Salesforce Pointed AI Agents at Leads No Human Contacted and Found Over $100 Million in Pipeline

Salesforce says it found more than $100 million in previously unseen pipeline by pointing AI agents at inbound leads no human had touched. Kris Billmaier of Salesforce describes the approach as the Untouched Lead Framework in a GTMnow roundup of five go-to-market frameworks: 75% of inbound leads scored 1-2 received no human contact, so the company assigned agents to that dead zone while human sellers stayed on higher-scoring prospects. The pipeline appeared within eight months.

The design choice is the lesson. Salesforce did not ask agents to replace sellers on the leads that already got attention. It split the work by where human time was already being withheld. Low-scoring leads were getting zero effort, so any agent follow-up was incremental, and the humans kept the accounts where their judgment moves a deal. That framing sidesteps the argument about whether agents sell as well as people and asks a narrower question: what is currently getting no one’s attention?

The same roundup shows why the human side still matters. Catherine Flynn of LinkedIn presents the Trust Gap Framework: 86% of B2B buyers prioritize seller expertise, while only 45% find sellers trustworthy. Her recommendations are practical: act on signals such as job changes, multithread early, respond within hours, and personalize more deeply. An agent that works the untouched leads has to clear that trust bar too, or it only converts a dead zone into an annoyed one.

The audit for a sales leader is simple and takes an afternoon. Pull inbound leads from the last two quarters, count how many scored low and received no human contact, and size the dead zone before buying any tooling. If the number looks anything like Salesforce’s 75%, the first agent deployment has an obvious target, and a baseline to measure against.