As shoppers flock to LLM-powered searches to learn more about products and set shopping budgets, Salesforce data reveals new opportunities for brands looking to have a firm hold on their customers' journeys.

“We see a lot of questions [from retailers and brands] being, How do my brand and my products show up where the shoppers start their journey?” said Caila Schwartz, Salesforce’s director of consumer insights and strategy for retail and consumers. “How do I take ownership of my brand presence in these spaces, and how do I plug into this audience space?”

Because of improvements like ChatGPT’s app partnerships, there’s more for consumers to discover in these searches.

• AI-powered search tools are projected to drive 21% of all holiday orders globally, accounting for $263 billion in sales, according to Salesforce data.
• 17% of consumers have used an AI assistant or LLM in the last year for product searches, according to Salesforce research.

Growing confidence in AI

The path to purchase is changing as consumers grow more confident this year in LLM-powered searches.

Eighty-six percent of users now trust AI recommendations, up from 46% in May, according to Salesforce research.

“The consumer has leaned into these third-party AI tools like ChatGPT, Perplexity, etc. and essentially what we’ve seen is really strong growth and interest on the consumer side in leveraging these tools for hyper-personalized product recommendations and for budgeting and planning,” Schwartz said.

In addition to asking about products, shoppers will give the genAI-powered chat a budget and ask the LLM to respond with a gift list or shopping list that fits it. That example, along with product recommendations, are the top two shopper use cases, Schwartz said.

Converting shoppers

Because LLMs keep improving, Schwartz says users are more likely now to find a brand website, or even a coupon code, in response to their queries. Increased trust in the LLMs means they’re more likely to click.

“This is a shopper that is getting a lot of information…and they’re more likely to buy once they get to the retailer or brand website,” Schwartz said.

• Traffic from AI assistants grew in volume 119% YoY in the first half of 2025, according to Salesforce research.
• Conversion rates from these channels were 700% higher than social media traffic and 200% higher than other traffic sources like traditional search, Salesforce found.

With the holiday season ramping up, shoppers show interest in using these tools. Over half (56%) say they plan to use the tools for Black Friday Cyber Monday (BFCM) this year, according to an August Klaviyo survey.

“It’s still early days, but they’re absolutely getting referrals,” said Schwartz. “We’ll see how that referral changes as a lot of the shopping capabilities are built out directly within the tool.”

Hyper-targeting LLMs

In order to show up in LLM-powered search results, retailers and brands need to pivot their content strategies. Hyper-targeted blog posts, product descriptions and catalogues will show up better in response to longer, hyper-personalized queries.

For example, a blog post or catalogue for toys might be titled, “a gift guide for a nine-year-old boy who loves Minecraft,” Schwartz said.

“Getting that level of specificity means a lot more content, in order to be pulled into those searches,” she explained. “So there’s a lot of work on the brand side to try and generate that content, even leaning into AI tools.”

This was originally featured in the Retail Daily newsletter. For more retail insights, statistics, and trends, subscribe here.