A pilot closes with good results, and nobody has yet decided who will notice when the new system starts to drift.
The point where most transformations end
The pilot in a global IT organization closed with positive final readings across the measures it had been designed to move. The sponsor called a close-out session to end the pilot and open the broader rollout. Before the rollout conversation began, she raised a different question: what happens now to the team that built it?
The internal change team had been operating as a project team for most of a year, and the work in front of it was no longer a project. When the CIO asked what ongoing sustainment actually meant, we answered with what usually happens instead. Most transformations end at exactly this point. The consultants leave. The internal team disbands. The new system rolls out, and then, slowly, the patterns that produced the original problems reassert themselves. Intake drifts. The prioritization forum loses discipline. The reporting layer accumulates exceptions. Within eighteen months, the organization is living with a new generation of pain points, often the same shape as the ones the work was meant to solve, because nobody was responsible for keeping the system alive.
From project team to standing function
The team's next form, we told them, was not a project team. It was a sustainment function with five jobs: watch the measures continuously, surface new gaps as they emerge, govern the system against drift, evolve the design as the organization changes, and teach the method to the people who would carry it next.
One member asked whether that meant they simply kept meeting. They did, but the cadence, composition, and purpose all changed. Weekly working sessions are project rhythm. What they needed was operating rhythm: a monthly review of system health against the measures, a quarterly session to surface new gaps, and an annual cycle to evaluate and redesign the system itself. Some members would stay for continuity, new people would join for fresh perspective, and the function would be built to outlast any individual in it, including the sponsor.
That last point landed harder than we expected. One member said what the others were thinking: they had built transformations before that depended on one person, and when that person left, the transformation collapsed within a year.
- Weekly working sessions
- Builds the new system
- Fixed membership
- Depends on its sponsor
- Monthly and quarterly reviews
- Watches, governs, evolves, teaches
- Rotating membership
- Built to outlast any one person
Two measures that slipped
The first health review took place six weeks after the pilot closed. We facilitated it so the sponsor could see the cadence modeled before running it herself. Two new members had joined: a delivery lead who had been one of the pilot's strongest advocates, and a senior business representative who had asked to join because her team had benefited from the work.
Most measures were holding or improving. Two were not. The share of work going through the new prioritization forum had begun to slip. The forum was still in use, but its use was eroding. And lead time for enhancements and change requests, which had fallen steadily through the pilot, had gone flat for the first time, with work from the largest business unit actually rising.
The sponsor called them what they were: not crises, but signals. We walked the team through the questions to ask. Was the drop concentrated in one category of work? Were leaders relitigating decisions outside the forum, the way they had before? Was the largest unit still submitting complete requests?
The new delivery lead supplied part of the answer. That business unit had been bypassing intake for weeks, sending requests through chat and email instead of the front door. The lead had assumed it was a one-time lapse. Hearing the lead-time data, it no longer looked like one. No single measure would have caught that pattern. It became visible when the numbers sat next to what someone in the room was already seeing. The sponsor gave the team two weeks to find out what was happening and come back with a recommendation for each signal.
-
Read the measures Monthly health review
-
The signal Spot the signal Forum use eroding; lead time flat, one unit rising
-
Pair it with what people see Largest unit bypassing intake by chat and email
-
Investigate for two weeks Find the underlying condition
-
What changed Fix small or escalate Targeted closure, or a larger finding upward
Stepping back to one session in eight
Six months into the sustainment function, the rhythm held. The quarterly sessions had produced three targeted interventions: re-anchoring the prioritization forum after its use eroded further, redesigning intake routing for the largest business unit, and refining the reporting layer to capture signals the original measures had missed. Each closed a specific gap before it could compound.
The sponsor was running the sessions now. We attended about one in eight, mostly as observers. The function had also started teaching. Two new areas of the organization launched pilots of their own, each with its own internal team, and the sustainment function served as their methodology coach, with members taking on mentoring roles.
In the seventh month, the team surfaced something it could not resolve at its own level: two business units being measured against partially overlapping outcomes, competing for the same delivery capacity without a shared basis for trade-offs. The intake bypass, the forum erosion, and the lead-time variance were all symptoms of it. When the CIO asked for the underlying condition, the sponsor answered without hesitation. The CIO concluded that the conversation belonged at the CEO level. What happened there is a different story. What mattered here was a team that could take a finding bigger than itself to the people who owned it.
-
Pilot close
A standing function forms
-
Week 6
First health review Facilitated by Agileon as a model; two signals found
-
Month 6
What changed Sponsor runs the sessions Agileon attends about one in eight; two new pilots coached internally
-
Month 7
A finding goes upward Overlapping business-unit outcomes taken to leadership
-
About a year in
Sponsor teaches the method Agileon in the room as guests

The work had become hers
About a year after the engagement began, the sponsor asked us to attend one final session. The sustainment function now included representatives from several pilot areas and a group of newer leaders she was bringing into the method for the first time. She walked them through the method at the pace of someone who had taught it many times. We were in the room as guests.
Asked what mattered most, her answer was about discipline. The method is not hard to understand, she said; it is hard to hold to, because the pressure to skip steps, fix symptoms, and declare victory early never goes away. And none of it was finished. The system they had built would not be the system they had in three years.
We left after her close. There were no more questions for us to answer. The work was no longer ours. It had become hers, and through her, the organization's.
A system is sustained by the people who notice it drifting, not by the people who installed it.
Share With Your Team
Choose who it is for. You send it yourself from your own email; Agileon never writes to your colleagues.


