Author: Chandresh Lad is a Product Manager at Anaplan.
Every mature Anaplan environment has models that eventually reach the end of their active life. The challenge is knowing what should happen next.
As Anaplan environments mature, the number of models naturally grows. New planning processes are introduced, existing solutions evolve, development and test models are created, and projects come and go. Over time, some of those models stop being actively used - a new solution replaces an older one, a project finishes, a proof of concept has served its purpose.
But the model remains, sitting unnoticed in a workspace or archive, consuming capacity and adding to an increasingly complex landscape. This is the model graveyard.
The problem isn't that models reach the end of their useful life - that's natural. The problem is letting them remain indefinitely without a conscious decision about what happens next.
Models have a lifecycle
An Anaplan model shouldn't be viewed as something built once and left to exist forever. Like the business processes they support, models have a lifecycle: created to address a need, developed and tested, deployed, maintained and enhanced, and eventually replaced or retired.
That final stage matters as much as the first — organizations are usually disciplined and particular about how a model is built and deployed, but few apply the same discipline to how it gets retired.
A mature environment has a clear approach to what happens when a model reaches the end of its active life: become read-only, be archived, or be deleted? Without an intentional lifecycle, the default is to keep everything, and that's how the model graveyard grows.
The "We might need it one day" problem
When someone asks whether an old model can be removed, the answer is often:
"We might need it one day."
It's understandable — the model may hold historical information, or something may still depend on it. But it isn't a sustainable retention strategy. Every model creates an ongoing responsibility: it needs to be understood, secured, governed and reviewed again. The longer models are kept without a defined reason, the harder it becomes to determine valuable assets from ones that simply haven't been deleted.
Not every inactive model is obsolete
There's an important distinction between inactive and obsolete. Some models are only used at particular points in the year; others hold historical information worth retaining. A model that hasn't been accessed recently isn't necessarily one to delete.
The more important question is:
Does this model still have a legitimate business, historical, audit or regulatory purpose?
If the model serves a distinct purpose, it should be managed accordingly. If not, it should be flagged for decommissioning and deletion. Deliberate model management requires clear retention criteria supported by concrete operational signals.
The archive is not the end
Archiving offers a useful transition between active use and deletion, reducing the risk of users working in an outdated solution while preserving history for an agreed period. But it shouldn't become a permanent destination, an archive holding every model ever created eventually becomes its own management problem.
Treat archive space as a limited resource. Before archiving a model, ask why you're keeping it, what value it provides, how long you need it, and what will eventually allow you to delete it. The objective isn't to preserve everything — it's to preserve what's valuable.
Retirement should be deliberate
Retiring a model shouldn't be an informal decision made when someone notices an outdated model, it should be part of the lifecycle. A typical process moves a model from active use through read-only access, archiving where appropriate, and ultimately deletion once retention requirements are met. The process will vary between organizations, but the principle holds: every model should have a deliberate end.
Deletion isn't a failure. If a model has served its purpose, its successful retirement is evidence the organization is managing its estate effectively.
The cost of keeping everything
Retaining models costs more than workspace capacity. Each additional model adds complexity — another asset to understand, another review, another question for future teams. A large collection of unclear or unused models makes it harder to focus on what actively supports the business.
Good model management isn't about maximizing how many models you keep. It's about maintaining an environment that's both understandable and sustainable.
Building a healthy model estate
A mature environment shouldn't be judged by how many models it contains or maintains, but by whether the organization understands its estate and makes deliberate decisions about what belongs in it.
Some models will stay in production for years. Some will be retired and retained for historical purposes. Some will sit in the archive for a defined period. Some will be deleted. All of these can be signs of effective management. The problem isn't models reaching the end of their lifecycle, it's failing to recognize when they have.
Final thoughts
The model graveyard isn't necessarily a collection of bad models. It's what happens when models outlive the need that created them and no deliberate decision is made about their future.
The solution isn't deleting everything old, it's managing the lifecycle with purpose. Keep what provides value. Retain what has a legitimate purpose. Archive what needs preserving. Delete what no longer needs to exist. And don't let uncertainty become a reason for retention.
A healthy Anaplan environment isn't one where every model survives forever. It's one where every model has a defined lifecycle.
……………
More from Chandresh: The model lifecycle: Managing Anaplan workspaces for long-term success
