Healthcare networks carried the most avoidable licence overlap of any regulated sector, and licensing to workforce profile rather than a blanket standard cut a renewal roughly 30 percent
The renewal is the one moment the committed quantities and the mix can be reset. Treated as a rubber stamp, it locks in three more years of overlap.
Prepared by Redress Compliance · August 19, 2026 · Microsoft agreement renewals in regulated sectors. 20 to 30 renewals worked, 2024 to 2025.
Executive summary
Suite overbuy: 20 to 35 percent of users held the top suite where a lower one plus a targeted add on covered the role. Standardizing everybody upward is the expensive default.
Duplicate security: standalone tools renewed alongside the security stack that had already replaced them. The suite includes the capability and the estate keeps paying for it twice.
Shift and shared clinical staff were licensed as full named users rather than as frontline workers, which is a fraction of the cost for the same access.
The work started with a profile census, not a price ask. Every user mapped to a role, then to the leanest licence that covered it.
What is actually being committed?
A three year commitment priced on a committed user or device count, reset only at renewal. The rules sit in the product terms, and the committed quantity is the number the whole agreement is built on.
Healthcare networks are complex because clinical, administrative and shift workers all need different profiles. That complexity is usually resolved by standardizing everybody on the top suite, which is simple to administer and expensive to own.
Three profiles, three very different prices
| Profile | Who it fits | Where it goes wrong |
|---|---|---|
| Knowledge worker suites | Office and administrative staff | Applied to the whole population as a standard |
| Frontline profiles | Shift, shared device and clinical floor roles | Never mapped, so these users sit on knowledge worker seats |
| Targeted add ons | A specific capability a lower suite lacks | Skipped in favour of upgrading the whole seat |
| Service accounts | Nothing, they are not people | Licensed as full named users because nobody flags them |
The frontline profile is documented on the frontline worker page, and the suite comparison on the plans and pricing page.
What produced the 30 percent?
A profile by profile census rather than a discount request. Every user mapped to a role, then to the leanest licence that covered that role, before anybody opened a commercial conversation.
That sequence matters. A discount applies to whatever quantity and mix you bring to the table, so bringing a corrected mix is worth more than bringing a harder negotiator.
The frontline profile is the largest single lever in a clinical estate. Shift and shared device staff genuinely do not need a knowledge worker licence, and the mapping is a data exercise rather than a judgement call. The comparison sits in the suite comparison guide.
The Microsoft agreement renewal guide
Workforce profile licensing, frontline seat math, and the buyer side moves across a renewal.
Get the brief →What 20 to 30 regulated sector renewals showed
Across roughly 20 to 30 Microsoft Enterprise Agreement renewals worked in regulated sectors between 2024 and 2025, the healthcare networks carried the most avoidable licence overlap. Three patterns recur.
- Suite overbuy: 20 to 35 percent of users held the top suite where a lower one plus a targeted add on covered the role.
- Duplicate security: standalone tools renewed alongside the security stack that had already replaced them.
- Clinical seat waste: shared and shift based staff licensed as full named users rather than frontline workers.
The renewal is the one moment the committed quantities and the mix can be reset. Treated as a rubber stamp, it locks in three more years of the same overlap.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call, so you can test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
Where does the duplicate security spend hide?
In renewals that were never revisited after the suite absorbed the capability. The top suite already includes a broad security stack, so standalone tools covering the same ground should be retired rather than renewed alongside it.
These renewals are usually small individually, owned by a different team, and invisible on the agreement itself. That is exactly why they survive three year cycles without anybody deciding to keep them.
Retire, do not just note
An overlap identified and left in place is an overlap paid for. The unbundling question, and which capabilities genuinely need a standalone tool, is worked through in the security unbundling guide.
Watch the briefing · 4:45Your 2027 Renewal Is Not Your 2024 RenewalThe discount levels are gone, the suites cost more, and support multiplies both.
How do the savings survive the term?
By keeping the profile census current rather than treating it as a renewal project. Roles change, people move between shift and desk work, and the mapping drifts within months of being built.
A census that is refreshed quarterly holds the saving across the full three years. One built once at renewal decays into the same overlap it corrected, and the next renewal starts from scratch.
A standing rule beats a periodic sweep
Service accounts deserve a standing rule rather than a periodic sweep. They should never be licensed as full named users, and nothing in the estate will flag one that is.
What the renewals measured, 2024 to 2025
Two cuts of the engagement file, both about the mix rather than the rate.
Holding the top tier where a lower one plus a targeted add on covered the role they actually performed.
Achieved by licensing to workforce profile rather than by standardizing the whole population upward.
Neither number came from a discount. Both came from a census that took weeks and was worth more than any commercial argument available at the table.
Your first five moves
- Run a profile by profile census before any price conversation, mapping every user to a role and then to the leanest licence that covers it.
- Move shift and shared device staff onto the frontline profile, which is the largest single lever in a clinical estate and a data exercise rather than a judgement call.
- Retire the standalone tools the suite already replaced, because identifying an overlap and leaving it in place is an overlap you keep paying for.
- Strip service accounts out of the named user count entirely, and make it a standing rule rather than a periodic sweep.
- Refresh the census quarterly so the mix holds across the term. The Microsoft practice builds it before the renewal, the renewal playbook and its companion paper carry the calendar, the 2027 series covers what changed, and the timing brief sets the window.
Frequently asked questions
What is being committed in this agreement?
A three year commitment priced on a committed user or device count, reset only at renewal. The committed quantity is what the whole agreement is built on.
What produced the 30 percent?
Licensing to workforce profile rather than standardizing everyone on the top suite. The work started with a profile census, not with a price ask.
How much suite overbuy is normal?
Between 20 and 35 percent of users held the top suite where a lower one plus a targeted add on covered the role. It is the most common finding in a regulated estate.
Which staff fit the frontline profile?
Shift, shared device and clinical floor roles. They genuinely do not need a knowledge worker licence, and the profile costs a fraction of one.
What is duplicate security spend?
Standalone tools renewed alongside the suite security stack that already replaced them. They are small individually, owned elsewhere, and invisible on the agreement.
Should service accounts be licensed?
Never as full named users. It deserves a standing rule rather than a periodic sweep, because nothing in the estate flags one that has been licensed incorrectly.
Why does the census come before the discount?
Because a discount applies to whatever quantity and mix you bring. Bringing a corrected mix is worth more than bringing a harder negotiator to the same table.
How do the savings survive three years?
By refreshing the census quarterly. Roles change and the mapping drifts within months, so a census built once at renewal decays into the overlap it corrected.
Is the renewal the only moment to reset?
For the committed quantities and the mix, yes. That is what makes a rubber stamped renewal a three year decision rather than an administrative step.
Where should the work start?
With the role mapping, because it is the input to everything else. The profile decides the licence, the licence decides the quantity, and the quantity decides the commitment.