memori, told Glossy she’s been pushing to earn more verified customer reviews on the brand’s website. She’s seen LLMs pull text word-for-word from her site’s reviews. She also said landing more media placements can improve the likelihood of being picked up by an LLM, even placements behind a paywall. LLMs are looking for expertise and will gravitate toward more authoritative content.
“Drill into the niche and use very specific industry terms that competitors aren’t using, everywhere from social media captions to podcasts,” Dowiak said. “Pretend you are talking expert-to-expert in some of your content, so the AI will look at you as an expert, not just a brand.”
Stat of the week
Usage of ChatGPT has jumped 70% from January 2025 to June 2025, according to Bain & Co. Shopping now makes up nearly 10% of all ChatGPT prompts, a 25% gain. Between March and June, click-through rates on links in ChatGPT results tripled, as well.
News to know
- Crocs’s sales fell by 6.5% this week, which CEO Andrew Rees attributed to caution among U.S. shoppers due to tariffs, as well as a “clear trend” for athletic footwear at the moment.
- On the flipside, wholesale profits are up at Asics, the parent company of the popular Onitsuka Tiger brand of sneakers. Tariff rates between Japan, where Asics is based, and the U.S. are still being negotiated, but Asics has raised its profit outlook for the year based on strong international performance.
- Gildan Activewear announced on Wednesday that it is acquiring HanesBrands for over $2 billion. Gildan hopes to double Hanes’s annual sales in the next few years. HanesBrands is the parent company of comfort-focused brands like Hanes, L’Eggs and Wonderbra.
- In other acquisition news, the Swiss watch brand Audemars Piguet announced it’s taking a majority stake in Inhotec, a Swiss precision manufacturer that supplies some of the biggest watch brands in the world.
Glossy’s fashion coverage
Inside On’s latest play for the luxury-lifestyle crown
Astrid & Miyu targets 50% US revenue growth with Madison Avenue store opening
David Allemann on On’s $853 million quarter, tennis push and Loewe lift
