Four complete A3 problem-solving case studies across four industries — manufacturing, software, healthcare, office — with every section filled in using realistic numbers. Each follows the canonical Toyota 7-section layout: Background · Current Condition · Goal · Root Cause · Countermeasures · Implementation Plan · Follow-up. Read these as templates for your own A3, not as theory. The A3 discipline is the one-page constraint: if your thinking does not fit, your thinking is not done.

The four case studies at a glance

Industry Problem Root cause Result after follow-up
Manufacturing Fastener cell cycle-time 18% over takt Tool-changeover variance + non-standard work +14% throughput, on-takt
Software / SRE On-call paging volume causing engineer burnout Alert thresholds set on raw metrics, not user impact Pages down 60%, sleep-disrupting pages near zero
Healthcare Emergency Dept door-to-doc time 47 min Triage bottleneck at single nurse station Door-to-doc 25 min (-22 min, -47%)
Office / Finance Re-bill rate at 8% (~$340k/quarter rework cost) No structured handoff between sales and billing Re-bill rate 2%, ~$255k/quarter saved

Each case study below is written as if you were reading the actual A3 sheet, section by section. Numbers and details are realistic but composite — the patterns reflect dozens of real Lean engagements, not any single client.

Case study 1 — Manufacturing: Fastener cell cycle-time reduction

Author: Cell production engineer · Coach: Plant Lean lead · Duration: 6 weeks

1. Background

Cell C-3 produces 14 SKUs of automotive fasteners (M6–M12 hex bolts and flanged screws) for a Tier-2 customer with weekly nominated quantities. The cell runs two shifts, five days, with a target takt of 38 seconds per part. Over the last quarter, cell C-3 has consistently run 18% over takt — observed actual 45 seconds — forcing weekend overtime to meet weekly commitment. Customer scorecard has flagged on-time-in-full at 91% versus the 98% target. Continued overrun risks losing 30% of the next-year contract.

2. Current condition

3. Goal / Target

Cycle time at or below 38 seconds (on takt) within 6 weeks, sustained for 4 consecutive weeks, with operator-to-operator variance under ±5%. On-time-in-full back to 98%.

4. Root cause analysis

Fishbone first to map the landscape (Methods, Machines, Materials, Manpower, Measurement, Environment), then 5 Whys on the dominant Methods branch.

Verified root cause: No process owner for standard work maintenance, leading to drift and operator-specific technique.

Secondary root cause: Tool-changeover variance (σ = 7 min) reflects no standard quick-changeover sequence — SMED principles never applied to this cell.

5. Countermeasures

  1. Reissue standard work sheets for all 14 SKUs with current process, validated by 2 senior operators and the manufacturing engineer.
  2. Assign cell production engineer as standing process owner with quarterly review cadence.
  3. SMED workshop on tool changeover — target reduce mean from 18 to 9 minutes, σ from 7 to 2.
  4. Daily start-of-shift standard-work confirmation by team leader (5 minutes, observation-based).
  5. Weekly cycle-time review at cell huddle with control chart on the cell-side board.

6. Implementation plan

ActionOwnerByVerify
Reissue standard work sheetsCell engineerWeek 12-operator + ME sign-off
Process-owner assignment postedPlant managerWeek 1Email + cell board
SMED workshop runLean leadWeek 2Before/after stopwatch study
Operator retraining (all 6)Cell engineerWeek 3Skills matrix updated
Daily standard-work confirmationTeam leaderWeek 3 onwardDaily checklist
Cycle-time control chart liveCell engineerWeek 4Board photo weekly

7. Follow-up

Week 6 measurement (n = 7,860 parts):

Sustaining actions: quarterly standard-work review by process owner; cycle-time control chart added to plant tier-2 daily metrics. Re-audit at month 3.

Case study 2 — Software: On-call paging burnout

Author: SRE team lead · Coach: Engineering director · Duration: 8 weeks

1. Background

The platform SRE team (8 engineers) rotates on-call for 12 production services (~140 alerts configured). In the last quarter, two engineers left the company citing on-call burden, exit interviews specifically called out the volume of after-hours pages. Engineering survey scores on "on-call is sustainable" dropped from 6.2 to 3.4 (out of 10). Hiring backfill is 4–5 months and the remaining 6 engineers are at risk of cascading burnout.

2. Current condition

3. Goal / Target

Reduce total pages per week to under 60 (a ~60% reduction), reduce sleep-disrupting pages to under 1 per engineer per week, and raise the "on-call is sustainable" survey score to ≥6 by week 8. Stop attrition driven by on-call burden.

4. Root cause analysis

Fishbone across Alerts, Services, Tooling, Process, then 5 Whys on the dominant Alerts branch.

Verified root cause: No team-level alerting policy or review cadence. Alerts proliferate based on infrastructure metrics rather than user-visible symptoms.

5. Countermeasures

  1. Adopt SLO-based alerting — page only when a user-visible SLO (latency, error rate, availability) is at risk of burn-rate violation. Move infrastructure-metric alerts to ticket queue, not pager.
  2. Quarterly alert review — team reviews every paging alert, kills the ones that did not produce a real incident response in the last 90 days.
  3. On-call retrospective at end of every rotation — departing on-call engineer reviews every page with the team, classifies as actionable / noise / informational, kills or downgrades the noise ones immediately.
  4. Auto-resolve for self-healing alerts — if the underlying signal returns to normal within 5 minutes, suppress the page entirely.
  5. Document on-call standards — one-page team policy: severity definitions, escalation rules, what should and should not page.

6. Implementation plan

ActionOwnerByVerify
SLO catalogue for top 5 servicesSenior SREWeek 2Approved by service owners
Migrate top 5 services to SLO alertingSRE rotaWeek 4Old alerts disabled, new alerts firing tested
End-of-rotation retrospective processTeam leadWeek 1 (every rotation)Doc + alert kill log
Auto-resolve config for noisy alertsTooling SREWeek 3PagerDuty rule deployed
On-call standards doc publishedTeam leadWeek 2Linked from runbook wiki
Repeat for remaining 7 servicesSRE rotaWeek 8All 12 services SLO-based

7. Follow-up

Week 8 measurement:

Sustaining actions: end-of-rotation retro is now standard process; quarterly alert review on the team calendar; on-call survey re-run quarterly.

Case study 3 — Healthcare: ED door-to-doc time

Author: ED nurse manager · Coach: Hospital Lean office · Duration: 10 weeks

1. Background

The Emergency Department of a 350-bed urban hospital sees 220 patients/day on average. Door-to-doc time (the interval from patient registration to first physician contact) currently averages 47 minutes — well above the 30-minute internal target and the published clinical-quality benchmark for an urban ED of this acuity mix. Patient satisfaction has dropped, left-without-being-seen rate climbed to 4.5%, and three serious-incident reviews in Q1 cited triage-stage delays as a contributing factor.

2. Current condition

3. Goal / Target

Mean door-to-doc time at or below 30 minutes within 10 weeks, sustained for 4 consecutive weeks. Left-without-being-seen rate at or below 2%. No triage-stage delays cited in incident reviews.

4. Root cause analysis

5 Whys on the triage queue (the dominant constraint visible in the value-stream walk).

Verified root cause: Triage capacity has not scaled with arrival rate; staffing and physical-space allocation reflect a 30%-lower baseline.

5. Countermeasures

  1. Open the second triage station during peak hours (10:00–22:00) using the renovated consultation room.
  2. Re-baseline nurse staffing against current census; add one nurse to the peak shift to staff the second station.
  3. Standard triage protocol — agreed ESI assessment sequence, target 6 minutes per patient (down from 8) without losing clinical rigor.
  4. Pull-trigger for opening second station — if waiting room exceeds 5 patients waiting for triage, charge nurse opens the second station immediately.
  5. Daily metrics board at nurse station with door-to-doc, LWBS, and patients-per-shift, updated each shift change.

6. Implementation plan

ActionOwnerByVerify
Re-baseline staffing modelED nurse manager + HRWeek 2Updated FTE plan signed
Renovate consult room → triage 2FacilitiesWeek 3Walkthrough sign-off
Standard triage protocol drafted + trainedLead RNWeek 4All triage RNs signed off
Pull-trigger rule liveCharge nursesWeek 5Posted at nurse station
Second station staffed peak hoursNurse schedulerWeek 6Schedule live, 4 weeks consistent
Daily metrics boardCharge nurse rotaWeek 6 onwardPhoto audit weekly

7. Follow-up

Week 10 measurement (n = 7,140 patients, weeks 7–10):

Sustaining actions: staffing model re-baseline annually against rolling 12-month census; daily metrics board permanent; LWBS reviewed weekly at ED leadership huddle.

Case study 4 — Office / Finance: Billing accuracy and re-bill rate

Author: Billing operations manager · Coach: CFO · Duration: 12 weeks

1. Background

The B2B services division of a mid-market firm bills 1,800 invoices per quarter (~$28M revenue). Currently 8% of invoices require a re-bill (correction and resend) due to errors in customer details, line items, contract reference, or pricing. Each re-bill costs an estimated $235 in finance team time and delays cash collection by an average of 11 days. CFO estimates the annualised cost of re-billing at ~$1.4M, plus 14 days of working-capital drag.

2. Current condition

3. Goal / Target

Re-bill rate at or below 2% within 12 weeks, sustained over 2 consecutive quarters. Estimated annualised cost reduction: ~$1.0M. Reduce DSO impact by 8 days on average.

4. Root cause analysis

Fishbone first across People, Process, Information, Tooling, then 5 Whys on the Information branch (the dominant cluster).

Verified root cause: No standard handoff between Sales and Billing, with no joint process owner. Errors are introduced at a free-form text boundary.

5. Countermeasures

  1. Standard order-to-cash handoff form in CRM — structured fields for contract reference, customer entity, line-item pricing, billing schedule. Sales rep cannot mark deal "closed-won" without completing it.
  2. Validation rules in CRM — contract reference must match an existing master-contract record; pricing must match contract pricing tables.
  3. Joint process owner — named Sales Ops + Billing co-leads for the order-to-cash handoff, monthly review.
  4. Daily exception report — any handoff form filed with missing/invalid fields routed to Sales Ops for fix before invoice issuance.
  5. Quarterly retrospective on top error categories with both Sales and Finance present.

6. Implementation plan

ActionOwnerByVerify
Handoff form designed (Sales + Billing)Joint process ownersWeek 3Approved by VP Sales + CFO
CRM custom object builtRevOps engineerWeek 5UAT pass
Validation rules deployedRevOps engineerWeek 6Test with 20 sample deals
Sales rep training (cohort of 24)Sales enablementWeek 7100% completed + signed off
Daily exception report liveBilling analystWeek 8Daily distribution + tracking log
Process owner monthly reviewJoint ownersWeek 8 onwardMeeting notes + action log

7. Follow-up

Week 12 measurement (Q-end, n = 1,810 invoices):

Sustaining actions: validation rules in CRM are now part of the deal-closure workflow; quarterly process-owner review continues. Next A3 cycle targeting the remaining 19% of error categories that the structured handoff did not address.

Common patterns across the four cases

Reading the four side-by-side reveals patterns that show up in nearly every A3, regardless of industry.

1. The current condition section is where most A3s fail

In all four examples, the current condition has at least 5 specific numbers with citation context (sample size, period, source). When a current condition reads "cycle time is too long" without a number, the rest of the A3 is built on sand. The discipline is to count, time, and measure before drafting the goal box.

2. Root cause hides one level deeper than you expect

The first three Whys are usually obvious to the team. The 4th and 5th Why — where ownership, organisational structure, or process design is the real cause — is where the actionable countermeasure lives. In the manufacturing case, "tool changeover takes too long" is the symptom; "no process owner for standard work" is the root cause. In the office case, "invoices have errors" is the symptom; "no joint process owner for order-to-cash" is the root cause.

3. Countermeasures should map one-to-one to root cause lines

Look at any of the four countermeasure lists — each numbered item directly addresses a verified root-cause finding. None of them are "train the team better" or "communicate more". Vague countermeasures are a sign that the root-cause work was not finished.

4. Implementation plans need owner, deadline, and verification — not just a deadline

The 4-column implementation tables (action / owner / by / verify) appear identical across cases. The verify column is what closes the loop — without it, "done" means "an email was sent" rather than "the change is in production and measured".

5. Follow-up uses the same units as current condition

Notice how each follow-up section measures the same metric, sample, and source as the current condition. This is non-negotiable: if your current condition is in minutes and your follow-up is in "significantly faster," you have not measured anything. The A3 closes when the metric you started with reaches the target you set.

Run the cause analysis live

Three of these four cases used 5 Whys or Fishbone (or both) in the root-cause box. Use the free tools to do the analysis interactively, then paste the result onto your A3 template.

Open the 5 Whys tool →

Common pitfalls these case studies avoid

Common questions

How do I scale this to my own A3?

Start by downloading the free A3 template. Then walk the 7 sections in order: do not write the countermeasure section until current condition and root cause are filled in with real data. The discipline is sequential — each section depends on the one before it. Treat the template as a forcing function, not a fill-in-the-blanks form.

What if my problem touches more than one team?

Cross-functional A3s are common — the office/billing case here is one. The convention is that the A3 has a single named author (the person closest to the problem) plus a coach (often a senior leader from a different function). Other functions appear in the implementation plan as named owners. The author owns the page; ownership of countermeasures distributes to the right teams.

Can I use A3 for a problem I already know the answer to?

You can, but the value drops considerably. A3 is most useful when the team thinks they know the answer but has not verified the root cause — the discipline forces them to slow down. If the answer is genuinely obvious and the data supports it, an A3 may be overkill; a one-page change request or kanban card is enough. Reserve A3 for problems where the cost of a wrong answer (or a too-shallow answer) is high.

How do these case studies relate to 8D?

All four are internal A3s — no customer involved, no external corrective-action obligation. If any of these problems hit a Tier-1 customer (say the manufacturing fastener-cell defects went out the door), the team would do the thinking on an A3 and then translate the finished thinking into the customer's 8D template for submission. See A3 vs 8D for the hybrid pattern.

What software should I use for digital A3s?

The simplest option is the downloadable A3 template in Excel, Word, or PDF. For collaborative real-time work, Miro, FigJam, and Mural all have A3-style templates. PowerPoint and Google Slides also work fine — the constraint is the one-page format, not the tool. Avoid tools that encourage scrolling; the visual gestalt of seeing the whole story at once is the point.

Do A3 case studies need to be this long?

The actual A3 sheet is one page — what you see here is the "A3 read aloud" with each section expanded for explanatory purposes. A real A3 in production fits all 7 sections onto a single landscape A3 sheet (297 × 420 mm). The verbosity here is for teaching; the discipline in practice is the opposite — severe compression onto one page.

What to read next