97% of Accenture employees report that Copilot helps them complete daily tasks up to 15 times faster. 89% active monthly usage among a cohort of 200,000 people tested. Microsoft has more than 450 million M365 enterprise users; only ~3% currently pay $30 per month for Copilot. Its shares are down 12% this year.
Microsoft is making its Microsoft 365 Copilot AI assistant available to all of Accenture’s approximately 743,000 employees. The contract extends the previous obligation Accenture to deploy the copilot 300,000 employees and expands it to the company’s full global workforce in more than 120 countries.
The timing is commercially important for Microsoft: Copilot is the company’s highest-profile enterprise AI product, but only about 3% of its more than 450 million Microsoft 365 enterprise users currently pay the $30-a-month Copilot premium. Microsoft shares have fallen nearly 12% this year as investors question whether the AI investment cycle will deliver the expected rapid revenue growth.
Accenture’s internal usage data, shared via Microsoft’s Newsroom, provides the most detailed real-world Copilot performance metrics published at scale by any enterprise. Among a cohort of 200,000 workers has been using Copilot for a long timemonthly active usage reached 89%, an extremely high adoption rate for any enterprise software, even a premium AI add-on.
97% of employees said Copilot helped them complete routine tasks up to 15 times faster. 53% reported significant productivity gains. And 84% said they would miss the tool if it were removed, reflecting habit formation rather than innovation.
Tony Leraris, CIO of Accenture: “If Microsoft 365 Copilot didn’t provide real value, our people simply wouldn’t use it, and our high adoption rate shows us it’s worth it.”
The rollout methodology is remarkable, as is the scale. Accenture didn’t just launch Copilot for 743,000 people at the same time. It started with a pilot of a few hundred senior leaders, expanded to 20,000 users, managed that cohort while improving data management and access controls, and then expanded in phases with a highly customized change management program, including one-on-one training for leaders, breakout sessions, and a structured internal community at Viva Engage where employees share use cases.
Leraris’s framework covers the lesson: “The real value from AI investments like Copilot doesn’t just come from making it work. It comes from investing in helping people understand how to use it, how to trust it, and how it fits their business.”
This is a direct rebuke to the deployment model, where enterprises purchase AI licenses and expect adoption to follow automatically. A commercial byproduct of the offering is Avanade’s D3 platform, a sales intelligence tool built by a joint venture between Accenture and Microsoft that uses Copilot to combine proprietary internal data, industry context and external sources into a real-time briefing for sales reps.
Research that once took days or weeks can now be produced in seconds. Avanade rolled out D3 to 25% of its resellers; active users generate 43% more leads than their non-using counterparts. This number, if at scale, makes D3 one of the most commercially compelling AI use cases published in 2026.
Accenture’s contract for Microsoft addresses a specific and well-documented problem. The company has more than 450 million Microsoft 365 enterprise users, the largest installed base of any enterprise productivity suite. Converting even a fraction of these users to the $30/month Copilot premium would represent significant incremental revenue at close to zero marginal cost.
But enterprise AI adoption has proven slower More than Microsoft’s initial predictions: Early Copilot deployments were characterized by high acquisition rates but low actual usage, as employees struggled to understand where the tool added value and where change management was inadequate.
Accenture’s presentation provides Microsoft with three commercially useful assets: a proof point for enterprise-wide adoption, a methodology blueprint for other large customers considering similar deployments, and a named reference to reference in every enterprise Copilot sales conversation over the next 18 months.
The wider context has been revised Microsoft-OpenAI partnershipthis gives Microsoft the flexibility to integrate multiple AI models into Copilot, including Anthropic’s Claude, rather than depending solely on OpenAI’s GPT family.
Microsoft has introduced a “Criticism” feature that checks outputs between models to improve accuracy. This multi-model strategy both reduces dependency on any one AI provider and allows Microsoft to direct various tasks to the best available model, which will become more important as enterprise customers demand more granular control over which AI systems handle sensitive workloads.
