Lean Change: turn uncertainty into a small, reviewable experiment

Lean Change is an approach to organisational change that connects insights, options and experiments. Instead of treating an implementation plan as a prediction of how people will respond, the team uses a limited intervention to learn what helps, what causes difficulty and what to do next. The approach’s owner places co-creation and feedback at its centre. Lean Change’s explanation

The useful unit of work is a question with an accountable next step. “Will a shared case card improve the afternoon handoff?” can become a test. “Make the warehouse more agile” cannot, until someone identifies the work, the people affected and the evidence that would matter.

Small experiments still need planning. They need permission to alter the work, a boundary on who is affected, support for participants and an honest review. The following process is an original practical adaptation of the approach, rather than an official Lean Change assessment or a claim that experimentation is always the right choice.

What the cycle adds to a change plan

Lean Change is associated with Jason Little. The organisation’s own history dates his book’s formal publication to 2014, following serialised chapters in 2013. That history establishes the approach’s provenance; it does not establish comparative effectiveness. The owner’s history

A conventional project plan can still describe dependencies, spending limits and contractual deadlines. The cycle adds a way to revisit assumptions about implementation. An insight changes the list of options; an experiment produces further insight. Completing an activity is therefore different from finding evidence that the change is workable.

Three questions keep the cycle concrete:

• Insights: What have we seen, heard or checked, and what are we still assuming?
• Options: What could we change, including options that address the environment?
• Experiments: What limited attempt could help us choose our next action?

The distinction matters when a team mistakes its preferred intervention for a settled need. A workshop might reveal that colleagues understand the new procedure perfectly but cannot access a required record. More communication would address the wrong problem. Begin with a change-readiness inquiry when feasibility and commitment remain unclear.

Find an insight before prescribing an intervention

Consider a hypothetical warehouse that transfers unfinished orders between morning and afternoon shifts. The operations manager thinks the afternoon team needs a reminder to check the shared spreadsheet. A packing colleague says the spreadsheet does not identify which trolley contains the unfinished order. Both accounts may describe part of the problem; neither explanation is established by the conversation alone.

Watch a normal handoff, with the workers’ agreement. Record what information is available, what question the recipient asks and where the answer comes from. Separate observation from interpretation: “the recipient asked which carton was waiting” is an observation; “the recipient does not care about accuracy” is an inference that the observation does not support.

Interview both sides of the handoff. Ask what makes a case difficult, what they already do to recover missing information and what else changes during a busy period. An existing workaround may be a clue to a missing design condition. It may also create a risk, so check it rather than immediately making it the standard.

In the invented warehouse, workers identify two problems: the order reference is sometimes absent from the trolley, and nobody is sure who can release a carton when one item is missing. One is an information problem; the other concerns decision authority. Treat them as separate insights with separate evidence.

Use stakeholder mapping to include the dispatch colleague who receives the carton next. Otherwise, a local improvement might simply move the confusion downstream. Keep customer details out of a public experiment board and agree how observation notes will be used.

Give participation a real decision boundary

Inviting colleagues to a discussion does not give them power to change a process. State which choices they can make, which need the process owner’s approval and which constraints the experiment must respect. Make the reason for a closed choice explicit.

For the warehouse, the operations manager approves a trial on one packing route. Workers can choose the card’s practical layout and report when it obstructs the work. They cannot change dispatch approval rules. A named duty lead can suspend the trial if cases become difficult to trace. The manager protects time for the review and owns unresolved authority questions.

CIPD’s employee-voice guidance describes voice in terms of expressing views and influencing matters that affect work. The practical implication here is to show how input can alter the experiment and how the team will receive a response. A suggestion inbox without a response owner leaves that influence uncertain.

Use a change contract when responsibilities are disputed. Agree who can stop the work, who pays for the intervention, how colleagues can raise concerns privately and what will happen to existing duties during participation. A mandatory operational change can still have meaningful choices, but describe those choices honestly.

Generate options that address different explanations

Produce at least a few plausible options before choosing the most familiar one. In our example, these might be a pouch holding one case card on each trolley, a short verbal handoff at a fixed time, or a designated handoff coordinator. Another option is to clarify release authority without altering the physical handoff.

Compare options against the particular problem. A coordinator adds coverage demands. A verbal exchange may fail when shifts do not overlap. A case card can travel with the order but needs a place to survive movement and an owner for updates. None should be called the best option simply because it is inexpensive or easy to announce.

Write down the assumption each option depends on. The card option assumes that a small amount of current information travelling with the trolley will prevent the recipient from having to reconstruct its status. Use assumption mapping when several uncertain conditions could undermine the proposal.

Check for objections that reveal another explanation. If workers say the trolley is already overloaded, examine that condition. The guide to understanding responses to change can help keep an operational concern from becoming a label attached to the person raising it.

Write a test that can change your mind

An experiment needs a prediction, a defined action and a decision it can inform. The Institute for Healthcare Improvement’s Model for Improvement similarly connects an aim and measures with small tests through Plan–Do–Study–Act cycles. That is a complementary improvement approach, not the same method or evidence validating Lean Change.

For the fictional warehouse, define the trial before starting:

Part of the experiment	Proposed record	Why it matters
Question	Does a current case card make the ordinary handoff easier to understand?	Keeps the test separate from release-authority disputes
Boundary	One packing route, both shifts, five working days	Identifies whose work changes and when the review occurs
Prediction	Recipients can identify the order, missing item and next owner from the card	Defines the behaviour being examined
Observation	Check each eligible handoff; record missing information and recovery work	Looks beyond whether a card was attached
Stop condition	A carton cannot be traced, or the pouch obstructs safe handling	Provides a practical route back to the approved process
Review decision	Revise, test under another condition, stop or propose expansion	Prevents automatic rollout

These choices are illustrative, not validated thresholds. Define “eligible handoff” and what counts as complete information so that two observers would describe the same event similarly. Also note the time required to maintain the card and the recipient’s experience. A gain for one shift may impose hidden work on the other.

For a fuller account of comparison, observation and stopping rules, use the small-experiment design guide. An informal workplace test often cannot isolate causation, but it can still expose a missing prerequisite or rule out an impractical option.

A hypothetical warehouse experiment

The team attaches a pouch to one trolley route and agrees three fields for the card. Both shifts rehearse a handoff with an invented order before the trial. They keep the established fallback available, and the afternoon recipient can ask for help without having to prove that the card is wrong.

Suppose the first review finds that cards travel reliably but become stale when an item arrives after the morning shift has finished. The team also notices that the duty lead quietly corrects cards before the recipient sees them. The original prediction has therefore not been demonstrated under normal conditions. Extra supervision is doing some of the work.

The useful response is to revise the update responsibility and test again when the lead is occupied elsewhere. The team records what remains uncertain rather than reporting that the experiment succeeded because every trolley had a pouch. An unresolved release decision remains with the operations manager; it is not added to the card as an unofficial instruction.

Keep a short review record: prediction, observations, unexpected conditions, participant concerns, decision and owner. Use the pivot-or-persevere decision guide when evidence conflicts. A limited change to the test may be sufficient; a disappointing first attempt does not require replacing the entire strategy.

Expand only after checking the conditions

A trial on an ordinary afternoon says little about peak dispatch, temporary staff or a route with several simultaneous orders. Identify the differences that matter before adding another group. Preserve the instruction and fallback while testing those differences.

Weiner’s readiness theory distinguishes shared commitment from collective belief in implementation capability, and relates those beliefs to task demands, resources and the situation. It is a conceptual theory. Applied here, enthusiasm for the pouch does not establish that other routes have capacity to maintain it.

At each expansion decision, check whether the intervention needs more support, changes who owns a task or requires another approval. Communicate what was learned and what is changing through the change communication plan. Avoid presenting a provisional trial as a final requirement simply because a launch date was previously announced.

Boundaries and evidence

Lean Change supplies an organising approach and a practitioner vocabulary. The owner’s materials support its stated cycle and history; they do not establish that it outperforms every other change method. The worksheet and warehouse case above are original guidance and hypothetical illustration, not a validated instrument or reported study.

Experimentation is unsuitable when the proposed variation exceeds the team’s authority, removes a required safeguard or exposes people to consequences the team cannot responsibly manage. In such cases, use simulation, approved alternatives or a properly governed decision first. A small number of affected people does not make an intervention harmless.

The cycle can also become busywork: many trials, few decisions and no examination of the underlying policy. When successive interventions fail for the same reason, use double-loop learning to question the governing assumption. A shared card cannot repair a system that gives contradictory instructions about responsibility.

End each cycle with a proportionate decision and a named next review. The value of the record is that another person can see what was tried, what remains uncertain and why the team chose its next step.