The news: Salesforce is acquiring AI customer service platform Fin, formerly known as Intercom, for approximately $3.6 billion.

• The deal, expected to close in Salesforce’s fiscal Q4 2027, will add Fin’s AI agent to the company’s Agentforce platform.
• Fin’s agent can handle customer queries across chat, email, WhatsApp, text, phone, and Slack using its proprietary Apex AI model.

Zooming out: Agentic AI is rapidly moving into production across customer service channels. Among customer service professionals surveyed globally in April 2026, 70% or more reported full production deployment of AI agents across chat, email, phone, and text, per a May 2026 Salesforce report.

Three factors will determine how much value Fin delivers for retailers:

Integration: Salesforce has a strong track record of embedding acquisitions into its core platform. Its $27 billion purchase of Slack in 2021 is now deeply integrated into Agentforce workflows, suggesting Fin could follow a similar path.

Channel fit: Fin’s coverage across chat, email, WhatsApp, phone, and Slack aligns with the channels retailers already use, which should ease adoption for brands on Salesforce’s CRM.

Consumer limits: While 88% of US adults who use AI-powered customer support say it resolved their issue, only 22% say it made them prefer the company, per Gladly’s 2026 Customer Expectations Report.

The challenge: Real-world performance doesn’t always match the promise.

When Rainbow Shops tested an AI voice agent, it successfully handled 40% of calls end-to-end but had little impact on overall agent workload, said Chief Digital Officer David Cost on EMARKETER’s "Reimagining Retail” podcast. That suggests many interactions were technically “closed” without fully resolving underlying issues.

Implications for retailers: While the Fin acquisition gives Salesforce clients a more capable, omnichannel AI service layer, the key question remains whether these agents actually resolve customer problems. Rainbow Shops' example shows that measuring conversation completion alone without fixing the underlying issue can create a misleading picture of performance while shifting the burden back to human agents.

That’s why retailers evaluating Fin should focus on resolution rates and customer satisfaction rather than surface-level metrics. As the Gladly data shows, resolving an issue with AI doesn’t automatically build loyalty, and using it primarily to cut costs can erode long-term relationships.

The more durable opportunity lies in augmentation. Salesforce says its own use of Agentforce saved 30,000 developer hours per month by automating repetitive tasks. Retailers that use AI to triage, route, and assist—while reserving human agents for more complex interactions—may be better positioned to improve both efficiency and the customer experience.