The Manager Is the Copilot Rollout: Why Middle Management Is Your Real Adoption Engine

Twenty million paid Copilot seats. And roughly a third of the people sitting on those seats actually use them.

Microsoft crossed 20 million paid Microsoft 365 Copilot seats this quarter. The workplace adoption rate — meaning people who actually open Copilot and do something with it — sits at 35.8%. Two out of every three assigned users are ghosting a license their company is paying full price for.

Every organization I work with has the same story underneath that number. They ran a launch. They sent the announcement email. They put on a lunch-and-learn. They stood up a Teams channel. Then usage flatlined by week six and the CIO started asking uncomfortable questions.

Here is the part that almost every internal comms plan gets wrong. The audience for your Copilot campaign is not the end user. It is the manager sitting one level above them.

The Data Everyone Ignores

Buried in nearly every recent Copilot adoption study is a finding that most rollout teams walk right past: when a direct manager actively uses Copilot, models it in meetings, and sets a visible expectation for the team, adoption in that team jumps to somewhere between 70 and 80 percent. When the manager is neutral or absent, adoption stalls in the twenties regardless of how much training you push at the team members.

That is not a small variable. That is the variable.

Think about how most Copilot rollouts are structured. Comms goes to "all users." Training is offered to "anyone who wants to attend." The champion network is built from volunteers who are usually individual contributors. Nobody has an explicit conversation with the front-line manager population about what is expected of them as leaders of Copilot-enabled teams. So they default to the same behavior they had before: tools come and go, this one will probably fade too.

Meanwhile the end user reads the launch email, opens Copilot once, gets a mediocre first result, and closes the tab. Nothing in their environment — their standups, their one-on-ones, their weekly reviews — tells them the tool is a real part of how their team now works. So it isn't.

Why This Is a Change Management Problem, Not a Training Problem

Most Copilot rollout plans I audit spend eighty percent of the budget on training content and twenty percent on communication, and treat the two as separate workstreams. That is backwards for what you are actually trying to achieve.

Copilot adoption is a behavior change, not a skill gap. The typical assigned user already knows enough to type a prompt into a box. What they lack is a reason to change how they open their morning, how they draft a document, how they prepare for a meeting. Those habits are shaped by their manager and their peers, not by a training video.

The mistake is running a technology campaign when what you need is a behavior campaign. A technology campaign says "here is the tool, here is what it does, here is how to use it." A behavior campaign says "here is what your team now does differently, here is how your manager will reinforce it, here is what shows up in your one-on-one because of it." Those are very different pieces of communication and they land with very different people.

The Manager-First Comms Playbook

If you are running the executive Copilot adoption plan for a mid-size organization, restructure your comms around the manager as the primary audience. Not additional. Primary.

Brief managers before you brief their teams. Two to three weeks before broad launch, run a manager-only session — sixty to ninety minutes, live, small groups. Not a training on features. A working session on how their team's workflow changes, three specific use cases their function should be doing in Copilot within thirty days, and what you expect them to do to reinforce it. If they walk out unable to articulate "here are the three things I want my team using Copilot for by end of month," you have not done your job.

Give managers a talk track, not talking points. The comms team's job is to draft the actual language a manager can use in their next team meeting and their next one-on-one. Two paragraphs each, in plain language, that they can paste into their own notes. Most managers will not compose this themselves. If you don't hand it to them, the message doesn't get delivered.

Put Copilot on the standing agenda. Work with a small set of pilot managers to add a two-minute "Copilot moment" to their weekly team meeting for the first sixty days — one person shares one thing they did with Copilot that week that saved time. Nothing else changes usage behavior faster than watching a peer, in your team, on your video call, describe a real workflow.

Track adoption at the team level, not the org level. The single most useful adoption report you can produce is not the tenant-wide dashboard. It is a leaderboard, by manager, of active Copilot usage in their team over the last thirty days. Show it to VPs. Managers will find out fast that their number is visible, and that alone reshapes their behavior. This is not surveillance — it is accountability, framed correctly. You would do the same thing for a new safety procedure or a new CRM.

Recruit champions from the manager population, not just individual contributors. A champion who is also a people-leader carries different weight than one who isn't. They can model behavior and set expectations. The individual-contributor champion network still matters — but a champion program without any managers in it is a comms channel to nowhere.

What to Push For This Week

If you take one thing from this piece into your Monday, make it this: go pull your Copilot usage data and cut it by manager, not by department or by license type. Rank managers by percentage of their team actively using Copilot in the last thirty days. You will find that most of your adoption is concentrated under a handful of leaders, and most of your dead-zone teams roll up to managers who have not personally used the tool in weeks.

That is your rollout status report. That is also your intervention list. The next thirty days of your change management work should be aimed squarely at the bottom of that leaderboard — not with more training content pushed to their team members, but with a focused conversation with the manager themselves about what they are expected to model.

The organizations that will look like Copilot success stories a year from now are not the ones with the best training library. They are the ones that figured out, early, that the manager layer is the change program. Everything else is warm-up.