Your Copilot Licences Are 30% Used. Here Is the 4-Week Relaunch.
Bottom line: independent surveys put weekly active use of paid Copilot seats at 20 to 36%. Most organisations already trained once, and adoption stalled anyway. The fix is not a second kickoff webinar. It is a 4-week relaunch that starts with a licence audit and ends with a real usage number.
I get asked why Copilot adoption fails a second time more often than I get asked before the first rollout even starts. By the time someone reaches out, the licences are already paid for, and the invoice looks the same whether anyone opens Copilot or not. That is the part that should worry a finance team more than it usually does.
The number nobody puts on a slide
Here are the Copilot adoption rate statistics nobody puts on a slide: independent survey data puts weekly active usage of paid Copilot seats somewhere between 20 and 36%. Read that next to Gartner's research: of 187 IT leaders surveyed, 40% were piloting Copilot, and only 5% of pilot completers had scaled it by the end of 2025. Piloting was never the hard part. Scaling is.
Ask why Copilot adoption fails on the second attempt almost as often as the first, and this is where the answer starts. The first rollout's targeting was probably off, and nobody went back to check who actually got value from it and who never opened the app again.
I have written up the specific reasons rollouts go quiet in the first place on why Copilot rollouts stall, so I will not repeat them here. This page picks up after that diagnosis, once the licences are already sitting there half used and someone finally asks the awkward question out loud.
What a relaunch does differently
A relaunch is not the first rollout again, louder. It is a 4-week programme that starts with a licence audit (who actually has a seat, and how little they are using it), works with a smaller group on their own real tasks, and ends with a usage number pulled from the same data everyone already has, not a self-reported survey.
(I am specific about the audit-first part because it is the step most relaunches skip. Nobody wants to admit the first rollout's targeting was off. It usually was, and that is fine. Fix the targeting before you fix anything else.)
Week by week
4 weeks. 4 working sessions. 1 clear structure, built entirely around what the audit finds rather than a generic curriculum.
Week 1The audit and picking the 20
Before anything else, a licence audit: who has Copilot, who has actually opened it in the last 30 days, and which team has the most to gain from getting it right. From that, we pick 1 group, up to 20 people, with a real repeating task worth fixing. No homework yet. This week is diagnosis, and it is the week most relaunches skip in their rush to get back in the room.
Week 2Working session 1
The group works on Copilot inside their own files, on the task the audit flagged. Not a generic tour of features, actual work. For a finance-adjacent team that might be turning a messy Excel export into something a manager can read in 2 minutes. For an ops team it might be drafting the weekly status update in Outlook from a pile of half-finished notes, with Copilot doing the first pass and a person finishing it. Homework: use it on the same task at least twice before the next session, in Word, Excel, Outlook or Teams, whichever is where the actual work lives, and note exactly where it got stuck.
Week 3Working session 2
We go through what got stuck. This is usually where the real change happens, because the questions people bring after a week of real use are completely different from the ones they had on day 1. Someone got a mediocre first draft in Word and did not know how to push back on it. Someone tried Copilot on a spreadsheet with 40 tabs and it did not go well. We work through the specific failure, not a new example. Homework: the same task again, this time without help, plus 1 new one, such as prepping talking points for a Teams meeting straight from the agenda.
Week 4Working session 3, and a usage number
The last session covers what is left, then we pull a usage number from the same data source everyone already has (Microsoft's own default adoption metric is one use in 28 days, a low bar worth knowing about). That number is a measurement, never a guarantee, and I will say so on the call too.
What this is not
This is not the first rollout run again with a new date on it. It is not a second kickoff webinar, and nobody in the room is hearing about Copilot for the first time. It is smaller, more targeted, and built entirely around what the audit found, not around a generic curriculum written before anyone looked at the actual usage data.
If the diagnosis turns out to be something other than training (permissions, licensing, a tool that was never the right fit for that team) I will say that on the intro call rather than sell a relaunch nobody needs yet.
Questions I get about the 4-week relaunch
Why is Copilot adoption stuck at 30% usage?
Independent surveys put weekly active usage of paid Copilot seats at 20 to 36%, which means most organisations already trained once and adoption stalled anyway. Ask why Copilot adoption fails a second time and the honest answer is usually the same reason as the first: nobody went back to check who actually got value from it. For the specific reasons a first rollout goes quiet, see why Copilot rollouts stall.
What is a 4-week Copilot relaunch?
A 4-week programme that starts with a licence audit, runs working sessions with a focused group on their own real tasks, includes homework between sessions, and ends with a measured usage number instead of another launch event.
What happens in each of the 4 weeks?
Week 1 is the licence audit and picking the group. Weeks 2 through 4 are working sessions on real tasks, with homework between each one, ending with a usage number pulled from actual data.
Do we need a new licence audit before relaunching?
Yes. The relaunch is built around who actually holds a licence and how little it is used, not a generic re-run of the original curriculum. Skipping the audit is the fastest way to repeat the first rollout's mistakes.
What should we expect to measure at the end?
A usage number, pulled from the same data everyone already has. Microsoft's own default adoption metric is one use in 28 days. That is a measurement, not a guarantee, and I will not pretend otherwise on the call.
Ready to run the audit and pick the 20?
The full 4-week programme is The Second Wave, from $8,440 / €7,280, live online, up to 20 people. Leave your details and I will call you back.