Why application rationalization should be continuous
The assessment is accurate the day it lands, and decaying by the next quarter.
A conventional application rationalization is a point-in-time assessment. It inventories the estate over several months, produces a disposition for each system, and is delivered as a document. From that moment it begins to decay, because the estate it describes keeps changing.
What is continuous application rationalization?
Continuous application rationalization keeps portfolio disposition decisions connected to live architecture, capability and value data, rather than fixing them in a document at a single moment. The portfolio view reflects what the organization runs now, and when something changes the affected decisions can be re-interrogated rather than re-derived.
How fast does a rationalization decay?
Faster than most sponsors expect, because decay is driven by ordinary business activity rather than by anything unusual. In the two quarters after delivery a typical enterprise will have: completed at least one acquisition or divestment; had a division buy a platform without central involvement; slipped a migration that the roadmap assumed complete; and seen a vendor acquired, changing the support outlook for a system marked retain.
None of those are failures. They are what organizations do. But each one invalidates part of the analysis, and because the analysis is a document there is no mechanism to identify which part.
The cost of the point-in-time model
- Re-commissioning. The most visible cost. Organizations pay for the same assessment repeatedly, usually when leadership changes and the incumbent analysis is no longer trusted.
- Decisions deferred past the window. Rationalization roadmaps are sequenced over years, but the analysis behind them was valid for months. Later-phase decisions execute against stale reasoning.
- Loss of reasoning. The document records what was decided. It rarely records the assumptions well enough for a successor to tell whether a decision still holds.
- Organizational scepticism. After the second or third assessment, teams reasonably conclude that rationalization is a recurring exercise rather than a decision, and engage accordingly.
What changes when the reasoning is retained
The shift is from storing conclusions to storing the reasoning that produced them. If a disposition records that a system was retained because it was the source of truth for a data domain, supported by evidence and reviewed by a named owner, then when a replacement assumes that data domain the affected decision is identifiable. The question becomes which of our assumptions just changed, and what depended on it.
That is a different artifact from a disposition matrix. It requires evidence links, editable assumptions, and change propagation — which is why it tends not to survive in a spreadsheet.
What it does not solve
Continuous rationalization does not make the decisions easier. The binding constraint in most programmes is not analysis quality but organizational willingness to retire systems people depend on. Keeping the analysis current removes one excuse for delay; it does not remove the politics, and any account suggesting otherwise is selling something.
Where EXOS sits on this
EXOS retains the reasoning behind each disposition in a governed environment the client keeps, rather than delivering a matrix and departing. See application rationalization.