Sustainment Architecture Check: The Architecture That Holds After the Engagement Ends

Two leaders study a diagnostic clarity layer held by sustainment architecture, with one marked weak joint where change can decay.

Twelve months from now, what survives the diagnostic?

Today’s Executive Briefing: The Architecture That Survives After the Engagement Ends

A diagnostic can be right and still fail to sustain.

The finding can be accurate. The plan can be clear. The first few weeks can even look better.

Then the organization returns to the baseline condition.

The meeting cadence thins out. The owner starts carrying the decision again. The new process becomes optional. The team remembers the language but stops using the standard. The old workaround quietly resumes its place inside the system.

That pattern has a name in the organizational change literature.

David Buchanan and colleagues reviewed the research on sustaining organizational change and named the problem directly. Some improvements decay. Gains disappear when new practices are abandoned. The NHS Modernisation Agency called this the "improvement evaporation effect." The review also makes the structural point that matters here. Sustainability depends on multiple factors at once, including managerial, financial, leadership, organizational, cultural, processual, contextual, and temporal conditions. (Buchanan et al., 2005)

In plain terms, sustainment is architecture.

McKinsey's 2023 transformation research shows the same problem at enterprise altitude. Fifty-six percent of respondents said their organizations achieved most or all transformation goals at first. Only 12% sustained those gains for more than three years. Respondents also reported that 42% of potential financial benefit was lost during execution and sustainment. The top performers had a different pattern. They maintained rigor in later phases, achieved people goals, and kept resources committed throughout. Organizations doing all three were 3.4 times more likely to sustain gains for more than three years. (McKinsey, 2023)

That is the architecture underneath the result.

The rehabilitation parallel is useful because it separates the original intervention from the system that preserves it.

A surgeon can repair the knee. The repair still needs rehabilitation. Without a post-surgical protocol, the body can stiffen, weaken, protect, compensate, or form scar tissue that limits motion. Clinical literature treats poor or missing physical therapy as its own risk factor, separate from whether the surgery itself was technically successful. Disuse research also shows how quickly the body begins to lose capacity under immobilization, with muscle loss measurable in days. (Journal of Applied Physiology, 2016; Noyes Knee Institute, 2022)

The surgery was real.

The rehabilitation architecture was missing.

Organizations work the same way.

The diagnostic names the condition. The sustainment architecture determines whether the organization can hold the correction after the diagnostic leaves the room.

Typically, in engagements, I find and diagnose the absence of sustainment architecture as an Inadequate Architecture condition. The organization may have real systems. Those systems may have worked for an earlier stage. They may even support ordinary execution. But they are inadequate for preserving the new decision rights, cadences, standards, ownership, and consequences the diagnostic introduces.

That is why the same pattern keeps frustrating capable operators.

The operator who built the organization's operating architecture over decades and now sees it reverting after every engagement built for a prior condition. The current condition requires a sustaining system that may never have been installed.

Use this before the next engagement, reset, operating-plan cycle, or post-diagnostic follow-through window.

Tool: Sustainment Architecture Audit

10-15 minutes. One page.

Choose three systems that matter right now.

A system can be a decision category, meeting cadence, ownership handoff, performance standard, client-delivery process, approval path, leadership behavior, operating metric, or strategic initiative.

For each system, write three lines.

  1. What is the system called?
  2. What sustains it today? A person, a process, or architecture.
  3. What happens if the sustaining mechanism disappears for 30 days?

Then mark the system.

Person-held means the system survives because someone compensates for the missing architecture.

Process-held means the system survives as long as people remember, comply, and make time for it.

Architecture-held means the system has decision rights, cadence, visibility, ownership, consequences, and review built into the way work runs.

Now look for the risk.

If the answer is a person, the system is vulnerable to absence, fatigue, delegation, succession, resentment, or scale.

If the answer is a process, the system is vulnerable to urgency, drift, exception making, and leadership inconsistency.

If the answer is architecture, ask whether that architecture still matches the condition the organization is in now.

The clearer sustainment question is whether the organization has somewhere for the diagnostic to live after the first improvement cycle ends.

© 2026 Lauren Carter. This instrument is proprietary. For individual diagnostic use only. Reproduction, adaptation, or redistribution in any form requires prior written permission from Lauren Carter.

The Rapid Executive Diagnostic reads the condition. The sustainment architecture check reads whether what the diagnostic produces will hold. Both reads happen within ninety minutes.

If this names the condition underneath what looks like the problem, there is more here. Subscribe for the next Briefing.

— Lauren

Clicking tells us you meant to answer. Hitting submit sends us the actual response. Without submit, we can’t see your choice.


If this is the conversation someone in your network has been needing, forward it. They'll know.



Do not click

Lauren Carter

A twice-weekly diagnostic on the structural conditions underneath how organizations actually perform. Each issue names a mechanism most strategy conversations skip, then gives you a tool to test it in your own operation. Built for executives, founders, and operators who already suspect the problem is architectural.