Service 02
Most Copilot seats sit idle. Yours can be measured.
Sustained weekly Copilot use settles at 30 to 55 percent of purchased seats in the first year. That gap is visible in your own admin center, and it is fixable without buying anything else.
The number most companies have not looked at
Microsoft 365 Copilot lists at $30 per user per month, and roughly half the seats a company buys never turn into weekly habit. At 50 seats that is about $18,000 a year producing value for around twenty people. The effective cost per genuinely engaged user lands somewhere between 1.8 and 3 times list price. None of this is hidden. It is in your usage reports.
Work out your own number
Move the sliders to your seat count and a realistic adoption rate. If you do not know your adoption rate, start at 40 percent, which sits in the middle of the observed range. The figure this produces is an estimate from list pricing. The real one lives in your tenant, and getting it takes about ten minutes.
Why adoption stalls
Copilot adoption stalls for reasons that have nothing to do with the model. The four we see most often are poor data hygiene, so answers come back wrong and trust collapses early. Permission chaos, so security says no. No role-specific training, so people try it once on a generic prompt and stop. And no owner, so nobody is accountable for whether it works.
What an engagement looks like
A Copilot enablement engagement runs four to six weeks and starts at $4,500. It begins with measurement, not with training, because training people to use a tool that is blocked or returning bad answers wastes everyone's time.
- Week 1. Baseline. Usage and adoption telemetry pulled from your tenant, broken down by department and role, with named low-adoption groups.
- Week 2. Clear the blockers. Permission and data hygiene issues that make Copilot unsafe or unhelpful. This often overlaps with AI governance work.
- Weeks 3 to 4. Role-specific enablement. Three live sessions built around what your finance, operations, and sales people actually do, not a feature tour.
- Day 60. Re-measure. The same report, run again, so the change is a number instead of a feeling.
What we will not promise
We will not promise a headcount reduction, and we will not guarantee an accuracy figure. AI output is probabilistic, so any engagement we run assumes a human reviews what it produces before anyone relies on it. That assumption is written into the agreement. If a vendor is promising you certainty here, read their contract carefully.
Questions
Common questions
How much of our Copilot investment is actually being used?
Typically 30 to 55 percent of purchased seats become sustained weekly users in the first year. We measure your specific figure from Microsoft 365 usage reports, broken out by department, before recommending anything.
Do we need to buy anything to run an assessment?
No. The measurement uses reporting already included in your Microsoft 365 tenant. If you have Copilot licenses, the adoption data already exists.
What does Copilot enablement cost?
Engagements start at $4,500 as a fixed fee, covering baseline measurement, blocker remediation, role-specific training, and a re-measurement at 60 days.
Is Copilot worth it for a small business?
It depends entirely on adoption, which is why we measure before advising. For a company where half the seats go unused, the honest answer is often to cut seat count and enable the remaining users properly, not to buy more.
Can you help with other AI tools, not just Copilot?
Yes. Workflow automation, document processing, and internal chat assistants are all in scope. Copilot gets the focus because most companies are already paying for it.
Also from Kildaris
Next step
Thirty minutes, and you will know where you stand.
No pitch deck. We look at what you are running, tell you what we would do first, and you decide whether that is worth paying for.