The contract vehicle decides your discount, your flexibility, and your exit. CSP and the Enterprise Agreement pull in opposite directions. This pillar gives the buyer side framework for choosing, splitting, and migrating between them in 2026.
The choice between Microsoft CSP and the Enterprise Agreement is not a price question. It is a portfolio question about how stable your seat base is, how much flexibility you need, and where Microsoft is steering the contract. This pillar frames the decision the way a buyer should.
CSP and the EA are two different commercial machines. They reward opposite behavior.
The EA rewards predictability. CSP rewards variability. Choosing badly means paying for the wrong one.
CSP is a subscription bought through a Microsoft partner. Under the New Commerce Experience, terms are monthly, annual, or triennial. Monthly term allows reductions. Annual term does not.
The EA is a three year volume agreement bought direct from Microsoft. The Enterprise Agreement program locks price and earns volume discount in exchange for a committed baseline.
Microsoft sets the enterprise floor at 500 users or devices. Below it, CSP or MCA is the practical route. The EA only earns its complexity well above the floor.
Microsoft CSP versus EA at a glance
| Dimension | CSP | Enterprise Agreement |
|---|---|---|
| Term | Monthly, annual, or triennial | Three years committed |
| Quantity flexibility | High on monthly, none on annual | Add only, no mid term reduction |
| Discount | Usually list or light partner margin | Committed volume discount and price lock |
| Buying channel | Through a partner | Direct with Microsoft |
| Best fit | Variable or growing seat base | Large stable seat base |
The cheaper vehicle depends on seat stability, not on a published rate.
A stable base of 2,400 or more users with predictable growth usually wins on the EA. The committed discount and the price lock outweigh the lost flexibility.
A volatile, seasonal, or fast growing base often wins on monthly CSP. You pay for what you use and avoid carrying committed seats you do not need.
Microsoft 365 Copilot is a per user add on on both vehicles, with prerequisites set out in the Copilot licensing guidance. While adoption is unproven, CSP lowers the commitment risk because you can scale Copilot seats without a three year lock.
The common advice, repeated by many resellers, is to pick the single vehicle with the lowest headline rate and move the whole estate onto it. We disagree. In nearly every estate we have modeled the lowest cost answer was a deliberate split, not a single choice, because no real seat base is uniformly stable. The buyer side move is to segment the estate into stable core, growth, and project pools, commit only the stable core to the EA for its discount, and run the variable pools on monthly CSP. Picking one vehicle for everything overpays on one pool to save on the other.
Source: Redress Compliance advisory engagement file, 2024 to 2025.
The vehicle question is not CSP or EA. It is how much of your estate is truly stable. Answer that honestly and the split writes itself. Answer it with optimism and you overpay either way.
The Microsoft Customer Agreement for enterprise, branded MCA E, is the contract Microsoft is moving toward.
MCA E is a digital evergreen agreement defined in the Microsoft Customer Agreement. It removes the fixed paper term and shifts the relationship to a rolling model.
MCA E removes the three year administrative burden, but it can also remove EA price protection. Read the price terms before accepting the migration at renewal.
The migration is almost always proposed at the EA renewal. Treat it as a negotiation point, not a formality, because the protections you give up are hard to win back.
The dual track is the move that beats a single vehicle in most estates.
Split the estate into stable core, growth, and project or seasonal pools. The segmentation is the work that makes the rest easy.
Review the pools each year. Move seats between vehicles as the base changes. The split is a living allocation, not a one time decision.
CSP is a monthly or annual subscription bought through a partner with flexible quantities. The EA is a three year volume agreement bought direct with committed quantities and a price lock. CSP trades discount for flexibility, the EA trades flexibility for discount.
Neither is cheaper by default. For a stable seat base of 2,400 or more users the EA usually wins on committed discount. For a volatile or growing seat base CSP often wins because you only pay for what you use each month.
Microsoft sets the enterprise threshold at 500 users or devices for commercial customers. Below that threshold CSP or the Microsoft Customer Agreement is the practical vehicle. Many mid market buyers sit just below the line and should not be pushed into an EA.
It depends on the term you chose. Monthly term CSP allows reductions at each monthly anniversary. Annual term CSP under New Commerce Experience locks the quantity for twelve months, so the flexibility advantage only applies to the monthly term.
Microsoft has narrowed EA eligibility and is steering many customers toward CSP and the Microsoft Customer Agreement. The EA is not gone for large enterprises, but the renewal is increasingly where Microsoft proposes a move to a different vehicle.
The Microsoft Customer Agreement for enterprise, branded MCA E, is a digital evergreen contract that replaces the paper EA. It removes the fixed three year term but also removes some EA price protections, so read the terms before accepting the migration.
Yes. A dual track is common and often optimal. Keep the stable core seat base on the EA for committed discount and run growth, seasonal, and project seats on CSP for flexibility. The split is the lever, not the choice of one vehicle.
No. The EA can include Microsoft Unified Support as a paid layer, while CSP support is delivered by the partner. Strong partner support can match or beat Unified Support, so evaluate the named partner, not the channel label.
Microsoft 365 Copilot is available on both vehicles as a per user add on. CSP lets you scale Copilot seats up and down more freely, while the EA locks the committed Copilot quantity for the term. For an unproven rollout CSP reduces the commitment risk.
Decide the split before the rate. Segment the seat base into stable core, growth, and project pools, then map the stable core to the EA and the variable pools to CSP. The vehicle decision is a portfolio decision, not a single answer, and the dual track usually beats committing the whole estate to one model.
Microsoft renewal moves, the EA framework, the M365 SKU framework, the Copilot framework, and the buyer side moves across the full Microsoft estate.
Used across more than five hundred enterprise engagements. Independent. Buyer side. Built for procurement leaders choosing a contract vehicle.
Engage independent buyer side Microsoft negotiation services. We do not resell. We do not implement. We sit on your side of the table.
See engagement scope, comparison vs Big4 and resellers, and the buyer side framework.
Visit page →Buyers ask which vehicle is cheaper. The honest answer is that the question is wrong. Match each pool of seats to the model that fits it, and the estate gets cheaper than either single answer.
