MPSA versus the EA, the agreement follows the buying pattern
The Microsoft Products and Services Agreement and the Enterprise Agreement answer different questions: MPSA answers buy now without committing, the EA answers commit volume for a lower price. The mistake we see most is inertia, buyers defaulting to the paper they already hold rather than the paper their buying pattern needs, and the wrong choice quietly costs 6 to 15 percent over a cycle.
Prepared by Redress Compliance · August 7, 2026 · Microsoft advisory. Based on 30 to 45 agreement reviews where MPSA was on the table, 2024 to 2025.
Executive summary
A third of estates held the wrong paper.
Across our reviews the agreement was a poor fit in about one estate in three, in both directions: MPSA buyers with steady annual volume left 6 to 15 percent on the table that an EA price level would have captured.
And EA holders with shrinking or irregular demand carried 9 to 18 percent of the enrollment as committed spend they could not use.
The agreement is a tool, not a status symbol.
The commitment line decides most cases. MPSA is a single perpetual purchasing agreement with no volume commitment, no term, and no annual true up: you owe nothing until you order, which protects falling or unpredictable demand.
The EA commits a baseline of qualified users or devices for three years, and the commitment is what unlocks the discount level and the price protection MPSA cannot reach.
Software Assurance splits the same way. On MPSA it is optional per purchase, added where the upgrade rights pay and skipped where they do not; on the EA it is bundled across the term, valuable for estates that upgrade often and wasteful for estates sitting on static versions.
The bundle is part of what the EA commitment buys, and part of what an idle EA wastes.
Cloud has left both agreements. Microsoft has pushed cloud purchasing toward CSP and the Microsoft Customer Agreement, and in close to 8 of 10 estates the cloud spend on MPSA had already migrated.
The 2026 decision is therefore three way: cloud subscriptions mapped to CSP or the MCA, perpetual on premises purchases without volume commitment kept on MPSA, and the EA held only where committed volume still earns its discount level.
The comparison, on the dimensions buyers ask about
| Dimension | MPSA | Enterprise Agreement |
|---|---|---|
| Commitment | None: buy as needed, no term, no true up | Three years, with a committed baseline and annual true up |
| Discount level | Transactional | Negotiated by volume, unreachable without the commitment |
| Software Assurance | Optional, per purchase | Bundled across the term |
| Price protection | None across years | Held for the full term |
| The best buyer | Irregular demand, no commitment wanted | Steady or growing volume that earns the level |
The EA is not the grown up choice; it is the committed one. The common framing treats MPSA as a stopgap for small buyers, and in a third of the estates we reviewed the EA was held out of habit while demand had flattened, with the committed spend partly unused.
MPSA paired with CSP for cloud would have cost less and carried no idle commitment. Hold the EA when the volume earns the discount, and step down when it does not.
The three questions that decide it
Is demand steady or irregular? Steady or growing favors the EA and its price level; irregular or falling favors MPSA with no commitment to strand.
How cloud first is the spend? If most of it is subscriptions, neither agreement is the right home, and the EA versus CSP decision guide and the Azure CSP versus EA analysis carry that fork in full.
Do you still buy perpetual on premises? MPSA remains the clean home for perpetual transactions without a volume commitment, the lane Microsoft's cloud program changes never closed.
The answers move over time, which is why the demand review belongs in every renewal cycle rather than in the one where someone finally questions the paper.
The Microsoft EA renewal playbook
The commitment decision inside the renewal: the sizing thresholds, the step down mechanics, and the levers that work whichever agreement wins.
Get the white paper →Software Assurance, per purchase or bundled
Software Assurance carries the upgrade rights and benefits that recur every year of an EA term, and its treatment is a real cost line in the comparison: on MPSA it is a per transaction decision.
Added where a product line upgrades often enough to pay and skipped where it does not, while the EA bundles it across the term whether each product needed it or not.
Estates that upgrade continuously get value from the bundle; estates on static versions pay for rights they never exercise, which is part of the 9 to 18 percent idle figure.
The annual count mechanics that police the EA side sit in the true up guide, and the renewal calendar that the whole decision times against in the 12 month renewal playbook.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across agreement reviews, 2024 to 2025
Across roughly 30 to 45 Microsoft agreement reviews between 2024 and 2025 where MPSA was on the table, buyers had defaulted to the paper they already held rather than the paper their buying pattern needed:
MPSA where steady volume earned an EA level, or an EA where flattened demand stranded the commitment.
The uncaptured discount on one side and the idle committed spend on the other, per cycle.
The demand review every renewal cycle is the cheapest control in the Microsoft estate: two years of purchase history classified as steady, growing, or irregular; the spend split into cloud subscriptions and perpetual licenses; the cloud mapped to CSP or the Customer Agreement.
And the EA commitment tested against whether the volume still earns the level.
The agreement you hold today is rarely the one your current buying pattern would pick, and finding that out costs an analysis rather than a cycle of overspend.
Your first five moves
- Pull two years of purchase history and classify the demand: steady, growing, or irregular, because the pattern picks the paper.
- Split cloud subscriptions from perpetual purchases, and map the cloud to CSP or the Customer Agreement, where it already belongs.
- Test whether the committed volume still earns the EA level, and price the step down honestly if it does not.
- Price the MPSA path against the EA path for the next cycle, with Software Assurance decided per purchase on one side and bundled on the other.
- Run the demand review every renewal cycle, not just this one. The Microsoft practice runs the comparison with you.
Frequently asked questions
What is the difference between MPSA and the Microsoft EA?
MPSA is a single perpetual purchasing agreement with no volume commitment, no term, and transactional pricing: you buy as needed across affiliates.
The EA is a three year enrollment committing a baseline of users or devices in exchange for a negotiated price level, bundled Software Assurance, and price protection across the term.
Which agreement is cheaper, MPSA or the EA?
It follows the buying pattern: steady annual volume on MPSA left 6 to 15 percent uncaptured that an EA level would have earned, while EA holders with flattened demand carried 9 to 18 percent of the enrollment as idle commitment.
The wrong choice costs 6 to 15 percent a cycle, and about a third of the estates we reviewed held the wrong paper.
Is MPSA still relevant in 2026?
In a narrower lane, yes: Microsoft has moved cloud purchasing toward CSP and the Microsoft Customer Agreement, and in 8 of 10 estates the cloud spend had already left MPSA.
What remains is a clean home for perpetual on premises licenses bought without a volume commitment, and for irregular buyers who want no committed baseline.
How is Software Assurance handled on MPSA versus the EA?
Optional and per purchase on MPSA, added only where the upgrade rights pay; bundled across the whole term on the EA.
The bundle rewards estates that upgrade continuously and quietly costs estates sitting on static versions, which is part of the idle spend carried by EAs whose demand no longer justifies them.
When should an EA holder step down to MPSA?
When the demand review shows the committed volume no longer earns the discount level: flattened or falling demand carried 9 to 18 percent idle commitment in our reviews.
The step down pairs MPSA for remaining perpetual purchases with CSP for cloud, costs less, and strands nothing, and the analysis belongs in every renewal cycle.
Where should cloud spend sit if not on MPSA or the EA?
On CSP or the Microsoft Customer Agreement, where Microsoft has steered cloud purchasing: close to 8 of 10 estates we reviewed had already migrated their cloud lines.
The 2026 structure is three way, cloud on CSP or the MCA, uncommitted perpetual purchases on MPSA, and the EA held only where volume genuinely earns the level.
