On March 9th, Microsoft made an important announcement. It introduced Microsoft 365 E7, which it named "The Frontier Suite." It is not a new product: it is a bundle within the Microsoft 365 family where four existing components have been gathered under a new name and a single price.
However, the name was chosen carefully: if your company is not on E7, the implicit message is that it is not at the frontier. It has been left behind.
Before making any decision, it is worth understanding what is actually inside this offer and which elements require careful evaluation.
E7 will be available from May 1st, 2026, at $99 per user per month, with or without Teams (thus respecting the EU unbundling requirements).
Inside, there are four components that already existed and could be purchased separately: Microsoft 365 E5 at $60, Microsoft 365 Copilot at $30, Entra Suite at $12, and the new Agent 365 at $15. If you sum them up, you reach $117. The bundle offers them at $99, with a declared saving of 15%.
The math works, but some licensing experts like SAM Expert have already highlighted a contradiction in the calculation. Entra Suite is included at $12, which is the additional price relative to Entra ID P1, already present in M365 E3. The problem is that E7 includes M365 E5, not E3, and M365 E5 already includes Entra ID P2, which covers some of the functionalities of Entra Suite.
Therefore, the real incremental value for those starting from E5 is less than $12, and the actual savings of E7 drop from the declared 15% to about 12-13%. It is not a huge difference in absolute value, but it is an element to keep in mind—especially on large-scale deployments.
In any case, the comparison with the à la carte price list is only the starting point. The comparison that really matters is with what is being paid today: if you are on E5 at $57 with Copilot at $30, you spend $87 per user. Moving to E7 means $12 more per user per month to add Entra Suite and Agent 365. The concrete question is whether these two components will generate real value for your organization.
There is also a contextual element that cannot be ignored. E5 will rise to $60 starting July 2026. E3 is already at $39. EA volume discounts were eliminated in November 2025 (we discussed this here as well). Many organizations are already absorbing effective increases of 15-23% before even considering E7. The next renewal is structurally more expensive than the previous one—E7 must be evaluated on this basis, not on the comparison with the à la carte price list.
To understand why Microsoft launched E7 right now, it is useful to look at what has happened with Copilot over the last two years.
Copilot was launched in 2023 at $30 per user per month as an add-on (here is our article on the subject). After two years of progressive discounts applied to partners and enterprise customers, lowering of minimum thresholds, and monthly billing to reduce adoption barriers, results show that Microsoft counts 15 million paid seats out of an installed base of 450 million M365 commercial users—approximately 3.3% of the installed base. This figure was disclosed for the first time in the January 2026 earnings call, after eight quarters without communicating the number, and analyzed by the industry newsletter Perspectives Plus.
This is not an isolated case in Microsoft's history. When a product does not reach expected adoption levels as an add-on, the recurring response is bundling. This already happened with Teams Phone and Power BI Pro, both absorbed into E5. E7 follows the same logic for Copilot and Agent 365.
The practical result is the possibility that with E7, Copilot could become a default, not a choice. When a company signs an EA standardizing on the E7 plan (perhaps to avoid being "demoted" to the MCA contract), the cost will fall on the entire user base, regardless of who actually uses Copilot or Agent 365 every day.
In a scenario where market protections are being discussed—see the Teams unbundling case—perhaps it is time to start providing more awareness and protection tools even to those signing the contract: progressive bundling could reduce the ability of organizations to understand what they are buying and why.
This is probably the most underestimated aspect of the entire offer.
$99 per user per month is not the complete cost for an organization that wants to use AI agents in a structured way.
Agent 365—one of the main components of E7—is a control plane. It serves to observe, govern, and monitor agents. It must be integrated with Copilot Studio—the tool used to design and publish agents which has separate consumption costs based on Copilot Credits. For more complex agents, requiring custom models or advanced orchestration, Microsoft Foundry is added—the underlying Azure platform—with even more variable consumption costs. These costs are not included in the E7 price and end up on the Azure invoice, managed by different teams, under different cost centers.
And in the absence of adequate FinOps practice, there is also a risk that those consumption costs may not be correctly entered, forecasted, or optimized relative to the existing Azure commitment.
The critical point is that as of March 9th, 2026, Microsoft has not published reference architectures or guidelines on the total cost of ownership for enterprise-scale agentic deployments.
Copilot Studio has a public price list, but real costs depend on how agents are designed, which models they use, and how frequently they interact with external systems—variables that no Microsoft document yet helps to estimate in a structured way. Those who sign E7 today commit to a cost model whose variable part is still difficult to quantify. It is an element that the CFO and the procurement team should keep in mind during budget planning.
A second aspect to know concerns the actual status of the product at launch.
As of May 1st, 2026, several security features of Agent 365 will remain in public preview. Runtime threat protection—one of the main use cases for those who want to govern agents in production environments—will enter preview in April 2026, so it will not be generally available at the official launch.
Anyone purchasing Agent 365 on day one should verify which features are actually in general availability and which are still in the preview phase before building deployment plans upon them.
There is a structural issue looking toward the coming years that deserves reflection.
Microsoft has publicly indicated that AI agents will need to be licensed similarly to employees—with their own identity, access to data, and compliance. The consequent logic is that an organization with 5,000 employees that deploys 2,000 agents could find itself managing licenses for 7,000 "workers."
The $15 per user for Agent 365 is the current starting point. On a three-year contract, one commits to a pricing model that Microsoft is still completing. It is an element of risk that must be quantified, especially for organizations that have significant automation plans in the coming years.
The "Frontier" brand is built to create a sense of urgency—a classic change management technique theorized by John Kotter: to make people accept a change, you must first convince them that not acting is dangerous. In a moment where boards of directors are asking IT management to demonstrate progress on AI, those who do not adopt E7 will find themselves having to explain the choice. The burden of proof tends to shift: it is no longer just the vendor who has to demonstrate the value of the product.
This context makes the next cycle of EA renewals particularly delicate. Arriving prepared—with real usage data, an honest assessment of necessary components, and an estimate of consumption costs—is the difference between a conscious commercial decision and one made under pressure.
A second aspect to know concerns the actual status of the product at launch.
As of May 1st, 2026, several security features of Agent 365 will remain in public preview. Runtime threat protection—one of the main use cases for those who want to govern agents in production environments—will enter preview in April 2026, so it will not be generally available at the official launch.
Anyone purchasing Agent 365 on day one should verify which features are actually in general availability and which are still in the preview phase before building deployment plans upon them.
There is a structural issue looking toward the coming years that deserves reflection.
Microsoft has publicly indicated that AI agents will need to be licensed similarly to employees—with their own identity, access to data, and compliance. The consequent logic is that an organization with 5,000 employees that deploys 2,000 agents could find itself managing licenses for 7,000 "workers."
The $15 per user for Agent 365 is the current starting point. On a three-year contract, one commits to a pricing model that Microsoft is still completing. It is an element of risk that must be quantified, especially for organizations that have significant automation plans in the coming years.
The "Frontier" brand is built to create a sense of urgency—a classic change management technique theorized by John Kotter: to make people accept a change, you must first convince them that not acting is dangerous. In a moment where boards of directors are asking IT management to demonstrate progress on AI, those who do not adopt E7 will find themselves having to explain the choice. The burden of proof tends to shift: it is no longer just the vendor who has to demonstrate the value of the product.
This context makes the next cycle of EA renewals particularly delicate. Arriving prepared—with real usage data, an honest assessment of necessary components, and an estimate of consumption costs—is the difference between a conscious commercial decision and one made under pressure.
At WEGG, we are SAM consultants specialized in optimizing license costs and negotiating with software vendors. We support organizations in structuring license management processes that ensure both compliance and a concrete lever for structural savings.
We do this through Flexera technology, which collects installed data on endpoints and compares it with what has been purchased—and which, through a SaaS connector to Microsoft APIs, brings usage data from the Microsoft 365 Admin Center and Viva Insights into the platform. The result is a single view per user: assigned license, real usage, and work patterns all in one place, without manually crossing exports from different systems.
Based on this, we work on three fronts. We help to:
Insights
OUR OFFICES
OUR OFFICES
PADUA
Via Arnaldo Fusinato 42, 35137
MILAN
Viale Enrico Forlanini 23, 20134
ROME
Viale Giorgio Ribotta 11, 00144
Copyright © 2025 WEGG S.r.l. • P.I 03447430285 • C.F. 02371140233 • REA 311023
Azienda Certificata ISO 9001:2015 – ITA / ISO 9001:2015 – EN
