OpenAI and Perplexity are launching AI shopping assistants, but competing startups aren’t sweating it

With vacation procuring on the horizon, OpenAI and Perplexity each introduced AI procuring options this week, which combine into their current chatbots to assist customers analysis potential purchases.

The instruments are markedly just like each other. OpenAI means that customers might ask ChatGPT for assist discovering a “new laptop computer appropriate for gaming underneath $1000 with a display that’s over 15 inches,” or they will share photographs of a high-end garment and ask for one thing related at a lower cost level.

Perplexity, in the meantime, is enjoying up how its chatbot’s reminiscence can increase shopping-related searches for its customers, suggesting that somebody might ask for suggestions tailor-made to what the chatbot already is aware of about them, like the place they dwell or what they do for work.

Adobe predicted that AI-assisted on-line procuring will develop by 520% this vacation season, which could possibly be a boon for AI procuring startups like Phia, Cherry, or Deft (rebranded as Onton) — however with OpenAI and Perplexity pushing additional into AI procuring experiences, are these startups at risk?

Zach Hudson, CEO of the inside design procuring device Onton, thinks that AI procuring startups with a specialised area of interest will nonetheless present a greater expertise to customers than general-purpose instruments like ChatGPT and Perplexity.

“Any mannequin or data graph is simply pretty much as good as its information sources,” Hudson instructed Trendster. “Proper now, ChatGPT and LLM-based instruments like Perplexity piggyback off current search indexes like Bing or Google. That makes them actually solely pretty much as good as the primary few outcomes that come again from these indexes.” (Perplexity famous to Trendster that it has its personal search index.)

Daydream CEO and longtime e-commerce govt Julie Bornstein agrees — she remarked to Trendster over the summer time that she all the time seen search as “the forgotten baby” of the style business, because it by no means labored notably effectively.

“Vogue … is uniquely nuanced and emotional — discovering a costume you’re keen on is just not the identical as discovering a tv,” Bornstein instructed Trendster on Tuesday. “That degree of understanding for vogue procuring comes from domain-specific information and merchandising logic that grasps silhouettes, materials, events, and the way folks construct outfits over time.”

AI procuring startups develop their very own datasets in order that their instruments are skilled on higher-quality information — one thing that’s simpler to realize if you’re making an attempt to catalog vogue or furnishings, quite than the sum of all human data.

In Hudson’s case, Onton developed an information pipeline to catalog a whole lot of 1000’s of inside design merchandise in a cleaner method, serving to to coach its inner fashions with higher information. But when AI procuring startups don’t pursue that degree of specialization, Hudson thinks they’re certain to be overshadowed.

“Should you’re utilizing solely off-the-shelf LLMs and a conversational interface, it’s very arduous to see how a startup can compete with the bigger corporations,” Hudson stated.

The benefit for OpenAI and Perplexity, nonetheless, is that their prospects are already utilizing their instruments — plus, their giant presence lets them ink offers with main retailers from the get-go. Whereas Daydream and Phia redirect prospects to retailers’ web sites to finish their purchases — generally incomes affiliate income — OpenAI and Perplexity have partnerships with Shopify and PayPal, respectively, permitting customers to take a look at inside the conversational interface.

These corporations, which rely upon mammoth quantities of high-priced compute energy to function, are nonetheless making an attempt to determine a path to profitability. In the event that they take inspiration from Google and Amazon, then it is smart to look towards e-commerce as an choice — retailers might pay them to promote their merchandise inside search outcomes.

However finally, that would simply exacerbate the present points that prospects have with search.

“Vertical fashions — whether or not in vogue, journey, or residence items — will outperform as a result of they’re tuned to actual client decision-making,” Bornstein stated.

Extra reporting by Ivan Mehta. Up to date 11/26/25, 11:30 AM ET with remark from Perplexity.