Supplement brand Olly says it is seeing a shift in how people arrive at its DTC site: more shoppers are asking ChatGPT-style questions (“best sleep supplement,” “best women’s multivitamin”) and landing directly on product pages instead of politely entering through the homepage like it’s 2016.
Jennifer Peters, Olly’s director of DTC, marketing technology and digital compliance, said these visitors are coming in lower-funnel: spending less time clicking around, but converting faster. Most of Olly’s traffic still comes from social media and digital ads, and most of its revenue still comes from in-store retailers like Walmart and Target. But its website is increasingly being treated as a discoverability layer.
To make product detail pages more legible to both humans and large language models, Olly used Contentsquare to identify friction points like vague ingredient and “solution” descriptions. It has started expanding product pages with FAQ sections that spell out ingredients and efficacy in a format that is easy to crawl, quote, and misquote slightly less often.
The FAQs name-check terms LLM users are likely to ask about — adaptogens, zinc, letter vitamins — and connect those questions to specific products like Olly’s Sleep gummies. Peters also pointed to the messier second bucket: third-party content on Reddit, social media, and influencer channels, where outdated or incorrect information can affect how an LLM summarizes the brand. Olly even rewrote its Wikipedia page, though the update has not been published yet because Wikipedia still has to vet company-submitted edits.
Separately, Olly has been doing more conventional e-commerce housekeeping. It clarified “add to cart” calls to action, surfaced high-performing products earlier on landing pages for New Year’s and back-to-school sales, and tightened language around subscription options after seeing customers wanted clearer subscribe-and-save discount percentages. Peters said new checkout callouts on product detail pages drove a 20% month-over-month increase in website revenue earlier this year and lifted average order value.
The broader traffic shift is not just Olly’s pet theory. Adobe data shows AI-driven retail traffic rose 393% year over year in Q1 2026, with AI-driven revenue per visit 37% higher than non-AI traffic. A year earlier, non-AI visits were worth 128% more than AI visits.
Olly says it is not trying to chase chatbot traffic by pumping out generative AI content. “We don’t want to publish fake content,” Peters said. For now, the bet is less “make the robot write” and more “make the existing product truth easier for the robot to repeat without improvising.”
Read more at Digiday.