Kaizen events
Business case
Business case is licensed CC BY 4.0. Attribution: Katafacts (katafacts.com).
Customise with AISkill file ↓1 · What it is
What it is
A business case puts a real number on why an investment is worth making — the cost of the status quo, what the fix actually costs, and the quantified value it returns, with payback period and 3-year ROI computed from those numbers rather than asserted. Every value driver carries how the number was derived and how confident you actually are in it — a driver with no real basis is named as a gap, not hidden behind a confident-sounding total. It's the resourcing step that naturally comes before a kaizen event charter: before committing a team's time to an improvement, this is what earns that commitment.
2 · When to use it
When to use it — and when not to
Use it when
- You're proposing a real investment — capital, a new system, dedicated time — and need more than 'this seems worth it' to get it approved.
- You want to know honestly which parts of the case are backed by real data and which are still an estimate worth flagging.
- The investment is heading toward a decision-maker who will ask how you got these numbers, not just what they add up to.
Not when
- The investment is trivial enough that building a formal case costs more time than the decision itself is worth — use judgement, not process, for a small call.
- You don't have any real data yet, just a hunch that something would help — go gather evidence first (a gemba walk, a VOC pass), or every driver in this case will be a guess dressed up as a number.
- You're building the countermeasure itself, not funding it — that's an A3's job; a business case exists to justify committing resources, not to solve the problem directly.
3 · How to fill it in
How to fill it in
- Cost of inaction
- What does the current state keep costing, in real terms, and why does that matter now rather than later?
- Investment
- Describe roughly what the fix costs — one-time and ongoing — and a first-pass itemized estimate gets drafted for you to correct with real numbers.
- Value drivers
- Drafted from your description of the proposal — each with how the number was derived and how confident it actually is. Don't fill these in directly; describe the proposal and let a first draft come back.
- ROI summary
- Computed for you from investment and value drivers — nothing to enter here.
- Assumptions
- Drafted alongside the value drivers — what has to stay true for these numbers to hold.
- Narrative
- Drafted executive summary — the shape of the case in words, never restating the computed numbers itself.
- Next steps
- What needs firming up before this case goes to whoever approves it, ranked by impact and effort, owned and dated.
4 · What good looks like
What good looks like
The example below justifies an automated inspection station for a solder-defect problem — two of three value drivers are grounded in real quality-log and time-study data, and the third, a faster product-launch estimate, is honestly left at low confidence rather than padded to make the total look stronger.
Same example, as a downloadable xlsx workbook.
Download .xlsxBusiness case · SMT Line 2 — AOI investment
Automated optical inspection — SMT Line 2 business case
Marcus Lee, Quality Manager · 2026-03-05
Team: Elena Cho, Process Engineer · Sponsor: Dana Ruiz, Operations Manager
Add a second automated optical inspection station immediately after reflow on SMT Line 2 to catch solder-joint defects before they reach final test and the customer, instead of relying on final-test inspection alone.
Cost of inaction
SMT Line 2 has averaged 3.5 solder-joint warranty claims per month for the last six months, each costing roughly $1,000 in parts, freight, and technician time to resolve — and a rework technician spends close to two hours a day on final-test rework a station-level catch would prevent entirely. Neither cost goes away on its own; it recurs every month the line runs without earlier detection.
Investment
| Description | Amount | Timing |
|---|---|---|
| AOI machine purchase and line integration | $85,000 | one-time |
| Annual service and calibration contract | $6,000 | annual |
Value drivers
Eliminate warranty claims from solder-joint defects escaping to the customer
cost savings · $42,000/yr · Confidence: high
Basis: Quality log shows an average of 3.5 solder-joint warranty claims per month over the last six months, at roughly $1,000 each in parts, freight, and technician time.
Reduce final-test rework labor currently spent catching these defects downstream
productivity · $18,000/yr · Confidence: medium
Basis: Rework technician time study shows roughly 2 hours/day on solder-joint rework, at a fully loaded $45/hr rate across 250 working days.
Faster new-product introduction from catching defects earlier in the line
Low confidencerevenue growth · $10,000/yr · Confidence: low
Basis: Engineering's rough estimate that earlier detection could shave about a week off each of two NPI cycles per year — not yet validated against a real product launch.
ROI summary
Payback period
15.6 months
3-year ROI
103.9%
3-year net value
$107,000
Total annual value
$70,000
1st-year investment
$91,000
Low-confidence, not yet proven
Faster new-product introduction from catching defects earlier in the line
Assumptions
- The defect rate stays consistent with the last six months' quality log, not a temporary spike.
- The rework technician's freed time is genuinely redeployed to other value-add work, not just idle.
- The NPI cycle-time saving needs validation against a real product launch before it's counted on in the final case.
Narrative
Two of the three value drivers are well-grounded — the warranty-claim reduction comes straight from six months of quality-log data, and the rework-labor saving from an actual time study. The third, faster new-product introduction, is honestly the weakest: it's engineering's rough estimate, not yet tested against a real launch. Even without it, the warranty-claim reduction alone covers most of the case on its own, which is worth stating plainly rather than leaning on the softest number to make the total look better.
Next steps
Validate the NPI cycle-time estimate against the next real product launch before counting on it in the final case presented to finance.
Linked finding: Faster new-product introduction from catching defects earlier — low confidence
Impact: medium · Effort: low · Owner: Marcus Lee · Due: 2026-04-15
Confirm with the operations manager that the rework technician's freed time will actually be redeployed, not left idle.
Linked finding: Reduce final-test rework labor — medium confidence
Impact: high · Effort: low · Owner: Dana Ruiz · Due: 2026-03-20
5 · Common mistakes
Common mistakes
Writing a value driver with no real basis behind it.
A number with nothing behind it gets challenged the moment someone asks how you got it — and once one number falls apart, the reader stops trusting the rest of the case too.
Padding the case with low-confidence drivers to inflate the total.
The honest total, with its weak points named, survives scrutiny. A total padded with shaky numbers just gives a skeptical reviewer a reason to distrust the whole case.
Skipping the cost of inaction and only presenting the upside.
A case that only shows what you gain, never what continuing to wait already costs, is missing half the argument for acting now instead of later.
6 · What it connects to
What it connects to
upstream
A3 problem solving
A countermeasure identified in an A3 that needs real resourcing is exactly what turns into a business case — the A3 finds the fix, this justifies committing to it.
downstream
Kaizen event charter
Once a business case earns the investment, a kaizen event charter is where the actual work gets planned and resourced day to day.
7 · Where AI helps
Where AI helps
Judgement — stays yours
- Deciding whether a drafted value driver's basis is actually solid enough to stand behind
- Deciding which low-confidence driver to firm up before presenting the case
Analysis — AI helps
- Drafting itemized investment and value drivers from a description of the proposal
- Drafting the narrative from the case actually built
Drudgery — automated
- Computing payback period and 3-year ROI from the numbers
- Tallying which value drivers are still low-confidence
- Exporting to xlsx in the house format
9 · Rate this kata