AI pricing needs to come down to increase adoption, says Jefferies' Brent Thill

Key Concepts

• AI Sales Growth Targets: Microsoft's reported adjustments to its targets for AI product sales.
• Growth vs. Sales Quotas: The distinction between overall business growth and specific sales targets.
• AI Demand Acceleration: The increasing customer interest and adoption of AI technologies.
• AI Agents: Software programs designed to perform tasks autonomously, often powered by AI.
• Pricing Model: The strategy companies use to price their products and services.
• Commoditization: The process by which a product or service becomes a commodity, leading to price competition and reduced differentiation.
• Enterprise AI: AI technologies specifically designed for business applications.
• Valuations: The assessment of a company's worth, often expressed as a multiple of its earnings or revenue.
• RPO Backlog: Remaining Performance Obligations, representing future revenue from existing contracts.

Microsoft's Response to AI Sales Growth Report

Microsoft has publicly refuted a report from The Information, which claimed the company had lowered its AI sales growth targets due to customer resistance to new products. Microsoft's official statement asserts that the report inaccurately conflated "growth" with "sales quotas" and that aggregate AI sales targets have not been reduced.

AI Demand and Adoption Trends

Brant Thill, a Jefferies analyst, shared insights from his work and observations at the Amazon conference in Las Vegas, indicating that AI demand is actually accelerating. He cited positive numbers from Snowflake and Salesforce as evidence of this trend. Thill noted that thousands of attendees at the conference are discussing the increasing adoption of AI and the future impact of AI agents, which are expected to generate more workloads for Microsoft and influence the adoption of its entire product portfolio. He stated, "We do not see a slowdown."

Supply and Demand Dynamics

Microsoft has reported a "50 plus percent RPO backlog," indicating that they are unable to keep up with demand and are facing supply constraints. This suggests a strong market for their offerings, contrary to the notion of customer resistance.

The Nascent Stage of AI Agents and Pricing Strategies

A key point raised by Thill, drawing a parallel to Amazon's comments at their conference, is that AI agents are still in their early stages of development, akin to "infants in a crib." He suggests that the software industry may need to adopt a "lower pricing model" to facilitate the widespread adoption of these agents. Once agents are live and integrated, this will drive further adoption.

Thill identified a "silver lining" in the reported challenges: many companies, including Salesforce and Atlassian, initially priced their AI offerings too high. His research suggests that a slight reduction in pricing is necessary to boost adoption. He emphasized that most enterprise AI products are less than a year old, and the typical product development cycle at Microsoft involves iterative improvements over one to two years. He expressed optimism for the rollout of AI agents in 2026.

Pricing Power and Commoditization Concerns

When questioned about the potential impact of lower pricing on margins and the long-term pricing power of these companies, Thill clarified that he is not suggesting Microsoft is currently lowering prices. However, he believes that charging a premium for new products is not always effective.

He highlighted the emergence of numerous "agent builders" and anticipates a "battle for who builds these agents." While acknowledging that commoditization will eventually occur, Thill argued that the industry is "way too early" for this to be a significant concern. He projected billions of agents in use, potentially reducing the need for human workers.

Thill believes that major platform companies like Amazon, Google, and Microsoft are well-positioned to weather commoditization due to their existing infrastructure. These platforms will serve as the "home" for AI agents, requiring ongoing maintenance and resources. He drew parallels to the internet and cloud SaaS booms of the past 30 years, suggesting that while some agent builders may not succeed, the core platform providers will remain strong.

Recalibrating Valuations in the Current Market

The discussion touched upon whether current valuations need to be recalibrated in light of AI's potential. Thill argued that software companies may be undervalued, especially if AI adoption accelerates as predicted. He views 2026 and 2027 as the "year of enterprise AI," anticipating significant revenue influx and accelerated growth for companies like Amazon, Google, and Microsoft (specifically Azure).

He suggested that current multiples are "pretty stable" and "should go higher." He contrasted this with the performance of application stack companies, noting that many are down year-to-date, while semiconductors are up significantly.

Conclusion

The prevailing sentiment from the discussion is that while a report suggested a slowdown in AI sales growth for Microsoft, the reality appears to be an acceleration in AI demand. The challenges lie in the early stage of AI agent development and the need for appropriate pricing strategies to drive mass adoption. Despite potential future commoditization, major platform companies are expected to benefit from the underlying infrastructure requirements of AI agents. Valuations for software companies may need to be re-evaluated upwards as AI adoption continues to grow.