← Insights

Continuous improvement

Kaizen that outlives the workshop

July 2026 · 6 min read

Everyone has seen a suggestion scheme die. It launches with posters and a pizza lunch. Ideas pour in for a month. Then the ideas stop being answered, the box gathers dust, and two years later a consultant recommends launching a suggestion scheme. The pattern is so reliable it has a lesson baked in: capturing ideas is trivial; keeping faith with them is the system.

Kaizen - continuous improvement through small, frequent changes - survives where suggestion schemes die because it treats each idea as a unit of work with an owner, a priority and a fate, not as a note in a box. In practice that requires three mechanisms, and all three fit in one honest spreadsheet.

Mechanism one: a priority that computes

The fastest way to kill an improvement pipeline is a monthly meeting that argues about which ideas to do. The argument repeats because it has no datum - loudest voice wins, and everyone else's idea sinks. The fix is old and unreasonably effective: score every idea on impact (1-5: how much it saves, in money, time or pain) and effort (1-5: how hard it is to do), and let the two scores place it in a quadrant:

• Quick win - high impact, low effort. Do these now; they fund the programme's credibility.
• Major project - high impact, high effort. Plan properly, resource properly - one or two at a time.
• Fill-in - low impact, low effort. Perfect for quiet Friday afternoons and new starters learning the process.
• Low priority - low impact, high effort. Park them without guilt; a recorded, honest "not now" keeps trust better than a silent never.

When the quadrant computes from the scores - automatically, visibly, the same way every time - the priority argument is settled at the moment of scoring. The meeting stops debating rank and starts assigning owners, which is the meeting you wanted.

Mechanism two: a lifecycle everyone can see

Every idea needs a status from a fixed lifecycle - raised, reviewed, approved, in progress, implemented, verified, closed (with rejected as an honest exit, used with a reason attached). The specific stages matter less than two properties: no idea is ever in limbo, and the raiser can always see where theirs sits. The moment ideas disappear into an unexplained void, contributions stop - that is the suggestion-box failure mode, and status visibility is its antidote. A status funnel on the dashboard (how many ideas sit at each stage) also shows management exactly where the pipeline clogs: lots of "approved" and little "in progress" means ideas are cheap and capacity is the constraint.

Mechanism three: verified savings, not promised ones

Estimated savings launch a kaizen programme. Verified savings keep it funded.

Record two numbers per idea: the estimated annual saving at approval, and the actual saving measured after implementation. The gap between the pipeline (sum of estimates) and the verified total (sum of actuals) is the programme's honesty metric. Verified numbers are almost always smaller - and infinitely more persuasive, because a finance director who has seen one honest £6,800 will believe the next estimate; one who has seen three inflated £50,000s will not. Verification also closes the learning loop: an idea whose saving did not materialise teaches you something about how you estimate.

Where the ideas come from - and where they graduate to

Seed the pipeline with the classic eight wastes - defects, overproduction, waiting, transport, inventory, motion, over-processing, and unused talent. Asking "where do you wait? what do you walk for? what do we make too early?" produces better ideas than "any suggestions?", and tagging each idea with its waste type gives you a Pareto that shows what your operation's signature problem is. A 5S audit's zeros and ones feed the same pipeline naturally.

And know the boundary of the tool. Kaizen ideas are improvements - better layouts, fewer touches, clearer standards. When an idea is really a problem - a recurring defect, an unexplained failure - it does not need an impact/effort score; it needs root cause analysis. That is the moment to graduate it to 5 Whys or a Fishbone, or a full 8D if a customer felt it. A healthy improvement culture runs both lanes and knows which is which.