For most of the past decade, Business Central release waves gave finance and IT leaders a dependable planning rhythm. Twice a year, Microsoft published a release plan describing what was coming to Dynamics 365 Business Central (BC), and organizations used that document to scope regression testing and prepare their users. That rhythm has now ended. Microsoft announced in August 2026 that it is retiring the release wave 1 and release wave 2 model, publishing new capabilities to the AI at Work roadmap as soon as plans are committed, and retiring Release Planner by November 15, 2026.
The change is narrower than the headlines suggest. Microsoft has stated that it affects how upcoming innovation is communicated, while products with established release schedules keep following them, so BC still receives major updates in April and October with minor updates in between. What disappears is the fixed checkpoint: a single release plan that told organizations what to expect for the next six months and made it reasonable to treat update testing as a twice-yearly project.
For executives accountable for ERP risk, the practical question is who now sets the planning cadence. BC runs invoicing, payables, and period close for many mid-sized organizations, and Microsoft is adding AI agents to those workflows release after release. With 2027 the first full year of continuous innovation for BC, organizations that treat update validation as a standing control will be better placed than those waiting for an announcement that is no longer coming.
Key Takeaways for Executive Leadership
- The end of Business Central release waves changes how Microsoft discloses new capabilities, while the twice-yearly major update and monthly minor update schedule remains in place.
- Without a vendor-set planning checkpoint, each organization needs its own review and validation rhythm, tied to the update calendar rather than to announcements.
- Repeatable regression testing that produces dated evidence is the most practical way to keep update risk visible to finance and audit as change arrives more continuously.
What Changes When Release Plans Give Way to the AI at Work Roadmap
The AI at Work roadmap, formerly the Microsoft 365 roadmap, now carries upcoming capabilities for Dynamics 365, Power Platform, Dataverse, Microsoft 365, and Copilot in a single view. Each roadmap item moves through In Development, Rolling Out, and Launched, and its status is updated as plans evolve. Microsoft's transition notice sets out the specifics BC customers should plan around:
- New release plans are no longer published on Microsoft Learn as of September 2026, although existing plans remain available for historical reference.
- There was no release wave 2 announcement in September 2026, and future capabilities will not be grouped into wave announcements.
- Release Planner retires by November 15, 2026, and personalized saved views do not carry over to the new roadmap.
- New BC capabilities appear on the roadmap once they are committed, so most will be visible earlier than they would have been under the wave calendar.
The most consequential line in Microsoft's guidance is its recommendation that organizations set their own review rhythm, monthly or quarterly. In governance terms, that moves ownership of the planning cadence from the vendor to the customer, and any testing cycle that was triggered by the release plan now needs an internal trigger instead.
What Stays the Same in the Business Central Update Cycle
Although Business Central release waves no longer frame the roadmap, the delivery mechanics have not changed, and they remain the firmest anchor for any testing plan. According to Microsoft's Business Central update cycle documentation, major updates become generally available in April and October, and administrators can schedule them at any point within a five-month update period. A one-month grace period follows, after which an enforced period begins in which extensions that prevent the update from completing may be uninstalled automatically. Minor updates arrive in each of the remaining ten months and carry critical improvements, including regulatory updates.
Feature Management adds a second layer, since some capabilities arrive as optional and become mandatory at a later major update. In version 29, for example, disabling SOAP web services on Microsoft UI pages became mandatory, affecting any integration still using that route. Microsoft strongly recommends testing critical business scenarios before production moves to any new major or minor version, and that advice carries more weight once no single plan summarizes what each version contains.
The following table compares the planning model under release waves with the model that replaces it, and sets out what each difference means for testing.
| Dimension | Release Wave Model | Continuous Disclosure Model | Implication for Testing |
|---|---|---|---|
| Roadmap disclosure | Twice-yearly release plan published ahead of each wave | Capabilities published as soon as plans are committed | Test scope can no longer be set once per half-year |
| Planning checkpoint | Set by Microsoft each spring and autumn | Set by each organization | The review rhythm becomes an internal control |
| Feature visibility | Static document at the time of announcement | Status tracked through In Development, Rolling Out, and Launched | Test scope needs a periodic refresh |
| Delivery schedule | Major updates in April and October, minor updates monthly | Unchanged | Update windows still set the testing deadline |
| Optional features | Switchable until the next major update, listed in the wave plan | Switchable until the next major update, tracked item by item | Mandatory switches need validation before they arrive |
Why Business Central Regression Testing Has to Become Continuous
Most update-related failures in ERP environments come from existing processes behaving differently after a change nobody was watching for, such as an ISV extension that no longer interacts with a standard page as expected. Business Central regression testing exists to catch that category of risk, and its value depends on running at the same cadence as the change itself.
That cadence is becoming harder to match manually. The version 29 update alone changed areas from reporting to supply chain, including purchase order matching in the Payables Agent. When AI agents act on payables, a change to their behavior becomes a change to financial controls, and finance leadership needs evidence that those controls still hold.
A workable operating rhythm for BC under continuous innovation usually includes four elements:
- A monthly roadmap review using a saved AI at Work filter for Business Central, attended by both finance and IT process owners.
- A maintained regression suite covering the processes that feed the financial statements, including order to cash, procure to pay, and period close.
- Execution of that suite against a sandbox copy of production before each major and minor update reaches the production environment.
- Dated test evidence attached to the change record, so audit can see what was validated, on which version, and with what result.
What Independent Research Says About the Cost of Untested Change
Independent research supports treating update validation as part of financial control. Uptime Institute's Annual Outage Analysis 2026, published in May 2026, reports that 57 percent of respondents said their most recent major outage cost more than $100,000, and that for the second consecutive year one in five put the cost above $1 million. The same research identifies failure to follow established procedures as the leading driver of outages attributed to human error, with inconsistent or unclear processes also common.
Uptime surveys IT and data center operators rather than ERP users, but the pattern carries over to BC. When validation depends on individuals working through a manual checklist under time pressure, procedural drift becomes the likely point of failure, and the cost of a disrupted payment run or delayed close lands on finance.
The capacity to close that gap remains limited across the market. The World Quality Report 2025-26, produced by Capgemini, Sogeti, and OpenText from interviews with 2,000 senior IT executives in 23 countries, found that just one-third of test cases are automated on average, and that nearly half of organizations remain in the planning stage of their automation efforts. Those respondents work in organizations with more than 1,000 employees. Mid-sized BC customers, where testing often falls to finance staff alongside their regular work, are unlikely to be further ahead, which leaves most update validation dependent on manual effort just as change becomes harder to anticipate.
How Business Central Test Automation Supports Continuous Governance
For organizations moving to a continuous validation rhythm, the operational constraint is capacity, and elevaite365 Test Automation is designed around it. Business users capture the processes they already perform as reusable test steps without writing code, which keeps ownership of test scenarios with the people who know what a correct invoice, posting, or approval looks like. Tests adapt to interface changes instead of breaking when a control moves, so the regression library stays usable from one update to the next. Suites can run unattended against sandbox environments on a schedule aligned to the update calendar, and each run produces step-level evidence, including pass or fail status, timings, and screenshots, that can be attached directly to the change record.
The result is validation that keeps pace with Microsoft's delivery model, with results finance and audit can review without translation from IT. Teams weighing the effort can model the cost of their current manual cycles with the elevaite365 ROI calculator, and related guidance is available on the elevaite365 blog.
Governing Business Central Updates Without Release Waves
BC will update about as often in 2027 as it did in 2026. The difference after the end of Business Central release waves lies in who is responsible for knowing what each update contains and for confirming that the business still works afterward. Under continuous innovation, the planning checkpoint Microsoft used to supply has to be built and owned internally, supported by a maintained regression suite and evidence that stands up to audit.
Organizations that put that structure in place over the coming months will enter 2027 with update governance that matches how Microsoft now delivers change. Those that continue to wait for a twice-yearly announcement will find their testing scope defined by whatever happened to be noticed, which is a difficult position to defend when the processes in question feed the financial statements.
Frequently Asked Questions
Are Business Central release waves ending?
Yes, as a way of announcing new capabilities. Since September 2026, Microsoft no longer publishes release plans or wave announcements for Business Central, and new capabilities appear on the AI at Work roadmap as soon as plans are committed.
The underlying update schedule has not changed. BC continues to receive major updates each April and October, and Microsoft Learn still refers to version 29 as part of 2026 release wave 2, so wave terminology may appear in documentation for some time even though the wave announcements themselves have ended.
Does Business Central still get two major updates a year?
Yes. Major updates become generally available in April and October, and minor updates are released in each of the remaining months.
Administrators can schedule each major update at any point within a five-month update period, followed by a one-month grace period. After that, an enforced period begins in which extensions that block the update may be uninstalled, so testing should be completed well before the window closes.
Where can we find upcoming Business Central features now?
On the AI at Work roadmap, filtered to Dynamics 365 Business Central. Message Center remains the source for tenant-specific change notifications, and Microsoft Learn continues to publish What's New pages for each version.
Microsoft recommends saving a filtered view and subscribing through RSS, and roadmap views can also be exported to CSV. Because personalized views from Release Planner do not migrate, teams that relied on them should rebuild their filters on the new roadmap and assign someone to review it on a fixed schedule.
How often should Business Central regression testing run under continuous innovation?
At minimum, before every major and minor update reaches production. Organizations with heavy customization, several ISV extensions, or AI agents in finance workflows benefit from more frequent scheduled runs.
Microsoft recommends testing critical business scenarios before production environments move to any new major or minor version. Aligning a scheduled regression run to each update, with results recorded against the version tested, gives finance and audit a consistent record of what was validated and when.
What happens if an extension prevents a Business Central update?
The update fails, the environment returns to its previous version, and Microsoft reschedules the attempt for seven days later. Repeated failures become a governance issue once the update period ends.
During the enforced period that follows the grace period, extensions that cause the update to fail may be uninstalled automatically so the update can proceed. The extension's data is retained, but any business process that depends on it can stop working until a compatible version is installed, which is why extension compatibility belongs in the regression suite instead of being checked after the fact.
elevaite365 is a Microsoft-focused solutions provider specializing in test automation for Dynamics 365 applications. Learn more at elevaite365.com.
A note on statistics used in this article:
The Uptime Institute figures (57 percent of respondents reporting that their most recent major outage cost more than $100,000; one in five reporting costs above $1 million; failure to follow established procedures as the leading driver of human error-related outages) are sourced from Uptime Institute's Annual Outage Analysis 2026 announcement, published May 13, 2026. The cost figures draw on Uptime's 2025 Annual Survey. Uptime Institute is an independent research and standards body focused on digital infrastructure reliability.
The World Quality Report 2025-26 figures (one-third of test cases automated on average; nearly 50 percent of organizations still in the planning stage of automation) are sourced from the 17th edition of the report, based on 2,000 interviews with senior IT executives in 23 countries conducted by Coleman Parkes Research in June and July 2025 and published in November 2025. The report is co-published by Capgemini, Sogeti, and OpenText. Capgemini is an independent global consulting and research firm.
Product and schedule details (release plan retirement, Release Planner retirement on November 15, 2026, update periods, and Feature Management behavior) are sourced from Microsoft's AI at Work roadmap announcement, the Business Central update cycle documentation, and the version 29 What's New page on Microsoft Learn.
Aline Andersson
Head of Global Marketing
Based in Stockholm, Aline leads global marketing at elevaite365. She brings a strategic background in product marketing and communications across B2B technology, healthcare, and enterprise data platforms, with particular strength in positioning complex, AI-enhanced solutions for regulated industries.