Copilot True-Up Is More Complex Than M365 True-Up
Enterprise Agreement true-up mechanics are well-understood for core M365 products. Microsoft 365 E3 or E5 seats are counted against the EA baseline annually, and the true-up invoice covers incremental deployment above the committed floor. Most enterprises have run enough M365 true-ups to understand the process, even if they have not always managed it well.
Microsoft 365 Copilot true-up introduces additional complexity that catches organisations off guard, for several reasons: Copilot seats may be under a separate commitment structure (adoption gate, phased deployment, pilot-to-production escalator) rather than standard EA true-up; Copilot's 38% active monthly usage rate creates a persistent gap between assigned seats and genuinely active users, making the standard true-up question — "how many seats were you deploying?" — commercially ambiguous; and the renewal conversation for Copilot is frequently structured by Microsoft as a seat count expansion discussion rather than a straight renewal, exploiting the true-up process to anchor a higher commitment level than actual usage justifies.
This article covers Copilot true-up mechanics in full: how seat counting works, the traps that inflate true-up exposure, the governance cadence required to manage Copilot seat count accurately, and how to use true-up data as a renewal negotiation lever rather than allowing Microsoft to use it against you.
True-Up Mechanics for M365 Copilot
Microsoft 365 Copilot is licensed as an add-on subscription to M365 E3 or E5. Under a standard Enterprise Agreement, Copilot seats follow the same true-up process as the underlying M365 subscription: annually, the enterprise reconciles its deployed seat count against the committed baseline, and pays for incremental seats at the contracted rate. The critical difference from core M365 true-up is that Copilot is almost always under a lower commitment level — because enterprises typically sign for a pilot cohort or an initial deployment tranche, not for their full seat population.
The true-up calculation for Copilot follows peak deployment mechanics, consistent with standard EA true-up. This means if you assigned 1,000 Copilot seats for a three-month pilot in Q1, reduced to 500 active users in Q2, and then grew to 750 by Q4, your true-up count is based on the highest point of deployment during the annual period — 1,000 seats — not the year-end count. Most organisations deploying Copilot in pilots do not configure seat assignment tracking against this mechanic, and discover at true-up that their highest pilot count becomes their committed baseline for the next year.
A 500-seat M365 E5 enterprise runs a 300-seat Copilot pilot in Q1. The pilot is deemed a partial success; the organisation plans to deploy 150 production seats in Q3. If they do not proactively revoke the 300 pilot seats before their Q2 true-up reference date, they will be billed for 300 seats — not 150. The excess 150 seats at £30/user/month represents approximately £54,000 in unnecessary annual spend. This scenario is extremely common and entirely preventable.
Seat Counting: Assigned vs. Active vs. Engaged
Microsoft's contractual true-up basis is assigned seats — the number of users to whom a Copilot licence has been allocated in the Microsoft 365 Admin Centre or via Entra ID group-based licensing. This is not active usage. A user who was assigned a Copilot licence during onboarding, never used it, and left the organisation six months later without their licence being revoked still counts as a deployed seat for the portion of the year they had an active assignment.
The gap between assigned, active, and genuinely engaged Copilot users is the largest source of preventable Copilot true-up cost in enterprise deployments. Understanding the three-level hierarchy is essential for true-up governance:
| Metric | Definition | Contractual Relevance | Benchmark (12-month post-deploy) |
|---|---|---|---|
| Assigned seats | Licences allocated in M365 Admin Centre; user has access to Copilot | True-up basis (peak count during period) | 100% of committed count (by definition) |
| Active users (MAU) | Users who opened a Copilot surface at least once in the calendar month | Not contractually defined; usage metric only | 38–45% of assigned seats at 12 months |
| Engaged users | Users with 10+ Copilot interactions per week (routine workflow integration) | Not contractually defined; productivity value metric | 15–22% of assigned seats at 12 months |
| Inactive seats | Assigned users with zero Copilot interactions in 30+ days | Count toward true-up peak; recoverable with proactive reclamation | 25–35% of assigned seats at 12 months |
The implication is that roughly a third of assigned Copilot seats in a typical enterprise deployment at 12 months are inactive — users who have not interacted with Copilot in a month or more. Those inactive seats are contributing to your true-up count and driving renewal baseline anchoring, without generating any productivity value. A systematic monthly seat reclamation process that identifies inactive seats, notifies users, and revokes licences for continued non-use is the single highest-value Copilot true-up management activity available.
True-Up Governance Cadence
The correct governance cadence for Copilot seat management combines monthly operational reviews with quarterly strategic reviews and a dedicated pre-true-up preparation process in the six weeks before your annual true-up date.
Monthly Operational Review (15–30 minutes)
Monthly, run the Microsoft 365 Admin Centre Copilot usage report (Reports → Usage → Microsoft 365 Copilot). This report shows assigned users, active users in the last 7 days, 28 days, and 90 days, and per-product activity breakdowns (Teams, Word, Excel, Outlook, etc.). Export the 90-day inactive list: any user with zero Copilot activity in 90 days should be flagged for licence reclamation. Send a notification to flagged users (or their managers): "Your Copilot licence has been inactive for 90 days. If you do not use Copilot in the next 30 days, your licence will be reclaimed and returned to the available pool." Execute reclamations at the next monthly review cycle. Track the reclaimed seat count — this is your governance dividend.
Quarterly Strategic Review (2 hours)
Quarterly, conduct a deeper Copilot ROI and deployment analysis. Review: active user trend (is MAU percentage improving or declining?); engaged user percentage (which user populations are integrating Copilot into daily workflows?); per-department adoption variance (identify departments where adoption is strong versus weak); and pipeline for next quarter's expansion or contraction. Use the quarterly data to update your deployment forecast — which feeds into your pre-true-up positioning. If your 12-month deployment trajectory shows consistent sub-40% MAU rates, that is evidence for a renewal commitment reduction argument.
Pre-True-Up Preparation (6 Weeks Before True-Up Date)
Six weeks before your annual Copilot true-up date, run a comprehensive seat reconciliation. The objective is to understand your precise true-up exposure — what is the peak deployment count for the year? — before Microsoft presents their count. Surprises in true-up reconciliation are always expensive; proactive knowledge is always cheaper.
Specific pre-true-up preparation steps include: export assignment history from M365 Admin Centre to identify peak deployment month; verify whether any pilot cohorts from earlier in the year created temporary high-water marks that are now materially above current active deployment; identify any seats that should have been reclaimed (departures, role changes, organisational restructuring) but were not; and prepare your true-up narrative — a documented deployment history that supports your count and pre-empts any Microsoft count discrepancy.
Commitment Structures and Their True-Up Implications
Not all Copilot deployments operate under standard EA true-up mechanics. Several alternative commitment structures have materially different true-up implications that must be understood before you sign.
Adoption gate structure: the initial commitment (e.g., 500 seats) converts to a higher commitment (e.g., 2,000 seats) upon hitting a defined adoption threshold (e.g., 70% MAU for 90 days). True-up implication: seats above the initial commitment are not billed until the gate is met. This is a favourable structure for buyers — if adoption does not meet the gate, you never trigger the expansion commitment. The trap is that gates are sometimes set at achievable-but-inflated levels, and achieving them triggers a sudden seat count increase that locks in a larger annual baseline.
Phased deployment structure: seat commitments increase in tranches over the EA term (e.g., 500 seats in Year 1, 1,000 in Year 2, 2,000 in Year 3). True-up implication: each tranche creates a new true-up baseline. If Year 1 adoption at 500 seats is below expectations, the Year 2 tranche increase to 1,000 seats may still be contractually required unless you negotiated a performance-based step-back clause. Always include step-back rights in phased structures.
Price cap structure: future expansion seats are priced at the original contracted rate (protecting against list price increases) in exchange for an expansion commitment. True-up implication: the expansion seats are voluntary but will be invoiced at the capped rate if you deploy above the original commitment. This is a valuable structure in an environment where Copilot list prices are increasing annually, but it requires careful modelling of whether you will actually deploy the expansion volume.
Pilot-to-production escalator: a 90-day pilot at no cost or low cost converts to a standard contract at the end of the pilot period, with the production seat commitment locked in at the start of the pilot. True-up implication: the production commitment clock starts at pilot conversion, and any seats assigned during the pilot that exceed the production commitment count toward true-up. This is the structure where the "peak deployment" mechanic is most likely to create surprise exposure.
Using True-Up Data in Renewal Negotiations
The annual Copilot true-up is not just a billing event — it is also the commercial moment when Microsoft initiates the renewal and expansion conversation. Microsoft's standard playbook is to use the true-up as an anchor: the true-up count becomes the starting point for the renewal seat conversation, and Microsoft's account team will typically recommend renewing at the true-up count (or above it, citing forecast adoption growth). The better-informed buyer uses the true-up data differently.
MAU data as a downward anchor: if your 12-month Copilot MAU rate is 38% (which is typical), you have legitimate grounds to argue that your renewal commitment should be based on actual engaged users, not the assigned seat peak. The argument is simple: "Our true-up shows 1,000 assigned seats, but our MAU data shows 380 active users. We are renewing for 400 seats — a slight uplift over active users to allow for adoption growth — not 1,000 seats." Microsoft will resist, but the data-driven position is more defensible than acquiescing to the true-up count as a natural renewal baseline.
Adoption rate as a pricing lever: sub-40% MAU rates are evidence that Copilot is not delivering the value Microsoft's productivity claims projected. Use that evidence to request price concessions at renewal — specifically, a lower per-seat renewal rate in exchange for a larger committed seat count. The framing: "We are prepared to commit to 750 seats for Year 2, but given our 38% MAU rate and the productivity gap versus Microsoft's modelled return, we need the Year 2 price at £X [below the current contracted rate]."
Competitive evaluation as leverage: Copilot faces real competition at renewal. GitHub Copilot (for developer populations), Google Workspace Gemini (in mixed-platform organisations), and standalone AI assistant deployments (ChatGPT Enterprise, Anthropic Claude for Business) are all credible alternatives that can be cited in renewal negotiations. Initiating a competitive evaluation 6–9 months before the renewal date — even a genuine one — creates pricing pressure that would not otherwise exist. For context, see our analysis of Copilot vs ChatGPT Enterprise on the competitive pricing and capability comparison.
The broader M365 true-up management framework — which covers both Copilot and core M365 products — is in our dedicated M365 true-up guide. The commercial architecture of Copilot renewal negotiations is covered in the Copilot seat pricing negotiation guide. For the full EA renewal framework within which Copilot renewals sit, see the EA renewal preparation guide for 2026.
Copilot True-Up Management Checklist
- Know your EA Copilot true-up date and add it to your Microsoft licensing calendar
- Confirm whether your Copilot seats are under standard true-up or an alternative commitment structure (adoption gate, phased, escalator)
- Configure monthly Copilot usage report export from M365 Admin Centre
- Establish 90-day inactivity threshold for seat reclamation triggers
- Run monthly seat reclamation process: notify, wait 30 days, revoke
- Track peak deployment count monthly — know your true-up exposure at all times, not just at year-end
- Conduct quarterly strategic review: MAU trend, engaged user percentage, department variance
- Begin pre-true-up preparation 6 weeks before annual true-up date
- Export 12-month assignment history to confirm peak deployment count before Microsoft presents theirs
- Prepare MAU and adoption data to use as renewal negotiation evidence 90–120 days before renewal
- Initiate competitive evaluation (ChatGPT Enterprise, GitHub Copilot, Gemini) 6–9 months before renewal as commercial leverage
- Document your right-sized renewal position (active users + growth buffer) before renewal conversation begins
Copilot true-up management is not a complex discipline — it requires consistent monthly operational attention and a clear renewal strategy built 3–4 months before the commercial conversation. The organisations that manage it well consistently renew at lower per-seat rates and more accurate seat counts than those who arrive at true-up without data and accept Microsoft's starting position. The return on 30 minutes of monthly governance is typically measured in five or six figures annually.