Why business improvements do not last and how to stop slipping back.

The project went well. The new process launched. People attended training, the presentation was circulated and somebody described the change as embedded.

Three months later, the old spreadsheet has quietly returned.

When an improvement fades, it does not necessarily mean the original idea was wrong or the team resisted change. It may mean the business never established the conditions needed to maintain it. This article explains what happens between project completion and dependable everyday performance.

1. Why launch day is the beginning of operational work

Buying exercise equipment is an event. Using it consistently is an operating routine.

A new process can encounter a similar gap between ownership and actual use, although it usually makes a less convenient clothes horse.

Implementation introduces a method. Sustained improvement requires that method to remain workable under normal demand, absence, exceptions and competing priorities.

Define what happens after the project closes. Who maintains instructions, reviews results, resolves defects and approves adjustments? Include those responsibilities in delivery planning rather than treating them as an administrative handover at the end.

A launch establishes availability. Performance over time establishes whether the change works.

2. Distinguish a temporary push from a sustainable improvement

A queue may fall while the project team works overtime and an experienced manager checks every difficult case.

That is an achievement. It is not yet proof that the normal operation can sustain the result.

Separate the effect of the changed method from the extra resources supporting implementation. Examine workload, specialist involvement, exceptional effort and deferred tasks.

A sustainable process needs sufficient ordinary capacity and clear responsibilities. If its performance depends on people repeatedly stretching beyond the planned arrangement, the design may still be incomplete.

I’d review what happens when temporary support ends. This avoids mistaking a successful mobilisation for a permanent increase in operational capability.

3. Assign ownership after the project team moves on

“The business owns it now” sounds reassuring until you ask which part of the business.

A process can spend a surprisingly long time being owned collectively and maintained by nobody.

Name the person accountable for keeping the method effective and clarify the responsibilities of managers, users and supporting functions. Include authority to resolve issues and access to the information needed.

The Institute for Healthcare Improvement’s work on sustaining improvement emphasises the daily work of frontline managers within a wider management system. That principle transfers usefully to business operations, although the practical routines must fit the setting.

Ownership needs to be executable. Giving someone a responsibility without capacity, authority or support can reproduce the original gap.

4. Check how the process is actually being used

Attendance at training does not demonstrate that people can use the process on an awkward Tuesday afternoon.

A completed learning register is evidence of attendance, which is useful information but has limited problem-solving powers.

Observe real work and sample the decisions, records and outcomes. Ask where staff have adapted the method and why.

A workaround may reveal an unresolved defect, a missing exception or instructions that no longer fit reality. Investigate before treating every deviation as a discipline issue.

Also check whether old tools and routes remain available. Running both methods indefinitely can split information and make performance difficult to interpret.

A clear working standard provides a reference for improvement, while feedback establishes where that standard needs revision.

5. Detect drift before the result deteriorates

A door that sticks slightly is easier to investigate before someone needs a shoulder to open it.

Operational drift often begins with small adaptations that appear reasonable in isolation.

Look for early indicators such as repeated exceptions, missing information, ageing work or increased dependence on a particular person. Review these alongside outcomes rather than waiting for the final target to fail.

Agree what pattern should trigger investigation and who responds. The control should support timely judgement, not create a report that nobody reads.

The aim is to recognise changes in operating conditions before the business becomes dependent on a new set of informal repairs.

6. Keep the management routine proportionate

Sustaining improvement does not require a weekly meeting about whether the weekly meeting is working.

A process can be over-governed as well as neglected.

Use the minimum useful rhythm for the work: a short operational check, periodic evidence review or a deeper discussion when conditions change. Match the frequency to consequences, variability and the speed at which problems develop.

Separate routine monitoring from investigation. Stable performance may need little intervention; an emerging issue may need focused analysis rather than another standing committee.

Keep actions clear and revisit them. Oversight that generates commitments without checking completion can become additional work while providing little control.

7. Adapt the process when circumstances change

A method designed for twenty requests a week may struggle with two hundred.

Telling people to follow the process more faithfully will not necessarily expand the doorway.

Demand, customer expectations, staff experience and systems can change after implementation. Review whether the assumptions supporting the design still hold.

Use a learning cycle to test adjustments and assess their effects. Preserve the purpose and essential controls while adapting the method to current conditions.

This is the difference between maintaining a document and maintaining operational capability. A process can be followed exactly and still become unsuitable for the business it now serves.

Our article on proportionate operational improvement explains how to test changes without automatically beginning a wider redesign.

8. Use independent challenge to maintain direction

A regular outside perspective can help when managers are too close to the daily work to see emerging patterns.

It can also provide somebody to ask the awkward but useful question before the quarterly result asks it more forcefully.

DeaconOps’s Operational Reviews offer ongoing senior operational support. That can include reviewing performance, discussing priorities, identifying problems, joining selected meetings and acting as a sounding board.

The service is broader than checking a completed project. It can support direction in an operation that needs experienced input at intervals.

Where a new issue needs substantial investigation or implementation, scope that separately through a Diagnostic or Improvement Project. Start a conversation about the level of support the business needs.

Questions about sustaining improvement

Does a return to the old process mean people resisted change?

Not necessarily. The new method may have unresolved defects, insufficient capacity or unclear ownership. Examine why people reverted before choosing a response.

How long should we monitor a new process?

Long enough to understand performance under relevant operating conditions, followed by an appropriate ongoing review. There is no single duration that suits every change.

Can Operational Reviews help even when we have not completed a project?

Yes. They can provide ongoing direction, problem identification, expertise and a sounding board, with the scope and time commitment agreed around the business.

---

About the author

Dean Cain is the founder of DeaconOps, with more than 15 years’ experience across operational delivery, leadership and regulated environments. Read about the DeaconOps approach.