Kaizen Savings Dashboard

What this tool does

CM8-360 follows improvement savings from the shop floor to the budget. It records what each improvement claimed, what finance validated, and whether a budget has actually been reduced; it splits the effect landing in this financial year from the full-year run rate; and it keeps cost reduction separate from cost avoidance, released capacity and working capital.

Everything runs inside this single file — no account, no upload, no network request of any kind.

Three numbers, never added together

The reason improvement programmes lose credibility is almost always the same: somebody presents a total that mixes what teams claimed with what finance agreed, and the finance director cannot find any of it in the accounts. This register keeps the three apart on every screen.

• Claimed — what the team that made the change says it is worth.
• Validated — what finance agrees, after looking at the evidence. Usually less.
• Banked — a budget has actually been reduced by it. This is the only figure a finance director recognises as a saving, and the gap between validated and banked is usually the largest of the three.

The reconciliation table exists to make the losses between stages visible and attributable.

What one row is

One row is one completed improvement. Record it when the change has been made, not when it was proposed — the funnel of ideas belongs in a different register. An improvement with two distinct savings in different cost categories is two rows, because they will be validated by different people and land in different budgets.

The five kinds of saving

• Cost reduction — spend actually falls. Somebody's budget can come down.
• Cost avoidance — spend that would have risen does not. Real, but no budget falls.
• Capacity released — hours or machine time freed up. Worth nothing until the freed time is used for something that earns.
• Working capital — stock or debtors turned into cash. A one-off cash effect, not a change to profit.
• Revenue — more sold or a price protected. Usually the hardest to attribute to an improvement and the easiest to overclaim.

Why avoidance is kept separate

Cost avoidance is genuine — replacing a discontinued part before it fails really does avoid an emergency rebuild — but it has a property that matters: no budget line will be lower next month. If avoidance is added to reduction in a headline figure, somebody eventually goes looking for the money, does not find it, and stops believing the whole programme.

The setting to count avoidance towards the target is off by default. Turn it on only if your finance team has agreed that is how the target is defined, and say so on the report.

Released capacity is not money

Four hours a week saved on a line that is not the constraint has produced four hours of standing about. That is not a criticism of the improvement — it may be the right thing to do, and it may become valuable the moment demand rises — but claiming it as a saving requires either that somebody left, that overtime fell, or that the hours went into work that earns.

The sample contains one of these: fifteen minutes off every changeover on a line that is not the constraint, validated as released capacity rather than as cost reduction. That is the honest treatment.

In-year against full-year

Both numbers matter and they answer different questions. The full-year effect is what the business is worth once the change has been in place a year — the right figure for judging the improvement. The in-year figure is what a budget holder will actually see this year — the right figure for a forecast.

Record the benefit start date separately from the completion date. They are often weeks apart, and using the completion date overstates the in-year effect every time. A change completed in the last month of a financial year contributes almost nothing to it.

What validation actually means

Validation is not a signature. It is somebody independent of the improvement agreeing that a measured before-and-after supports the figure. The tool enforces three things on any row marked validated: a date, a named person, and the evidence written down.

"Finance" is not a person, and the tool says so. A validation nobody's name is against is a validation nobody will defend when it is questioned six months later.

Set the hurdle so small savings can be claimed and accepted without ceremony, and large ones cannot. Anything above the hurdle still sitting as claimed is flagged, because unchecked large claims are what eventually get the programme audited.

The realisation rate

Most programmes land somewhere between 60% and 85% the first time this is measured. A rate above 95% usually means validation is not really happening; a rate below 50% means sizing is being done optimistically and the fix is to teach people how to size, not to shout at them.

Once you know your own rate, use it. A programme that says "we have claimed 400,000 and our claims historically validate at 70%" is far more credible than one quoting 400,000.

Banked

Mark a saving banked only when a budget has actually been reduced, a headcount removed, a contract renegotiated or a standard cost changed. That is a specific, checkable event.

Validated savings that are never banked accumulate quietly, and they are the reason a plant can run three years of improvement and show no change in its cost base. The register flags every one of them, and that list is the right thing to take to a budget conversation.

What this does not do

• It does not post anything to a ledger, and it is not an accounting record.
• It cannot tell whether a saving is real. That is what the evidence field and a named validator are for.
• It does not detect double-counting. Two improvements claiming the same hour will both count it, and only a person who knows both can spot that.
• It does not account for a saving decaying. Many recurring savings quietly stop after a year as habits return; re-validating the big ones annually is worth the effort.
• It assumes a month is a twelfth of a year for phasing, which is close enough for a forecast and not exact.

Printing and sharing

The Report tab prints the tiles, charts and both tables with a title block you fill in. The reconciliation table is the page to take to a finance review; the category chart is the one to take to a budget holder.

Saving your work

The register is held in this browser, on this computer, and stays there between visits. Use the backup button to write a JSON file you control; the spreadsheet download gives finance the same rows.

Accuracy & disclaimer

Every figure here is entered by you and the tool simply phases and totals it. Whether a saving is genuine, whether it has been counted twice, and whether it will still be there next year are all questions the register can prompt and cannot answer.