After the last big upgrade
For decades, enterprise IT lived in a comfortable rhythm. You ran a version for five or seven years, then a migration project arrived, with a budget, a war room, and a party at the end. That rhythm is quietly dying. The major workplace and collaboration platforms now update themselves continuously, on the vendor's schedule, and the next big upgrade is never coming, because the upgrade is always already happening.
Change as an operating condition
Evergreen means change stops being an event and becomes weather. Features appear, features retire, defaults shift, and the environment you validated last quarter no longer exists in quite that form. Nobody asks whether you are ready. The question left to you is narrower and more interesting: how your organization absorbs a permanent stream of external change without drama.
What breaks first
Budgets break first. A migration was a project with a funding line and an end date; a change stream needs standing capacity, which finance processes built around projects are structurally bad at granting. Testing breaks next: the grand test phase before go-live has no go-live to precede anymore, so validation has to become a routine, sampled and risk-based, concentrated on the few paths that must never fail. Documentation and training follow: the screenshot manual is dead on arrival when the screens change monthly, so enablement becomes continuous too. And then the quiet one: ownership. A project owned the migration. Who owns the perpetual stream? If the honest answer is nobody, changes land unreviewed until one of them hurts, and then the answer becomes everybody, briefly, in a crisis call.
The operating model answer
The pieces are unglamorous, and they work. A named owner per evergreen platform, accountable for the change stream and not merely for the service. An intake routine in which reading release notes is a job, not a hobby, and every announced change gets classified: adopt, configure, defer, or block where blocking is even possible. A standing, largely automated check for the short list of things that must never break: identity and access, security baselines, the handful of core workflows the business stands on. And communication as a routine of its own, small and frequent rather than big and rare, because users forgive change they were told about and remember change they discovered.
The sober upside
Continuously updated platforms are continuously patched platforms, and the security argument in that sentence is not marketing. There is also a second-order prize. An organization that builds the absorption routine for one platform acquires a general capability: digesting external change without heroics. That capability outlasts any single product decision, and it transfers to whatever arrives next.
The last big upgrade has, for many landscapes, already happened; most organizations simply have not noticed yet. The ones that do well from here will not be the ones with the best migration projects. They will be the ones that no longer need any.