title: “Understanding Six Sigma — and Why Your Portfolio Company’s Close Cycle Is a DPMO Problem” date: 2026-05-26 slug: understanding-six-sigma-applied-to-finance type: concept quoted_authority: Jack Welch
Understanding Six Sigma — and Why Your Portfolio Company’s Close Cycle Is a DPMO Problem
A portfolio company tells you their monthly close lands “around the 15th, give or take.” That’s a 3-sigma answer to a 6-sigma question. The factories that invented Six Sigma — Motorola in 1986, GE in 1995 — would have called that variation uncontrolled, and they would have been right. Six Sigma exists because give or take is what kills enterprise value at exit.
The methodology, in one number
Six Sigma is a statistical methodology for eliminating defects and reducing variation. Its namesake target is 3.4 defects per million opportunities — 99.99966% yield. The sigma scale isn’t linear; it’s exponential.
| Sigma level | DPMO | Yield |
|---|---|---|
| 3σ | 66,807 | 93.3% |
| 4σ | 6,210 | 99.38% |
| 5σ | 233 | 99.977% |
| 6σ | 3.4 | 99.99966% |
A factory running at 3-sigma quality looks fine to a casual visitor — 93% of units pass. But 93% means 67,000 defects per million chances, which is why Motorola was hemorrhaging warranty costs in the mid-1980s before Bill Smith formalized the framework under CEO Bob Galvin. Within four years of implementation, Motorola had cut manufacturing defects by over 90% and saved $2.2 billion.
Six Sigma is not about cost reduction; it’s about customer satisfaction and process improvement. Costs fall as a byproduct. — Jack Welch
That quote is the whole point. Welch turned Six Sigma into GE’s operating system in 1995 and reported $12 billion in savings in the first five years — not by chasing costs, but by chasing variation. The cost savings were the residual.
Variation has two faces
Before you can attack variation, you have to know which kind you’re looking at.
Common-cause variation is inherent to the process. Normal wear, minor temperature shifts, small differences in how people do the work. It’s “in the system.” You don’t fix it by yelling at the operator — you fix it by redesigning the system.
Special-cause variation is assignable. Something specific happened: a vendor changed batches, a sensor drifted, someone skipped a step. You hunt it, name it, and eliminate it.
A bottling plant fills one-liter bottles. Common cause: a few millilitres of spread around the spec. Special cause: a new batch of caps with the wrong torque rating throwing the whole line off. Mistake the second for the first and you’ll waste a quarter retraining your operators when the actual fix is a phone call to the supplier.
The diagnostic tool is the 6Ms of cause analysis — Man, Machine, Method, Material, Measurement, Mother Nature — typically arranged on a Fishbone (Ishikawa) diagram, with 5 Whys drilling each branch down to root.
What this looks like in a finance department
Here’s the translation PE buyers don’t usually make: a finance department is a factory. Inputs (transactions, invoices, employee hours, bank feeds) get transformed into outputs (financial statements, AR aging, payroll, tax filings). Every transformation step has defect opportunities.
Take a monthly close at a $50M portfolio company. Roughly:
- 2,000 transactions posted per month
- 15 GL accounts that demand attention each close
- 8 reconciliations (cash, AR, AP, payroll, intercompany, inventory, accrued expenses, prepaids)
- 3 management reports generated
If even 0.5% of postings need a correcting JE — which sounds tiny — that’s 10 defects per month, or roughly 5,000 DPMO against the universe of (2,000 transactions × 1 critical attribute) opportunities. That puts the close cycle at about 4-sigma quality: better than typical, far from world-class.
Now overlay the symptoms a QoE team actually sees:
- Close lands on the 15th, not the 5th. (Cycle-time defect.)
- The trial balance reopens twice in audit. (Defect rate on first-pass close.)
- AR aging buckets shift $200K month-over-month with no narrative. (Measurement-system defect.)
- 6% of the income statement sits in Uncategorized at month-end. (Method defect — no chart-of-accounts discipline.)
Run the 6Ms on those four symptoms and the root cause is almost never the controller. It’s Method (no documented close checklist), Measurement (no DPMO scorecard, so nobody sees the variation), and Material (vendor invoices arrive in three different formats, two of which are unstructured PDFs). That’s a system problem, not a people problem.
When Bank of America ran Six Sigma against its operations in the early 2000s, customer complaints fell 24% and processing errors dropped 10%. When American Express applied it to card issuance, cycle time went from 22 days to 7. The same playbook works on a close cycle, an AR aging, or a 13-week cash forecast.
What this means for diligence
The thing PE buyers should ask is not “How fast is your close?” — that’s a vanity metric. The thing to ask is:
“What is your defect rate, what is your variation, and what is your process for hunting special causes?”
Concretely, in a Q&A or operational diligence session, three questions surface the truth:
- How many correcting journal entries did you post against last quarter’s books? A real number means they’re measuring; “we don’t track that” means they aren’t.
- What’s the standard deviation of your close-cycle days over the last 12 months? Six Sigma’s whole point is that the spread matters more than the mean. A close that lands “5 ± 7” days is worse than one that lands “12 ± 1.”
- When you find an error, do you trace it back to a 6M category, or do you just fix it? The first answer is a Six Sigma shop. The second is a cleanup shop.
A target whose CFO can answer all three crisply is running a finance function with statistical-process-control discipline. That target’s working-capital surprises, post-close adjustments, and audit-period drift will be measurably smaller than the average diligence experience. That’s enterprise value you don’t have to claw back at the True-Up.
A target whose CFO cannot answer those questions is offering you a 3-sigma close cycle dressed up in 6-sigma vocabulary. Price accordingly.
This is the discipline we apply to every Kaizen CFO engagement. Talk to sales →