Between physical retail and e-commerce is the booming “third shelf” that is agentic commerce where personalized recommendations add another layer of competition by determining which products show up the answer to shopper queries.
Consumer use on agentic AI shows no signs of slowing down, with Spate data showing that ChatGPT activity skyrocketed 30 times since January 2025.
One of the defining features of agentic commerce is real-time personalization from AI agents’ relationship with consumers – which could be a huge benefit to brands depending on how they set up their data, according to Jessica Wright, Spins’ SVP of Foundry, an agentic discovery solution for CPG brands.
AI platforms are trained to customize their responses to consumers’ preferences, like products with a specific certification or nutrient content – which can change how products are surfaced and unlock a “level of personalization that I think the industry has been striving to achieve,” Wright said.
And the industry is picking up the pace with adoption. More than half (62%) of respondents in Spins’ 2026 Executive Pulse survey said they are actively exploring agentic shopping tools, while 31% report their products are not optimized for discovery.
Likewise, 9 in 10 retailers are adopting or piloting AI, and 68% of retailers look to generative AI to improve their marketing and content, per Spins.
The SEO evolution into GEO
Instead of broad targeting via traditional SEO keywords, backlinks and rankings, brands must ensure their product attributes are optimized for AI agents and generative engine optimization (GEO) that drives their searches – and like anything in the digital space, it all starts with structuring the right data.
Rather than GEO replacing SEO, Wright says brands can leverage their existing SEO framework.
“[GEO] not only helps you in an SEO standpoint, but it also helps for your discoverability in AI responses as well,’ she said.
But, brands must now translate it into formats that machines can interpret.
The new dataset for the third shelf
In the agentic commerce ecosystem, a brand’s dataset – marketing copy, advertisements, reviews and user generated content, among others – is “essentially up for grabs for these AI assistants to retrieve that information and use that to formulate its own context and perception around the product and brand,” Wright explained.
Essentially, AI platforms blend the datasets they find to form what Wright describes as a “mental model” of a brand.
AI agents are catching structured data – rather than vague, aspirational marketing speak – that highlight a products’ attributes and how it delivers value to the consumers
“Fluffy marketing content [is] not going to be as easily interpretable for an AI assistant,” Wright added.
She illustrated this with an example of unclear messaging for a hypothetical energy drink brand that’s marketed as “an active sense of being throughout the day.”
The copy doesn’t provide any details about the product – and that ambiguity is problematic for an AI system, she said.
“There’s no mention that this is an energy drink, there’s no mention of the caffeine content, there’s no mention of any other health benefits,” she added.
Instead, brands must communicate in explicit, structured attributes “that are very clearly interpretable by a machine,” Wright said.
The data shift is toward precision: clearly stated ingredients, benefits, certifications and functional attributes.
Where is agentic commerce headed?
Currently, AI is most heavily influencing discovery, but that will expand, Wright said.
In the future, more utilitarian purchases – like staple products that need replenishment – may be fully automated, she explained.
At the same time, differentiation will become even more important as AI assistants continue to learn their users’ needs – and brands that can translate their products’ attributes into a format that is easily understood by AI agents will be more visible.
