With summer in the rearview mirror, marketers are preparing for a 2026 holiday season that will look very different from last year. According to Salesforce, 20% of holiday ecommerce traffic is expected to come from AI agents this season. By Cyber Week, Salesforce predicts that one in three ecommerce sites will have an AI shopping agent of their own. Salesforce put it plainly: “Agentic search will be king of the holidays.”

AI is also changing the fundamentals of consumer product discovery. In a separate Salesforce survey, 79% of commerce leaders reported an increase in traffic attributed to large language model (LLM)-powered search. This momentum is why the holiday season will be decided by AI-generated recommendations.

A growing share of those recommendations are coming from retail-native agents, like Alexa for Shopping and Walmart Sparky. In July, Amazon CEO Andy Jassy said that more than 350 million customers used the company’s shopping agents over the last year. Weeks later, Walmart CEO John Furner said the number of customers using Sparky is up 70% from 2025.

Their ability to quickly shepherd consumers from consideration to purchase makes retail-native agents a critical channel for holiday marketing strategies. Agentic marketing platforms (AMP) like Bluefish are designed to help marketers understand and optimize how shopping agents recommend their products.

How to optimize for retail-native shopping agents

So how should brands optimize for this fast-growing channel? Brands are learning that product detail pages (PDPs) need to do more than inform human shoppers. They can also serve as structured knowledge hubs that help LLMs understand and recommend products. Most PDPs have an average of 300–500 words, but LLMs need clear information to truly understand a product, its primary use cases and benefits, and how it fits the needs of different shopper segments.

Traditional search relies heavily on matching queries with relevant keywords on a webpage. Shopping agents work differently: they interpret what a shopper wants, then compare products based on the factors relevant to that decision. Brands need to optimize product pages around the topics AI uses to evaluate and recommend products, not just the search keywords.

A global media agency recently used Bluefish to identify why a leading consumer packaged goods account was losing Alexa for Shopping recommendations to competitors. The agency discovered that competitor citations were often driven by active ingredients listed on rivals’ PDPs.

Using that signal, the agency restructured the brand’s PDPs, adding ingredient comparison modules and ingredient-benefit FAQs based on topics the Bluefish AMP identified as influencing Alexa for Shopping’s responses. Following the changes, every optimized PDP improved its performance in the shopping agent, with one product leapfrogging more than 500 competitors to become a top performer in its category.

Key takeaway: Retail-native shopping agents are becoming an audience in their own right, with different information needs than human shoppers. Brands that understand those differences will be better positioned to earn recommendations, and the corresponding revenue, this holiday season.