Full narration of the briefing. Click a section heading to jump the player to that moment.
Copilot will be the largest single new line in most 2027 renewal proposals, and it will arrive at tenant scale, because that is how it is sold. I want to be clear at the start: this is not an argument against Copilot. Organisations are getting real value from it, and Microsoft passed twenty million paid seats by April 2026 with daily use up roughly tenfold in a year. It is an argument against buying it the way it is usually offered, which is for everyone, at once, at renewal, before anybody has proved which of your people will use it.
I am Tom, Claire is with me.
Start with the true cost, because the thirty dollar headline is only part of it. Copilot is thirty dollars per user per month with an annual commitment, and it sits on top of a qualifying Microsoft 365 licence. After the July 2026 rise, that means the all in cost is about sixty nine dollars per seat per month on E3, and about ninety on E5. Now multiply by your headcount and by twelve.
For five thousand users on E3 that is roughly four point one million dollars a year, all in. That is the number the business case has to carry, not the thirty.
Now the part your account team will not lead with, and it is well documented. Organisations that buy Copilot for their whole workforce typically find thirty to forty percent of those licences unused within the first ninety days. And where full tenant deployment was committed at a renewal without a validated pilot behind it, forty to seventy percent sits unused through the first year. Read the condition in that second number carefully, because it is the whole lesson.
The failure is not the product. The failure is committing at renewal scale before anybody knew who would actually use it.
The causes are consistent and none of them are the technology. Change management that never happened, so people were given a tool and no new way of working. Training treated as an email rather than an investment. Prerequisites missing underneath: permissions too loose, so the assistant surfaces documents people should never see, or content so disorganised that its answers are poor and trust evaporates in a fortnight.
And roles where there is genuinely no daily use case. Every one of those is knowable before you buy, which is exactly why a pilot is not a delay tactic. It is the cheapest possible way to find out which of them apply to you.
A pilot that settles the question has four properties. It is narrow: two or three job families, not a cross section of volunteers, because volunteers always look like success. It is long enough: ninety days, so novelty wears off and habit is what you are measuring. It is measured on use, not sentiment: weekly active users as a share of licences, and how many stay active in month three.
And it has a decision rule agreed in advance, in writing: this is the threshold at which we roll out, this is the threshold at which we stop. Set that number before you start, because after you have spent the money everybody discovers reasons the number was too strict.
Then buy it in a shape that matches what you actually know. Tranches rather than a tenant wide commitment: a first block sized to proven demand, with the right to add more later. A price hold on those later blocks for the full term, so being careful now does not cost you more per seat in year three, which is the concession most buyers forget and the one that quietly punishes discipline. Reassignment rights, so a licence follows the person rather than dying with the leaver.
And a genuine reduction right at anniversary, because the whole point of measuring adoption is being able to act on what you find.
The move from this briefing: if Copilot will be in your 2027 proposal, start the pilot now, because ninety days of real usage data is only leverage if it exists before the negotiation rather than after it. Two job families, one decision rule, written down. You will either walk in able to say exactly how many seats you want and why, or able to say no with evidence, and both of those are strong positions. Next session: what happens beside the seat, with Cowork, agents and Copilot credits.
See you there.
Redress Compliance works on contingency: our fee is 25 percent of what we save you. Nothing saved, nothing paid. Independent, buyer side only, never vendor funded. Want Redress to contact you? Reach out and we respond the same day.
Talk to a Microsoft negotiator