AI is no longer a future consideration for eCommerce brands. It is actively deciding which products get discovered and which ones disappear. In this episode of Shero Live, Beth Shero is joined by Cande Rybecky, VP of Marketing and Growth at Shero Commerce, and Molly Sishton, Co-founder of SO-ME, an AI-powered consumer shopping app.
Molly learned the hard way that most product data simply is not ready for natural language search, and she is here to share what she found.
In this episode
-
Why product data is the single biggest factor determining whether AI recommends your brand or skips it entirely
-
How to balance AI optimization and conversion on your PDPs without keyword stuffing or endless scrolling
-
What the shift from "being found" to "being understood" means for how you structure collections and content
-
Why ChatGPT ads are generating so many questions, and what we actually know so far
Watch the Full Conversation
Episode highlights
Molly Siston built SO-ME after trying and failing to find a black-tie dress with a simple natural-language search. What felt like a straightforward task exposed a fundamental problem: product data across major retailers was not structured for how people actually describe what they want. That experience sent her team deep into hundreds of thousands of product records, standardizing and enriching data so it could be matched to real conversational queries. Her takeaway for brands is clear.
The question is no longer how to be found. It is how to make your product understood.
The conversation gets practical around product detail pages and collections. Molly's advice is to avoid cramming every possible keyword onto a single PDP. Instead, she recommends building niche collections that answer specific natural-language questions, such as "power dressing for work" or "what to wear to a summer wedding." These collection pages give AI agents more structured, contextual information to work with, and they keep individual PDPs clean enough to convert. Tags and backend data matter too, since crawlers are pulling signals from more than what is visible on screen.
Cande draws a direct parallel between AI optimization and SEO, arguing that brands need to make the same mindset shift they made when organic search first emerged. You are not going to see immediate returns. The ROI is slower, more like a marathon than a paid media campaign. But brands that are doing the right things now, investing in case studies, rich content, and structured product data, are already showing up in ChatGPT results. Shero closed a client in Q1 who found them through ChatGPT, which Cande holds up as proof that the work compounds over time.