Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / Why GE Failed at Six Sigma
Transcript
- Lucas: So, everyone knows the Six Sigma success story at General Electric. Jack Welch called it 'the most important initiative in the history of the company.' But I want to talk about why that narrative is incomplete — and how GE's fanatical embrace of Six Sigma actually planted seeds for its later collapse. Luna: That's a bold take. I mean, GE was the poster child. Every consulting firm used them as the case study. Are you saying the quality revolution was a mirage? Lucas: Not a mirage — more like a distortion. Six Sigma gave GE a lot of genuine operational efficiency. But look at the numbers: by 1998, Welch claimed the initiative saved the company $350 million. Impressive, right? But when you dig into the details, those savings were mostly from incremental tweaks to existing processes — not new products, not new markets. Luna: So you're saying the cost-cutting was real, but the innovation pipeline dried up? Lucas: Exactly. And it's not just my opinion. After Welch retired, several former GE executives went on record saying that the relentless focus on reducing variance made the culture allergic to experimentation. If a project couldn't be measured in defects per million, it didn't get funded. And that's lethal when you're running a technology-driven conglomerate. Luna: I remember reading that GE's R&D budget actually shrank as a percentage of revenue during the late 1990s. Meanwhile, competitors like Siemens were investing more heavily in fundamental research. Lucas: That's a key datapoint. By 2000, GE's R&D spend was around 2 percent of revenue. Siemens was closer to 7 percent. The difference shows up in their patent output and, more importantly, in their ability to pivot into new industries. GE never really capitalized on the internet boom the way a company its size should have. Luna: But Welch's defenders would say Six Sigma is why GE Capital became so profitable — they applied the same rigor to financial processes. Lucas: They did, and that's actually the other side of the problem. GE Capital's success under Welch fed the narrative that you could apply manufacturing discipline to any business. But financial risk doesn't follow a normal distribution. When you treat loan defaults as a defect rate you can control through process, you miss the fat tails — and we all saw how that ended in 2008. Luna: So the same toolkit that worked on assembly lines was dangerous in a finance context. That's a really important distinction. Lucas: And it's not just finance. Look at GE's appliance division. They were famous for reducing defects on refrigerators and washing machines. But by the 2000s, they were late to the market with smart appliances and energy-efficient designs. The competition from LG and Samsung ate their lunch. Why? Because those companies were willing to launch products with a few bugs, iterate fast, and learn from customers. GE's process wouldn't allow a product to ship until it hit six sigma quality — which meant they were always a generation behind. Luna: There's a great quote from a former GE engineer: 'We spent so much time making the perfect mousetrap that we forgot the mice had evolved.' Lucas: That's exactly the problem. And here's the thing — Six Sigma is not bad. It's a powerful tool for reducing variation in high-volume, repetitive processes. But when it becomes the religion that overrides all other business logic, you get a culture that punishes failure so severely that nobody attempts anything risky. Luna: So what's the alternative? How do you balance process rigor with the need for breakthrough innovation? Lucas: There's no single answer, but I think the best companies build a 'two-speed' culture. You have a core business that runs on process excellence — think Toyota's production system — but you also carve out separate spaces where experimentation is protected. Amazon does this with their 'two-pizza teams.' Google had their famous 20 percent time. Apple under Jobs was notoriously anti-process in product development, but incredibly disciplined in manufacturing. Luna: I want to talk about Toyota, because they're often lumped in with Six Sigma, but they're really different. Their lean manufacturing is more about continuous improvement and respecting people, not just statistical process control. Lucas: Right. And that's a critical distinction. Toyota's system empowers workers on the line to stop production if they see a defect. That's about quality, but it's also about autonomy and problem-solving. Six Sigma, as implemented at GE, was much more top-down. You had 'Black Belts' and 'Master Black Belts' who were essentially process cops. It created a hierarchy of expertise that often alienated the people doing the actual work. Luna: So Welch's GE was more about control than empowerment. And that might have worked in the 1980s when they were restructuring, but it became a liability as the economy shifted toward services and software. Lucas: Exactly. And the cultural legacy persisted long after Welch left. Jeff Immelt tried to pivot GE toward digital and industrial internet, but the muscle memory of Six Sigma was so ingrained that every new initiative got bogged down in process. Immelt himself said in his book that GE's 'process orientation' made it hard to move fast. Luna: There's a great case study in GE's healthcare division. They had a breakthrough MRI technology in the lab, but it took years to bring to market because the Six Sigma team kept demanding more data on defect rates. Meanwhile, Siemens and Philips launched similar products faster. Lucas: I hadn't heard that specific example, but it fits the pattern perfectly. The cost of delay is something Six Sigma doesn't measure. If you obsess over the defect rate of a new product, you might miss the fact that your competitors are already capturing the market with a product that's 'good enough.' Luna: So for listeners who work in operations or quality — should they abandon Six Sigma entirely? Lucas: Absolutely not. But you have to use it as one tool in a larger toolkit, not as the operating system of your entire business. If you're running a factory that makes the same widget a million times a year, Six Sigma is your best friend. If you're building a new software platform or entering a new market, you probably want a more agile approach. Know when to optimize and when to explore. Luna: That's a great framing. And it reminds me of something I read about 3M. They famously had a '15 percent rule' — engineers could spend 15 percent of their time on projects outside their core responsibilities. That's how Post-it Notes were invented. But when they hired a CEO from GE who brought Six Sigma thinking, that rule got squeezed, and innovation suffered. Lucas: Yeah, the 3M story is textbook. Jim McNerney, who came from GE, tried to apply Six Sigma discipline to 3M's R&D. The result? More efficient, incremental improvements, but fewer breakthrough products. The board eventually reversed course after McNerney left. It's a cautionary tale about assuming that what worked at one company will work at another. Luna: So the lesson is: culture eats process for breakfast. And if your process is designed to eliminate all variation, you're also eliminating the kind of variation that leads to discovery. Lucas: Exactly. And that's the big takeaway for anyone in business execution. Every tool has a domain where it's valuable and a domain where it's destructive. The art is knowing the difference. Six Sigma is a fantastic way to make a mature process more efficient. It's a terrible way to invent the future. Luna: If today's episode was useful to you and you want to keep it ad-free, buy me a coffee dot com slash fexingo helps. Lucas: Yeah, that's the link. Even a few dollars makes a real difference in keeping these conversations going. Luna: So, let's come back to a practical question: if you're a manager at a mid-sized company that's considering a Six Sigma program, what should you watch out for? Lucas: First, ask yourself what problem you're trying to solve. If it's high defect rates in a stable process, Six Sigma makes sense. If it's lack of innovation or slow time to market, you need a different approach. Second, don't let the black belts take over. Make sure the program is owned by line managers, not a separate quality police force. And third, keep a separate budget for high-risk, high-reward projects that don't meet the Six Sigma criteria. Protect your moonshots. Luna: Are there companies today that you think are doing this well — balancing process rigor with innovation? Lucas: I think Toyota still does it well, despite their recent challenges. Their production system is process-oriented, but it's built on respect for people and continuous improvement, not just defect reduction. Also, companies like Danaher have built a culture around the Danaher Business System, which borrows from both lean and Six Sigma but emphasizes growth and innovation alongside efficiency. Luna: And on the tech side, I'd point to companies like Intuit. They've adopted a 'design for delight' approach that uses data and experimentation, but they also have rigorous processes for their financial products. They've managed to keep both cultures alive. Lucas: That's a great example. The key is to have a portfolio approach to improvement. Not every project needs to be a six sigma project. Some need to be lean startup projects, some need to be R&D moonshots, and some need to be cost reduction initiatives. The mistake is treating one methodology as the universal answer. Luna: So the takeaway for our listeners: be skeptical of any single management fad, no matter how many success stories it has. Look for the hidden costs — in innovation, in culture, in agility. Lucas: And remember that the best-managed companies are the ones that know when to be a machine and when to be a garden. Machines produce consistent output. Gardens produce surprising growth. You need both. Luna: That's a great note to end on. Thanks for the conversation, Lucas. Lucas: Thanks, Luna. And thanks to everyone listening. We'll be back tomorrow with another episode of The Operator's Manual.