Jefferies Maintains Buy Rating on Microsoft, Bullish on Copilot Upgrades Broadening Monetization Potential
nashnova research
Jefferies analyst Brent Thill reiterated a Buy on Microsoft with a $575 target, arguing three new Copilot features widen both subscription and consumption monetization — but warning that enterprise AI budget discipline remains the key variable.
What exactly did the new Copilot upgrade?
Microsoft's refreshed Copilot adds three features — Cowork, Code, and Autopilot — extending coverage from simple assistance to end-to-end delegated work.
Autopilot — which lets AI run tasks proactively, continuously, and in loops without step-by-step human prompts — is the headline change. It moves AI from "ask-and-answer" to "run the whole workflow."
This means → both the frequency and intensity of enterprise AI usage could rise, giving Microsoft more billable surface area.
Why can Microsoft compete with Meta on enterprise AI?
Thill frames the Autopilot-equipped Copilot as an enterprise counterpart to Meta's Muse.
His core argument: Microsoft already has deep integration with enterprise proprietary data and the M365 workflow stack.
In plain terms = Meta's AI leans consumer-side; Microsoft holds the enterprise's email, documents, identity, security, and compliance layers — a full stack that is hard to displace.
Does a stronger product automatically mean more spending?
Thill flags a clear risk: stronger features ≠ meaningful consumption growth.
Enterprises are tightly managing AI budgets, restricting usage to workflows with clear ROI. Microsoft's own credit-approval and usage-tracking mechanisms confirm that consumption will be actively governed.
This means → even if Copilot's capabilities improve, spending is gated by enterprise budget controls, not by Microsoft's release cadence.
How does monetization play out, and what is the key to watch?
Thill expects consumption growth to follow a gradual ramp, not a step-change.
The Cowork + Code + Autopilot combination widens both subscription revenue and usage-based billing, lifting the long-term ceiling on average revenue per user (ARPU).
This reflects a deeper call: the pace at which AI-agent workflows move from pilot to production will determine whether this thesis delivers — the loosening or tightening of enterprise AI budgets is the central verification point.
市场有风险，内容仅供研究参考，不构成投资建议。