Most business processes weren’t designed around what’s best. They were designed around what was possible. When AI removes the constraint, most businesses keep the process anyway, and that’s one of the main reasons AI projects fail. Here’s what a postwar statistician, Toyota and a car plant full of vodka thermoses can teach us about it.
This came out of my conversation with Ed Ortega on Slideshow with Dave Hayward. Ed is a partner at Machine & Folk, an AI strategy and development firm, with 20 years of Silicon Valley behind him. He’s relentlessly curious, brilliant without being annoying about it, and what we call in New Zealand a bloody good bloke.
Deming and the idea of building quality into the process
In 1950, an American statistician named W. Edwards Deming flew to Tokyo. Japan’s factories were rubble, and its business leaders turned to him to help rebuild. His big idea: stop inspecting for quality at the end of the line. Build quality into the process from the start, and fix problems where they happen, not downstream after they’ve multiplied.
The Japanese took him extremely seriously. Toyota won the Deming Prize in 1965, and its chairman later said: “There is not a day that goes by that I do not think about what Dr. Deming means to our management.” America, busy winning the postwar economy, largely ignored him.
The line goes… stop
Out of Deming’s thinking came the Toyota Production System, and with it the Andon cord. Any worker on the line could pull it the moment something went wrong. Teammates would swarm, fix the problem at the source and keep moving. Across Toyota’s plants, the cord was pulled every few seconds. Stopping constantly made them faster.
American factories ran on the opposite logic: the line never stops. Faults were carried forward, and cars rolled into holding yards with problems baked in from station one. Workers told stories of engines put in backwards.
Same people, different results
In 1984, GM handed Toyota the keys to its worst plant, in Fremont, California. It was notorious for absenteeism and worse. On any given day, one in five workers didn’t show up, and some brought thermoses of orange juice and vodka to the line.
Toyota kept almost the entire workforce and flew them to Japan in groups of 30 to learn a completely different way of working. Same people, same plant, and defect rates dropped to match Toyota’s plants in Japan.
That thinking became Kaizen, which informed a 1986 Harvard Business Review article on iterative teams, which directly shaped the Agile Manifesto in 2001. A straight line from a postwar Tokyo lecture hall to your sprint planning. And now, to AI.
Why AI projects fail: the constraint is gone, the process remains
On the episode, Ed and I watched a tool called Pencil build a fully designed CRM interface from a single prompt, in about ten minutes.
The more interesting conversation was about why the designer and developer handoff exists at all. Designers work in vectors. Developers work in code. The briefs, the mark-ups, the tickets and the “this corner should be 8 pixels, not 4” back-and-forth weren’t design decisions. They were a constraint: two professional languages with no shared ground, and an expensive process built around the gap. Tools like Pencil close that gap, and teams working this way are sprinting in a straight line while everyone else jogs the long way round.
Ed also described a project where his team automated an enterprise data extraction system. Halfway through, they realised the whole workflow had been built around a human bottleneck that AI had just removed. The database structure, the roles, the process: all of it was shaped by a limitation that no longer existed.
The people involved were freed from repetitive work that machines do better, and moved to higher-value work that’s more satisfying and more billable. Every decision draws on the same limited pool of mental energy. When people are freed from work that only existed because of a constraint, you don’t just get time back. You get thinking back.
The real risk with AI
GM sent its own people to Fremont. They watched the transformation happen with their own workforce, and had every opportunity to take it back to their other plants. Most didn’t, or couldn’t. The constraint was gone, but the habits, assumptions and org structures stayed.
That’s the real risk with AI right now. The tools are changing fast. The processes, less so. The Fremont workers weren’t bad workers. They were good workers trapped in a bad system.
If you want to think through where AI and marketing can do strategic work in your business, book 30 minutes with me.
Questions people ask
Why do so many AI projects fail?
Often because the tool is added to a process that was designed around an old constraint. The constraint disappears, but the process, roles and habits stay the same, so the gains never arrive.
What can AI adoption learn from lean manufacturing?
Fix problems at the source and redesign the system, not just the tools. Toyota’s turnaround of the Fremont plant came from changing how people worked, with the same people.
Where should a business start with AI?
Look for work that only exists because of a limitation, such as handoffs between teams or manual data wrangling. Those are the processes most worth redesigning.
This article first appeared in Bright Objects, my fortnightly newsletter on brand, marketing strategy and AI for business and tech leaders. Subscribe here.
Sources and further reading
- Ira Glass, “NUMMI”, This American Life, Episode 403 (2010)
- “Toyota’s Management History”, The Deming Institute
- Amy Edmondson, “Psychological Safety #79: The Andon Cord”
- Machine & Folk and Pencil
- Slideshow with Dave Hayward, Ed Ortega (Season 2, Episode 2)
If you know your business is all steak and needs some sizzle….. we’d love to hear from you.
Dave Hayward
Subscribe to Bright Objects Your essential marketing & tech insights
Warm personal stories wrapped around solid business, revenue and marketing strategy, how-tos, technology discussion (especially AI), philosophies and tactics. Occasionally, we’ll talk about personal productivity and things important to us (like astronomy and dogs).
