Sustaining Change After Launch: How to Keep New Ways Alive

Why changes fade after go-live and how to keep them working: define the core, build it into systems, reinforce it and read backsliding as information.

Leader Chaos · Last reviewed October 23, 2026 · 5 min read

Launch day feels like the finish line. The new system is live, the reorganization is announced, the training is done, and the project team is ready to move on. In practice, launch is closer to the middle. The months after it decide whether the change becomes "how we work here" or quietly fades into workarounds and old habits.

This guide covers why changes fade after launch, how to define what "sustained" means for your change, and the practical moves that keep a new way of working alive without freezing it in place.

Why changes fade

Much of the research on sustaining change comes from health care. In a 2012 review in Implementation Science, Shannon Wiltsey Stirman and colleagues examined 125 studies of what happens to new programs and practices after initial implementation. They found that partial sustainability, where some but not all elements of a program continue, was relatively common. Among the few studies that used rigorous evaluation methods, they write, "a small number reported full sustainment or high fidelity."

The review grouped influences on sustainability into organizational context, capacity, processes, and factors related to the new program or practice itself. Those settings differ from most workplaces, but the categories translate well: does the environment support the change, are there people and resources to keep it going, are there routines that keep it alive, and does the change itself fit the work?

Decide what "sustained" means

"Make it stick" is too vague to manage. Before the project team disbands, write down:

The core of the change: the few elements that must stay for it to deliver its purpose.
The adaptable parts: details teams can adjust to fit their work.
How you'll know it's in use, measured by behavior and results rather than availability. "Everyone has a login" is availability; "invoices are approved in the new system within two days" is use.
Who owns it after launch, by name.

Separating core from adaptable matters because some drift is healthy. In a 2013 paper in Implementation Science, David Chambers, Russell Glasgow and Kurt Stange proposed a Dynamic Sustainability Framework. They challenge two common assumptions: "voltage drop," the expectation that benefits shrink as a program moves from research settings into routine practice, and "program drift," the assumption that any departure from the original design reduces benefit. Their framework instead emphasizes continued learning and problem solving, ongoing adaptation with a focus on the fit between the change and its context, and an expectation of ongoing improvement rather than diminishing results.

Build the new way into systems

A change that depends on people remembering to do it will fade. A change built into the systems around people tends to last. The change firm Kotter describes the final step of its model, institute change, as connecting new behaviors to organizational success and evaluating systems and processes so management practices reinforce the new behaviors.

Achilles Armenakis and Stanley Harris, in a 2009 article in the Journal of Change Management, list strategies change agents can use through adoption and institutionalization, including formalization (structure and procedures), human resource management practices, and rites and ceremonies. In everyday terms:

Update job descriptions, onboarding materials and templates so new hires learn only the new way.
Reflect the change in goals and performance conversations.
Put it in the budget, so it doesn't depend on leftover project money.
Retire the old way. If the old spreadsheet still works, someone will keep using it.

Reinforce it, visibly and often

In Prosci's ADKAR Model, the fifth element is Reinforcement, which Prosci describes as sustaining the new way and preventing backsliding. Reinforcement can be as simple as a manager noticing and naming good use of the new process, sharing results that came from the change, or fixing a pain point people raised.

Leaders' own behavior is the loudest reinforcement. Armenakis and Harris describe the belief in principal support: that leaders are committed and that the change is not another passing fad or "program of the month." When leaders keep using the old reports or stop mentioning the change after launch, people draw the obvious conclusion.

Keep attention after the project team moves on

In the current version of Kotter's model, which the firm calls accelerators, the seventh step, sustain acceleration, urges leaders to press harder after the first successes. The risk after launch is not open opposition; it's that attention moves to the next priority.

For example, imagine a hypothetical 60-person property management company that moved maintenance requests from phone calls to an online system. The project team declares success at go-live. Three months later, half the requests arrive by phone again, because a few busy staff found it faster to call a favorite technician. A simple post-launch rhythm would have caught it: a named owner who reviews request data monthly, a 30-day check-in with each team, and a 90-day review that asks what is getting in the way.

Read backsliding as information

When people drift back to the old way, it's tempting to treat it as discipline. Sometimes it is. More often it's a signal about fit. Watch for:

Shadow tools, such as personal spreadsheets or side channels that duplicate the new system.
Workarounds that skip a step of the new process.
New hires being taught the old way by colleagues.
Results stalling after an early gain.

Ask what the workaround is solving. If the new process genuinely doesn't fit a situation, adapting it may be better than enforcing it. That is the Dynamic Sustainability Framework's point: keep the core, and keep improving the fit.

Know when to adapt or retire

Not every change should last forever. Wiltsey Stirman and colleagues note that discontinuing a program is sometimes the result of finding more effective, efficient or compatible practices. Researchers Stephen Cummings, Todd Bridgman and Kenneth Brown note that the familiar "refreeze" image fits awkwardly with Kurt Lewin's own view that groups are always adapting. Sustaining a change means keeping what works working, not locking it in place.

Key takeaways

Launch is the middle of a change, not the end; research on new programs finds partial sustainment is common.
Define the core of the change, the adaptable parts, how you'll measure real use and who owns it after launch.
Build the change into systems: job descriptions, onboarding, goals, budgets, and retiring the old way.
Reinforce it through recognition, results and leaders' own consistent behavior.
Treat backsliding as information about fit, and adapt where that serves the change's purpose.

Sources

Spotted something out of date? Let us know.