TMTPOST -- Microsoft Corp. closed down 2.5% Wednesday after a volatile trading session triggered by conflicting reports about its artificial intelligence (AI) sales quotas. The stock dropped as much as 3% following the Information's report that the company had lowered growth targets for certain AI products after salespeople missed targets. The decline narrowed after CNBC reported Microsoft's denial of the claims.
AI Generated Image
The Information reported that several Microsoft divisions reduced quotas for how much salespeople should increase sales of certain AI products after many missed their targets in the fiscal year ending in June. The report cited two salespeople in the Azure cloud division and focused on Microsoft's Foundry marketplace for AI models and agents.
Microsoft pushed back strongly against the report. "The Information's story inaccurately combines the concepts of growth and sales quotas," a company spokesperson said in an emailed statement. "Aggregate sales quotas for AI products have not been lowered."
The episode highlights mounting questions about whether massive tech spending on AI infrastructure is justified by current customer adoption rates. While Microsoft and rivals including Alphabet Inc.'s Google and Amazon.com Inc. continue investing heavily in data centers and chips, some businesses complain it's hard to measure savings from AI tools and worry about costly mistakes the technology still makes.
Sales Target Adjustments
One Microsoft sales unit had asked salespeople to increase customers' spending on the Foundry marketplace by 50% in the last fiscal year, according to the Information. Fewer than one-fifth of salespeople in that unit met their targets, and in July Microsoft lowered the goal to about 25% growth for the current fiscal year, the publication reported.
In another unit, the quota was set to double Foundry sales but was dropped to 50% after most salespeople didn't meet it, according to the Information. The publication cited two salespeople in Azure's cloud unit. The size of Foundry quotas per salesperson can range from tens of thousands to hundreds of thousands of dollars annually.
Foundry is Microsoft's hub for sales of AI models and tools to build and deploy automated AI agents. It's distinct from the company's Copilot-branded products, which represent Microsoft's main vehicle for selling AI tools to office workers.
Customer Resistance
Corporate customers have complained that it's hard to measure savings from using AI for routine tasks and note that the technology still makes mistakes that can be costly, according to the Information. The publication cited the example of private equity firm Carlyle, which started using Copilot Studio last year but later reduced spending on the tools after having trouble getting the AI to reliably connect data from other applications.
Carlyle representatives told Microsoft they were having difficulty getting the AI to tap data from applications such as Salesforce's customer relationship management app, according to someone with direct knowledge of the situation cited by the Information. The change was part of a broader company effort to choose which AI tools to keep spending on, though a Carlyle spokesperson said the firm is still increasing overall AI spending.
Analyst Response
In a research note, Jefferies wrote that the Information "completely missed the point in its article," adding that its analysts had spoken with Microsoft's management. Jefferies said Microsoft urged investors to focus on accelerating remaining performance obligations, an indication of future revenue. The analysts also said their checks showed robust adoption of Microsoft's Copilot line of AI assistants.
Much of the demand for Microsoft's AI computing power has come directly from OpenAI, a relationship separate from the Foundry model marketplace. Microsoft and other Big Tech companies have been spending heavily on chips, servers and data centers to fuel demand for AI computing, but signs are emerging that the market is growing skeptical about whether costs are justified by technology that remains unproven in many business applications.
