Microsoft Charges More For Copilot While Only 1% Use It Weekly

With just 1% of M365’s 477 million users actively engaging weekly, Microsoft raised Copilot prices to $30 per seat anyway

Annemarije de Boer Avatar
Annemarije de Boer Avatar

By

Image: Deposit Photos

Key Takeaways

Key Takeaways

  • Only 1% of Microsoft 365’s 477 million users actively engage with Copilot weekly.
  • Microsoft raised M365 pricing while Copilot’s effective cost runs three times its listed price.
  • Anthropic’s Claude now powers Copilot features, meaning users pay for integration, not proprietary AI.

The numbers Microsoft would prefer enterprise buyers skip past are hiding in plain sight. Out of roughly 477 million commercial Microsoft 365 users, about 20 million have paid Copilot seats, according to independent tracking reported by AI Business Weekly. That’s approximately 4.4% penetration. Layer in the uncomfortable second stat: only 20–30% of those purchased seats see weekly use, according to enterprise deployment analyses from Redress Compliance and AI Business Weekly.

Run the multiplication. Sustained weekly Copilot usage lands somewhere around 1% of the entire M365 customer base. Microsoft responded to this reality by raising Business Standard pricing from $12.50 to $14 per user per month — and still charging $21–$30 per user per month for the Copilot add-on on top.

Broad But Shallow

Fortune 500 “adoption” mostly means pilots and phased rollouts, not company-wide deployment.

About 70% of Fortune 500 companies have technically adopted Microsoft 365 Copilot, according to Lighthouse Global. That sounds impressive until you learn what “adopted” means in practice: small pilots, select departments, cautious rollouts. Enterprise licensing specialists put it bluntly, via Redress Compliance: “Real adoption trails the purchase. Companies buy broad and use narrow.”

Microsoft’s own definition of “active user” doesn’t help clarify things. One Copilot action in a month qualifies. As one adoption consultant noted, “If a user uses Copilot just once in a month and only in Copilot for Outlook, I personally wouldn’t really consider them an active user.” At 30% genuine weekly adoption, the effective cost per active user runs roughly three times the list price — like paying for a gym membership your whole office ignores.

The real twist: Microsoft’s core selling point is deep Graph integration — reading your emails, meetings, and files — not model quality. Anthropic’s Claude now runs inside Copilot’s Researcher, Copilot Studio, Copilot Chat, and Excel, with admin opt-in through the M365 admin center. Industry commentary suggests Claude outperforms Microsoft’s default model on most tasks, according to AI Business Weekly. You’re increasingly paying for a compliance wrapper and workflow shell, not proprietary AI brilliance.

What You’re Actually Paying For

The gap between free Copilot Chat and the paid add-on is real — but only matters if people actually use it.

Free Copilot Chat, already bundled with eligible M365 plans, handles web-grounded chat and reasoning over uploaded files. The paid tier unlocks Graph-connected reasoning across your full organizational data — emails, meetings, Teams chats, shared files. That distinction matters on paper. But when two-thirds of licensed users aren’t touching it regularly, the $30-per-seat invoice starts looking like a streaming subscription nobody in the house watches.

For context, ChatGPT claims roughly 400 million weekly active users, according to reporting cited by LinkedIn analyst Jukka Niiranen. Copilot has 20 million paid seats — and most of those sit idle. Satya Nadella has claimed weekly Copilot engagement now matches Outlook-level activity, but that benchmark applies to the minority of active users, not the broader M365 base.

Before the next renewal conversation, pull the Copilot usage dashboard and count sustained weekly active users — not monthly logins. The cost-per-active-user math is unforgiving. If those numbers don’t justify the seat count, that’s the real conversation worth having with Microsoft.

Share this

At Gadget Review, our guides, reviews, and news are driven by thorough human expertise and use our Trust Rating system and the True Score. AI assists in refining our editorial process, ensuring that every article is engaging, clear and succinct. See how we write our content here →