Accenture to move 743,000 employees to Microsoft 365 Copilot after pilot with 89% usage

Accenture will roll out Microsoft 365 Copilot to all 743,000 employees after a large pilot that showed unusually high engagement. For Microsoft, this is not just a big contract but a rare public case that is supposed to prove to the market that enterprise AI can deliver measurable returns rather than remain an expensive add-on to an office suite.

Scale and reach of the rollout

Microsoft and Accenture are expanding the earlier Copilot deployment plan from 300,000 people to the entire global workforce: roughly 743,000 people in more than 120 countries. Microsoft calls it the largest enterprise deployment of Copilot. The scale matters for more than the record: this is a company where digital tools get a quick usefulness test, because consultants, managers, developers and internal service teams use them every day.

What stands out is not the size of the licensing but actual usage. Among the 200,000 employees who have been working with Copilot for some time, monthly activity reached 89%. For enterprise software, and especially for a paid AI add-on at $30 per user per month, that is a very high figure. Even more important, 97% of pilot participants said Copilot helps them complete routine tasks faster, in some scenarios up to 15 times faster. And 53% noted a noticeable gain in productivity.

Why the pilot worked

Accenture did not switch Copilot on for hundreds of thousands of accounts at once. The company took a more expensive but workable path: first a small pilot for top executives, then an expansion to 20,000 users, tuning of access and data governance policies, and only then a phased large-scale launch. In parallel, it ran separate work on training, use cases and internal sharing of practices. This is an important detail: in enterprise AI, it is not only models that fail but also organizational habits.

• The pilot started with several hundred executives
• Reach was then expanded to 20,000 employees
• Access and data governance rules were refined before scaling
• Separate training sessions were held for executives and teams
• The company launched an internal community where employees shared use cases

This approach, it seems, produced the high adoption rate. Accenture says plainly that value appears not when a license is issued, but when people understand where Copilot saves them time, how far they can trust it and how it fits into daily work. For the market, this is an uncomfortable but important conclusion: enterprise AI requires not only a subscription budget but also separate operational discipline.

"If

Copilot did not bring real benefit, our people simply would not use it," said Accenture CIO Tony Leraris.

Why this matters to Microsoft

The Accenture case matters to Microsoft for financial reasons. The company has more than 450 million commercial Microsoft 365 users, but as of publication on April 28, 2026, only about 3% of that base paid for the Copilot add-on. At $30 per user per month, even a small increase in conversion could bring billions in additional revenue. So Microsoft needs more than license sales; it needs stories in which AI genuinely enters the daily workflow.

The context for the company is not easy: at the time of publication, Microsoft shares were down about 12% year to date, and investors were increasingly asking when the huge AI spending would begin to show up clearly in revenue. The Accenture deal gives Microsoft three arguments for the market: a large reference customer, a clear deployment method and a set of figures it can show other corporations. In parallel, the company is developing Copilot into a multi-model platform and adding mechanisms such as Critique, where answers can be double-checked by different models.

There is also a practical effect. Avanade, the joint venture tied to Accenture and Microsoft, uses Copilot in its D3 sales analytics platform: the system pulls together internal data, industry context and external information into short briefs for salespeople. According to the company, active D3 users create 43% more sales opportunities than colleagues without the tool. If that metric holds at scale, it will be one of the strongest commercial arguments for enterprise AI this year.

What it means

The Accenture story shows that the enterprise AI market is entering a new phase: what matters now is not the mere fact of buying a model, but the ability to build it into processes, training and access control. If the adoption figures are confirmed at full scale, Microsoft will have a powerful case for selling Copilot to other large companies, and businesses will get a more realistic template for launching AI without the "bought the licenses and forgot" effect.