Salesforce still sees a place for live customer service agents after massive cuts

Article Date: 2025-09-10T14:43:00+00:00
Source URL: https://www.hrdive.com/news/salesforce-future-customer-service-people-ai/759615/
Image: Main image

Summary

Salesforce says its AI agents have handled millions of customer conversations recently, including about 1.5 million inquiries on help.salesforce.com over nine months, as the vendor scales agentic AI across support channels. CEO Marc Benioff, while noting AI’s growing role and cost benefits, emphasised that human agents remain essential for complex, empathetic or escalated interactions. The company reduced its customer support headcount from roughly 9,000 to about 5,000 as AI adoption rose, yet customer satisfaction (CSAT) scores have remained steady.

Partners such as DirecTV and Reddit report significant efficiency gains — fewer handling hours and much faster resolution times — and retailers like Williams Sonoma are deploying Salesforce-powered AI assistants. Salesforce’s broader business also shows growth: revenue rose 10% year over year to $10.2bn for Q2 FY2026, and Data Cloud and AI annual recurring revenue increased about 120% to $1.2bn.

Key Points

Salesforce’s AI agents have managed millions of conversations and 1.5m inquiries on its help site in nine months.
CEO Marc Benioff says AI agents reach roughly 90% accuracy, but human agents are needed for the remaining complex cases and for empathy.
The vendor cut its customer support workforce from about 9,000 to about 5,000 as AI usage expanded.
Partner case studies: DirecTV saved nearly 300 agent hours and executes ~50,000 AI-driven actions weekly; Reddit cut average resolution time from 8.9 to 1.4 minutes.
Salesforce’s revenue grew 10% YoY to $10.2bn in Q2 FY2026; Data Cloud and AI ARR rose 120% to $1.2bn, signalling strong commercial traction for AI products.
Surveys cited in the piece show many customers still prefer human support and report higher satisfaction with human-led digital interactions versus AI-only interactions.

Context and relevance

This story sits at the intersection of HR, customer experience and AI adoption. It illustrates how a major enterprise vendor balances efficiency gains from AI with continued reliance on human labour for complex and emotionally sensitive work. For HR and operations leaders, the article highlights the kinds of workforce shifts and upskilling choices organisations face when deploying agentic AI at scale. It also underlines the commercial case for AI within enterprise software while flagging limits around accuracy and customer preference.

Why should I read this?

Short version: if you hire, run or supply customer service, this is one to skim. Big vendor, big cuts, big AI wins — but humans still matter. We read it so you don’t have to: it gives the numbers, the trade-offs and partner examples you’ll want when planning headcount, training or tech spend.

Author style: Punchy — quick, focused take that saves you time and flags what to watch.