Key Takeaways
- CRM shares have plummeted approximately 30% year-to-date amid concerns that artificial intelligence will undermine traditional SaaS business models
- Marc Benioff, CEO, maintains that Salesforce’s AI potential is reaching unprecedented levels
- The company plans to unveil Agent Albert, an advanced AI platform, before the end of the year
- Currently, 23,000 out of 150,000 total customers are utilizing Agentforce, with some reporting up to 40% reductions in support and IT tickets
- The platform handled 2.4 billion Agentic Work Units during the most recent quarter, representing a 57% increase from the prior period
Salesforce (CRM) has encountered significant headwinds throughout 2026. Shares have declined approximately 30% since the start of the year, pressured by mounting concerns that artificial intelligence capabilities could fundamentally undermine the software-as-a-service model the company pioneered.
The market’s anxiety centers on a fundamental business model issue. Salesforce generates revenue primarily through per-user licensing arrangements—billing organizations based on employee count. The fear is that AI-powered automation could enable businesses to accomplish more work with smaller teams, thereby reducing the number of licenses required.
CEO Marc Benioff rejects this narrative entirely. In comments to the Wall Street Journal, he stated: “People think we have our back against the wall when in fact the opportunity has never been greater.”
The pain isn’t isolated to Salesforce. The iShares Expanded Tech-Software Sector ETF (IGV) has fallen 20% during the same timeframe, with certain SaaS companies experiencing declines roughly twice as severe as CRM’s losses.
The company’s AI transformation began in earnest during early 2023, when Benioff convened approximately 40 senior leaders for an intensive three-day planning session at Salesforce Tower. The objective was to fundamentally restructure the organization’s annual strategy around artificial intelligence. For months afterward, these leaders continued meeting on Saturdays to refine the approach.
This strategic realignment yielded Agentforce, which debuted in late 2024. The platform enables organizations to deploy autonomous AI agents capable of executing tasks including customer service inquiries, lead qualification, and IT support requests. Adoption has reached 23,000 customers out of Salesforce’s total base of 150,000.
Tangible performance improvements are emerging from early implementations. Pearson deployed Agentforce agents to address order status inquiries, process refunds, and resolve access code issues—boosting the percentage of customer questions resolved without human intervention by 40%. PenFed Credit Union achieved a 40% reduction in IT support tickets by implementing an agent to manage password resets and account access issues.
Agentforce Limitations and Challenges
The technology hasn’t proven perfect across all use cases. Pandora’s chief digital officer reported that Agentforce experiences difficulty with ambiguous or sophisticated customer requests—such as providing jewelry recommendations when a customer mentions “my wife likes dogs.” Intricate problems continue to require human judgment.
Initial customer implementations also revealed that considerable time was required to prepare and structure data before AI agents could effectively utilize it. Salesforce addressed this friction by developing a data-integration infrastructure within its technology platform and pursuing strategic acquisitions in the data management and AI-powered sales domains.
Agent Albert Launch and Revenue Model Evolution
Before 2026 concludes, the company intends to release Agent Albert, a sophisticated AI platform designed to analyze user patterns and execute actions autonomously. Named as a tribute to Einstein, Salesforce’s long-standing mascot, the platform represents three years of internal research and development.
Regarding pricing strategy, Salesforce departed from its traditional seat-based model approximately twelve months ago. The company now employs a hybrid framework—maintaining user licenses while adding usage-based charges for Agentforce activities. A proprietary metric termed Agentic Work Units (AWUs) quantifies platform output: the previous quarter recorded 2.4 billion AWUs, reflecting a 57% sequential increase.
Benioff further contends that enterprises cannot simply code their own CRM systems using generative AI tools. The data security infrastructure, brand protection mechanisms, and regulatory compliance features Salesforce has developed over decades represent significant barriers to replication, he argued—even with tools like Claude Code or OpenAI’s Codex.
Salesforce has committed over $300 million to Anthropic since 2023. When the companies announced in February that Claude Cowork would integrate with Salesforce applications, CRM shares surged 4%.
Stifel analysts observed that “CIOs and CTOs prefer a unified platform that integrates agents, actions, data, and workflows”—a positioning Benioff is emphasizing as debates surrounding SaaS viability intensify.
Data from Andreessen Horowitz indicates that business customers with significant AI initiatives have increased their median Salesforce expenditure by 3% over the past three months.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
