Oxford, Ohio · Butler County & Greater Cincinnati
You are paying $30 a seat for software nobody opens
Microsoft 365 Copilot is thirty dollars per user per month on top of your existing licences. Industry reporting puts unused or barely-used seats somewhere between 30 and 60 percent. On a forty-person company that is somewhere north of five thousand dollars a year buying nothing at all.
This is probably you if
- You bought Copilot seats and cannot say who uses them
- Adoption spiked at launch and then flatlined
- The rollout was a licence purchase and an email
- People say they tried it and it was not useful
- Renewal is coming and you cannot justify the spend
Why it stalls
Copilot gets bought as a licensing decision and it is actually a habit-change problem. Everyone gets a seat, IT announces it, and then people go back to working the way they worked last month, because nothing about their actual Tuesday changed.
The standard fix is a training session covering every feature equally, which is close to useless. An hour of Copilot demos in the abstract teaches nobody what to do with the report they have to write on Thursday.
What actually moves the number
Training built around specific jobs people already do. Your estimator gets shown how it drafts a scope of work from a site visit note. Your office manager gets shown how it turns a meeting into actions. Your bookkeeper gets shown what it can and, importantly, cannot be trusted with.
That means starting by looking at how your team really works, then teaching four or five things that matter to them, rather than forty things that matter to Microsoft.
And where Copilot is the wrong tool
This is the part your licence reseller will not tell you. Copilot is good at drafting, summarising and searching across your own documents. It is not the answer for a repeatable process that runs the same way every week: that should be automated properly, not prompted by a person every time.
Plenty of the value in an engagement like this is working out which of your problems is a training problem and which is a build problem, and being honest about the split.
The same goes for the seats you bought elsewhere
ChatGPT Team, Gemini for Workspace, whichever AI subscription got approved last year: the pattern is identical. A handful of enthusiasts use it constantly, most people tried it twice, and nobody has measured anything. Worth auditing before renewal rather than after.
Questions people actually ask
- Can you just tell us whether to keep paying for it?
- Yes, and that is a reasonable engagement on its own. Usage data plus a short conversation with each team usually shows within a week whether the seats are earning their keep, and which ones should be reassigned or dropped.
- Are you a Microsoft partner?
- No, and that matters here. I get no margin on your licences either way, so I have no reason to tell you to buy more seats or keep the ones you are not using.
- We are twelve people. Is this worth doing?
- At twelve seats you are spending roughly four thousand three hundred a year. It is worth an hour to find out whether that is working, and if it is not, whether the money is better spent building the thing you actually needed.
Worth a conversation?
Fifteen or twenty minutes about how your business actually runs. No deck, no discovery fee. If I do not think there is anything here worth building, I will tell you that.
Book a callWorking with businesses in Oxford, Hamilton, Fairfield, West Chester, Liberty Township, Middletown, Monroe, Trenton, Mason and Cincinnati.
Last reviewed .