Microsoft Licensing in 2026: What Changed While You Weren't Looking

Your Microsoft spend increased again this quarter.

Not dramatically. Just enough that it caught your eye during budget review. Office 365 seats, Azure consumption, some Dynamics subscriptions, maybe Copilot now. The usual.

Except when you dig into the detail, the line items don't quite reconcile with what you remember approving. And the per-user costs seem higher than they should be given your organisation's size.

If you're a CTO, IT General Manager, or technology leader in an Australian organisation with more than 200 Microsoft users, this pattern will feel familiar. You're not imagining it. Something structural shifted in how Microsoft licensing works, and most organisations are still operating on assumptions that no longer hold.

The discount you expected probably isn't there

Three years ago, scale meant better pricing.

Organisations with 500+ users expected meaningful volume discounts on Enterprise Agreements. More users meant lower per-seat costs. That was the trade-off for committing to three-year terms and annual true-ups.

That assumption no longer reliably holds for cloud services.

Microsoft has progressively moved away from tier-based volume discounting for Online Services. Pricing for Microsoft 365, Dynamics, Defender, and similar cloud products now sits much closer to list rates than it used to, regardless of user count.

On-premises licensing still behaves differently. Hybrid scenarios can too. But for cloud-first organisations in Australia, the EA pricing advantage largely disappeared.

This changes the entire commercial calculation. If an Enterprise Agreement doesn't deliver materially better per-seat pricing than CSP, why lock in three years?

Enterprise Agreement versus CSP in 2026

The gap between EA and CSP pricing narrowed considerably over the past 18 months.

In many cases, both sit close to Microsoft's published rates unless specific concessions are negotiated based on strategic factors or transaction size.

The decision between EA and CSP is now about structure and operational fit, not assumed discounting.

Enterprise Agreements still make sense when you need:
Three-year contractual certainty with annual payment cycles rather than monthly reconciliation.
Annual true-up rather than continuous seat-level adjustments.
The ability to negotiate bespoke commercial terms in genuinely large or complex environments.
Continued support for on-premises and hybrid licensing scenarios.

CSP makes sense when:
Your workforce size changes frequently and you need flexibility without penalty.
You want to avoid multi-year commitments for cloud services that evolve constantly.
You're cloud-first and don't need on-premises constructs.
You want pricing that's broadly comparable to EA without the commitment weight.

For mid-sized organisations and technology teams managing variable headcount, CSP is increasingly the better structural fit. For very large, stable organisations with significant on-premises complexity, Enterprise Agreements may still deliver value.

The key point is that pricing outcomes are no longer automatic. Scale alone doesn't secure better terms anymore.

One practical consideration that often gets missed is reversibility.
While organisations can move from CSP back to an Enterprise Agreement, doing so is not always frictionless. CSP structures, particularly under New Commerce Experience terms, can fragment commitment timelines and reduce the clean, centralised licensing footprint that EA negotiations rely on. This doesn’t make a return to EA impossible, but it does make timing, sequencing, and forward planning more important if an Enterprise Agreement remains a future option.

It's also worth understanding where complexity can arise when moving between models.

Microsoft doesn't charge exit fees or penalties for moving from CSP to an Enterprise Agreement. The friction comes from timing and structure rather than financial penalties. User-based licences can transition cleanly if aligned with CSP commitment end dates under New Commerce Experience terms.

Azure is more complex. Moving Azure subscriptions between CSP and EA often requires billing realignment, and in some cases, manual migration of resources. Reservations and savings plans typically don't transfer automatically between licensing models, which means timing these transitions poorly can result in lost commitments or overlapping costs.

These are operational and timing considerations rather than formal barriers, but they create real cost and effort if not planned properly.

Why Microsoft licensing became complicated quickly

Microsoft doesn't sell isolated products anymore.

It sells platforms designed to unlock future capability. Bundles that cross traditional product boundaries. Ecosystems where one purchase decision influences the next.

The commercial model reflects that complexity.

Per-user licences. Per-device licences. Consumption-based services. Capacity commitments. Reservation discounts. Enterprise true-ups with delayed cost exposure. CSP subscriptions with flexibility but material cost differences depending on term and partner margin.

This makes like-for-like comparison difficult and increases switching costs once you're operationally embedded.

That's manageable when value is clear and usage is understood. It becomes problematic when organisations pay for capability they don't use, or when licensing structures drift away from how the business actually operates.

What changed over the past 18 months

Several developments reshaped how medium and large Australian organisations approach IT procurement for Microsoft services.

Copilot introduced new cost layers
Copilot licences sit on top of existing Microsoft 365 subscriptions. Some organisations are seeing strong early value. Others committed to incremental cost before usage patterns or return on investment were well understood.

Security requirements pushed organisations into higher tiers
Regulatory expectations, cyber insurance requirements, and internal risk standards increasingly expect security capabilities that only exist in higher-tier licensing. What was optional two years ago is now often treated as baseline.

Azure consumption became harder to forecast
Multi-cloud strategies and workload migration increased variability in Azure spend. Commitment-based discounts remain valuable, but only where consumption patterns are stable enough to forecast with confidence.

Enterprise Agreement true-ups are surfacing later costs
User growth or licensing changes during EA terms can emerge as significant true-up liabilities. The commercial impact often appears long after the original operational decision was made.

CSP relationships are being re-examined
Many organisations are reassessing whether their CSP arrangements deliver value beyond billing and provisioning, particularly where strategic optimisation support is limited or absent.

The procurement imbalance that matters

Microsoft enters renewal discussions with complete visibility.

Usage data across your environment. Growth trends. Licensing history. Renewal timelines. Platform dependency levels.

Most organisations don't have equivalent insight into what comparable organisations are paying, which terms are genuinely negotiable, or how alternative licensing structures would perform over time.

That imbalance often results in renewals that feel pressured, opaque, and difficult to challenge.

If this feels complicated, it's because it actually is. The dynamics changed materially, and the gap between what organisations assume and what's commercially achievable widened. Understanding where you actually have leverage matters more now than it used to.

What effective ICT procurement looks like in 2026

Strong Microsoft licensing procurement starts with understanding actual usage rather than relying on entitlement counts.

It means separating immediate requirements from future optionality. Identifying dormant features. Being realistic about adoption timelines. Stress-testing growth assumptions against actual workforce behaviour.

It also means understanding your leverage points.

Time before renewal creates leverage. Structural alternatives create leverage. Willingness to model EA versus CSP properly creates leverage. Understanding what's genuinely negotiable versus what's theatre creates leverage.

The question is whether that leverage is being used, or whether renewals are treated as administrative exercises that happen under time pressure.

Common mistakes that quietly increase cost

Certain patterns appear consistently across Australian organisations managing Microsoft estates:

Rolling Enterprise Agreements without properly testing CSP alternatives.
Licensing for projected growth that never materialises.
Underestimating true-up exposure during the EA term.
Treating CSP partners as interchangeable when margin structures and support levels vary materially.
Comparing models on list price alone instead of total cost over time.

These issues arise in well-run organisations. Complexity and inertia create blind spots. Similar patterns emerge in technology contracts more broadly, where commercial terms drift away from operational reality over time.

What to focus on now

If you manage Microsoft licensing for a medium or large Australian organisation, a few things matter more than the rest:

Understand actual usage, not just licensed quantities.
Model EA versus CSP based on real workforce behaviour and growth patterns.
Test CSP options rather than assuming commercial equivalence.
Consider hybrid approaches where they genuinely fit operational needs.
Know your renewal timeline and create space to negotiate properly.

Microsoft renewal cycles align to vendor timelines, not yours. Negotiating under time pressure materially weakens outcomes. Most of the commercial value in Microsoft procurement comes from creating structured time to model alternatives and test assumptions before you're inside the renewal window.

Why this matters more than it used to

Microsoft's commercial model continues to evolve. Bundling is tighter. Platform dependencies are deeper. Pricing structures are more nuanced. And historic assumptions about volume discounting no longer hold reliably for cloud services.

Organisations that manage this well treat Microsoft licensing as a commercial negotiation, not an administrative exercise.

They model total cost. They benchmark terms. They create competitive tension where possible. They understand usage well enough to make informed trade-offs between immediate cost and future flexibility.

Those that don't often assume Microsoft spend is fixed and unavoidable.

That assumption compounds over time. The difference between good outcomes and poor outcomes in ICT procurement often comes down to whether you tested the assumption or accepted it.

This article provides general commercial and procurement commentary only and does not constitute legal, financial, or professional advice. Microsoft's licensing programs, pricing structures, and commercial terms change frequently. The author makes no representations or warranties as to the accuracy or currency of information presented. Readers should verify all licensing and pricing details directly with Microsoft or an authorised partner before making procurement decisions.