The pitch usually arrives twice. First as a banner inside Word or Teams, then as an email from whoever sells you Microsoft licensing, and it is a good pitch. You try it on a real task, ask it to summarize a meeting you missed or turn four rambling emails into a clean client update, and it does a genuinely useful job in about eight seconds. You think, well, that was easy.
Then you open the quote. Because the question was never whether the tool is impressive on one task. The question is whether it is worth paying for every month, for forty people, some of whom will use it daily and some of whom will forget it exists by Thursday. That is a per seat math problem, not a technology problem, and the answer is genuinely different depending on what a given person does all day. So let us do the math honestly, role by role.
What You Are Actually Buying, and What It Costs
Microsoft 365 Copilot is an assistant built into the Microsoft apps you already pay for. It writes and rewrites in Word and Outlook, summarizes and catches you up in Teams, builds first-draft slides in PowerPoint, and answers questions about files stored in your organization’s SharePoint and OneDrive. SharePoint and OneDrive are simply where Microsoft 365 keeps your company’s documents.
On price, take it straight from the source. As of this writing in 2026, Microsoft’s own pricing pages list the Microsoft 365 Copilot add-on for enterprise at 30.00 US dollars per user per month with an annual commitment, and Microsoft 365 Copilot Business, the small and mid-sized business version, at 21.00 US dollars per user per month, currently promoted at 18.00 dollars annually or 25.20 dollars month to month. Microsoft also sells bundles, listing Business Standard with Copilot at 23.50 dollars and Business Premium with Copilot at 32.00 dollars per user per month on annual terms. Every one of those add-on options requires a qualifying Microsoft 365 subscription underneath it, so check what you have before you compare.
Promotional pricing moves, so verify the number the week you buy. What does not move is the shape of the decision.
Do the Per Seat Math Before the Demo Wears Off
Multiply the seat price by twelve and you have the annual cost per person. At 30 dollars a month that is 360 dollars a year for one seat, or 9,000 dollars a year across 25 people. At 18 dollars it is 216 dollars a year per seat. Those are just Microsoft’s published prices times twelve, nothing clever.
Now run the break-even with your own numbers instead of somebody’s case study. Take a person’s fully loaded hourly cost, meaning wages plus payroll taxes and benefits. Divide the monthly seat price by that hourly cost. The result is how many hours per month Copilot has to save that specific person to pay for itself. Do it for a partner and do it for a receptionist and you will notice the two answers are nowhere close, which is exactly the point. Then be skeptical about the number you get, because time saved only becomes money when it gets spent on something billable or something you would otherwise have hired for.
Roles Where It Usually Pays Off Today
- Owners, executives, and anyone drowning in meetings. The Teams meeting recap alone is the strongest feature in the product. If you sit in six meetings a day and half of them could have been a summary, this is the clearest yes on the whole list.
- Salespeople and account managers. Turning a messy call into a follow-up email, catching up on an account thread before a renewal call, and drafting proposals from prior documents are all high frequency tasks with low accuracy risk, because a human is reading it before it goes out anyway.
- Marketing and internal communications. First drafts, tone rewrites, and repurposing one long document into six shorter ones. Volume work where the output was always going to get edited.
- Operations managers and executive assistants. Document heavy roles that live in Word, Outlook, and Teams simultaneously, and that are constantly asked to find where something was decided.
- Project managers. Status roll-ups, summarizing what changed since last week, and writing the update nobody wants to write.
Roles Where It Is Not Worth It Yet
- Anyone doing serious analysis in Excel. Excel is where the gap between the demo and daily reality is widest, especially against real spreadsheets with merged cells, external links, and fifteen years of history. Finance teams tend to be the most disappointed group.
- Frontline, field, and shift staff. Technicians, drivers, warehouse, retail, and clinical floor staff who barely open Word are not going to find 360 dollars a year of value in a writing assistant.
- Highly specialized technical work. Engineering calculations, coding, and regulated filings usually need purpose-built tools, and general assistance in a Word document is not the bottleneck.
- Anybody whose output cannot be reviewed by a human before it ships. That is not a Copilot rule, it is an AI rule. Treat it like a junior staff member. Fast, tireless, occasionally confidently wrong, and never the last set of eyes.
The Prerequisite Nobody Mentions in the Demo
Copilot answers questions using files the person asking already has permission to open. That sounds reassuring until you remember how permissions actually accumulate in a growing company. If half your SharePoint is shared with everyone because it was faster that way in 2019, Copilot will cheerfully surface salary spreadsheets and acquisition notes to anyone who asks a reasonable-sounding question. It is not a leak. It is your existing permissions, finally made searchable in plain English.
So before you buy seats, spend a couple of hours reviewing who can see what, especially in HR, finance, and leadership folders. This is worth doing whether or not you buy Copilot. It just becomes urgent the day you do.
How to Buy It Without Overbuying
- Start with a pilot group of five to ten. Pick from the roles in the “pays off” list, not the loudest volunteers. Mixed pilots produce mixed conclusions.
- Clean up file permissions first. Thirty minutes of prep prevents a genuinely uncomfortable conversation later.
- Decide what success looks like in advance. Pick two or three specific recurring tasks and check in ninety days whether those tasks actually got faster. Vague enthusiasm is not a renewal justification.
- Watch the commitment terms. Annual pricing is cheaper per seat and harder to walk away from. Month to month costs more and buys you the option to stop.
- Reassign seats instead of buying more. Someone in your pilot will not use it. Move that license rather than assuming everyone needs one.
- Write a one page usage policy. What can go into it, what cannot, and who reviews output before it leaves the building.
The Bottom Line
Copilot in Microsoft 365 is worth it for meeting-heavy and document-heavy roles right now, and not yet worth it for spreadsheet analysts, frontline staff, and specialized technical work. Buy it for the first group, skip the second, and revisit in a year, because it is the dumbest it will ever be today and the roles on the “not yet” list will not stay there permanently. The mistake is not buying it. The mistake is buying it for everyone at once and calling that a strategy.
If you want help sizing a pilot, checking your file permissions first, or just getting a straight answer about your specific license mix, talk to Harrison Ward Technology. We support businesses across Denton County and DFW, and we have no interest in selling you seats nobody will open.
Comments are closed