All coins

WiremarketsDextape brief

Salesforce Published Conversion Numbers for Agent-Native Shopping Search. They Cover One Thing.

Salesforce has published merchant-side conversion numbers for agentic commerce product discovery, marking a shift from theoretical agentic potential to reported performance metrics. According to the Winter ’27 Release, the company’s Agentic Commerce Search capability is now delivering a 13% conversion lift and a 17% increase in add-to-cart rates.

Dextape brief

By Dextape Desk ·

Salesforce has released merchant-side conversion figures for its Agentic Commerce Search tool, reporting a 13% conversion lift and a 17% rise in add-to-cart rates. The numbers appear in the Winter '27 Release, and the company describes them as early results. The feature uses a small language model trained on each merchant's catalog and shopper history to interpret natural-language queries. Salesforce has made the capability standard for all Shopper Agent customers. The article notes that the vendor has not disclosed its methodology, sample size or baseline metrics, and that the release does not say who bears liability when a search misleads an agent. Salesforce also plans to charge for each call a registered agent makes, with a November target and a multiplier yet to be announced.

Why it matters: It shows what agent-driven product discovery reportedly does to sales funnels, while the missing methodology and liability terms leave merchants without a clear basis to judge the figures.

Summary by Dextape, based on reporting by Forkast. Written with AI from the publisher’s text and checked automatically for copied wording and unsupported figures. Not investment advice.

Read the full story

Headline and excerpt from Forkast’s public feed. Forkast holds the copyright and publishes the full story; Dextape links to it rather than reprinting it.

Brief by

Dextape DeskEditors

The Dextape Desk signs Dextape's own work: briefs on wire stories, the daily market report and desk analysis. Briefs are written by an AI model from the publisher's article text and checked automatically for copied wording, unsupported figures and investment advice before they appear. AI drafts from the editorial desk need corroborating sources and an editor's review before release.

Share this story