Lean Six Sigma Integration
Lean Six Sigma integrates two disciplines that look similar but solve opposite problems. Lean attacks waste and flow (speed, inventory, handoffs) — its instinct is 'do less.' Six Sigma attacks variation and defects (consistency, capability) — its instinct is 'do it the same way every time.' A pure-Lean shop will speed up a defective process. A pure-Six-Sigma shop will perfect a slow process nobody cares about. Integration means using Value Stream Mapping to find where waste lives, then DMAIC where defects in those waste-points have a dollar number attached. The right sequence: Lean first to remove the easy 30-50% of cycle time, then Six Sigma where defects remain in the simplified flow.
The Trap
Treating Lean Six Sigma as a single rebranded methodology run by a 'Belt' org. In reality the disciplines pull against each other in early projects: Lean wants you to remove the inspection step (waste), Six Sigma wants you to add measurement (data). Teams that don't pick which problem they're solving end up with bloated charters, 9-month projects, and savings on a slide deck nobody trusts. The other failure: applying Lean Six Sigma to processes that aren't yet stable enough to benefit. If your process changes monthly because the product is still finding fit, neither discipline applies — you're optimizing a moving target.
What to Do
Sequence: (1) Walk the process at the gemba and Value Stream Map it — annotate cycle time, wait time, and defect rate at every step. (2) Calculate Process Cycle Efficiency = Value-Add Time / Total Lead Time. If under 25%, Lean problem (flow). If over 50% but defect-laden, Six Sigma problem (variation). Most processes are under 5% PCE — start Lean. (3) After Lean removes obvious waste, re-measure. (4) Charter DMAIC ONLY on remaining bottlenecks where defects cost over $250K/year. (5) Embed control charts and standard work together — Six Sigma keeps the gain, Lean keeps the flow. Refuse projects that can't name a sponsor and a dollar target.
Formula
In Practice
GE under Welch (1995-2001) ran pure Six Sigma and reported $12B savings, but executives privately admitted that mature divisions had hit diminishing returns by year 4 — defect rates were already low and the next sigma cost more than it saved. Honeywell merged with AlliedSignal in 1999 and combined Lean (from AlliedSignal's TPS-influenced operations) with Six Sigma (from Honeywell's quality program) into 'Lean Six Sigma' — the first formal integration. The combined methodology produced $1.2B in annual savings by 2003 and became the dominant operating system for Fortune 500 ops shops. Critics note the same thing as pure Six Sigma: it optimizes existing operations brilliantly but doesn't defend against disruption.
Pro Tips
- 01
If you've never mapped your value stream, don't run a single DMAIC project. You'll fix a defect in a step that should have been deleted. Walk the process first; measure variation second.
- 02
Belt certification is NOT the integration. The integration is in project SELECTION: every charter should explicitly state whether the primary lever is flow (Lean) or capability (Six Sigma). 'Both' usually means 'neither rigorously.'
- 03
The classic 80/20: in service operations, 80% of available savings come from Lean (waste, queues, batch sizing) and 20% from Six Sigma. In high-volume manufacturing, the ratio inverts. Don't import a manufacturing-belt template into an office.
Myth vs Reality
Myth
“Lean Six Sigma is one unified methodology”
Reality
It is two methods with overlapping vocabulary and opposite default moves. Practitioners who blur them end up running consulting theater. The integration is project-portfolio level — knowing which method to deploy where — not a fused toolkit.
Myth
“You need belts to do Lean Six Sigma”
Reality
Toyota — the original lean enterprise — has no Black Belt program. Process owners run improvement. Belts make sense when you need to industrialize the methodology across thousands of people fast (GE's situation in 1996), but they create a parallel quality bureaucracy that competes with line ownership when overdone.
Try it
Run the numbers.
Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.
Knowledge Check
A loan-origination process has a total lead time of 18 days, with only 6 hours of actual value-add work. What is the Process Cycle Efficiency, and which discipline should you apply first?
Industry benchmarks
Is your number good?
Calibrate against real-world tiers. Use these ranges as targets — not absolutes.
Process Cycle Efficiency by Industry
Lean Enterprise Institute benchmarks across industriesWorld-Class Manufacturing
> 25%
Good Service Operations
10-25%
Typical Office Process
1-10%
Bureaucratic / Permit Process
0.1-1%
Fully Broken
< 0.1%
Source: Lean Enterprise Institute / Michael L. George, Lean Six Sigma
Real-world cases
Companies that lived this.
Verified narratives with the numbers that prove (or break) the concept.
Honeywell (Lean Six Sigma origin)
1999-2003
Following the AlliedSignal-Honeywell merger in 1999, the combined company had two operating systems: AlliedSignal's TPS-influenced Lean program and Honeywell's Six Sigma program. CEO Larry Bossidy formally fused them into 'Lean Six Sigma' as a single methodology — the first time the combination got an official name. By 2003, the program reported $1.2B in annual savings. The integration model became the template Fortune 500 ops shops still use today. Critically, Honeywell did not build a separate quality bureaucracy — they trained line managers in both methods and forced project charters to declare whether the primary lever was flow or variation.
Annual Savings (2003)
$1.2B
Methodology Birth
1999-2000 post-merger
Project Charter Discipline
Required Lean vs Six Sigma declaration
Belt Bureaucracy
Avoided — line ownership instead
The Honeywell model worked because it kept project ownership with the process owner. The integration was in selection logic, not in a fused training program.
GE under Welch (Pure Six Sigma)
1995-2001
GE ran pure Six Sigma without a Lean component for the first 5 years and reported $12B in cumulative savings. By year 4-5, internal benchmarking showed mature divisions hitting diminishing returns: defect rates were already under 4 sigma in many lines, and the next sigma cost more in process investment than it returned. GE quietly added Lean tooling around 2000 (called 'Lean Six Sigma' internally by 2002) but the program had already lost momentum because it had been pitched as a defect-reduction silver bullet. The lesson: ignoring flow means you optimize the variation in slow, wasteful processes — and you run out of defect-savings before you've removed the bigger waste.
Pure Six Sigma Period
1995-2000 (5 years)
Reported Savings
$12B cumulative
Diminishing Returns
Visible by year 4
Lean Added
~2000-2002 (too late)
Six Sigma without Lean is a fixed playbook against a moving constraint. GE spent 5 years optimizing variation in processes that were structurally bloated.
NUMMI Plant (Toyota-GM Joint Venture)
1984-2010
The NUMMI plant in Fremont, California was a joint venture where Toyota took over a GM plant infamous for terrible quality, absenteeism, and labor disputes. Toyota brought TPS — an integrated lean-and-quality system that predates the formal 'Lean Six Sigma' label by 20 years. Without firing a single line worker, Toyota raised defect-free first-pass yield from 35% to over 90% within 18 months and cut absenteeism from 25% to under 3%. They did NOT use belts, certifications, or DMAIC. They used andon cords, standard work, gemba walks, and process-owner accountability. NUMMI is the proof that integration is cultural, not bureaucratic.
First-Pass Yield (Before)
35%
First-Pass Yield (After 18mo)
90%+
Absenteeism (Before)
25%
Absenteeism (After)
Under 3%
The deepest Lean Six Sigma integration in history had no belts, no DMAIC, and no quality department. It had standard work, andon cords, and managers who walked the floor.
Decision scenario
The Lean Six Sigma Program Design
You are a newly hired COO at a $1.2B specialty insurer. The CEO has approved a $2M Lean Six Sigma program based on a McKinsey deck and wants 50 Black Belts trained in year one. Customer complaints concentrate around slow claim turnaround (avg 22 days) and inconsistent settlement amounts on similar claims. Your first VSM walk shows 19.5 of the 22 days are queue time across 6 handoffs.
Annual Premium
$1.2B
Avg Claim Lead Time
22 days
Process Cycle Efficiency
Under 1%
CEO-Proposed Spend
$2M (training)
Defect Categories
Settlement variance + lateness
Decision 1
You can spend the $2M on the Belt program as proposed, or counter-propose. The CEO has political capital invested in the announcement. Your VSM data says flow is the binding constraint — but settlement variance is a real Six Sigma problem too.
Execute the CEO's plan — train 50 Black Belts in year one and deploy them on top complaint categoriesReveal
Counter-propose to the CEO: 6-week Lean kaizen first to collapse handoffs and queue time. Then a smaller, focused Black Belt program (8 belts) targeting the residual variation problems with audited dollar charters.✓ OptimalReveal
Related concepts
Keep connecting.
The concepts that orbit this one — each one sharpens the others.
Beyond the concept
Turn Lean Six Sigma Integration into a live operating decision.
Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.
Typical response time: 24h · No retainer required
Turn Lean Six Sigma Integration into a live operating decision.
Use Lean Six Sigma Integration as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.