Toyota’s $20-an-Idea Experiment–and the Risk of Fixing the Wrong Problem
Twenty years ago this month, I recorded the first episode of Lean Blog Interviews. My guest was Norman Bodek, who spent decades studying Toyota and introducing American audiences to the work of people, including Taiichi Ohno and Shigeo Shingo.
I'm sharing a video excerpt from that conversation because the story still holds up–and because I think about it whenever a client asks how much they should pay employees for improvement ideas.
Here's the story Bodek told.
According Norman, someone in Toyota's North American operations decided to pay $20 per idea.
Not per idea implemented. Per idea submitted.
As Bodek described it, an employee could move a piece of paper from one side of a desk to the other, call that an idea, and collect $20. One employee became prolific enough to pay for a swimming pool for the summer.
The payment worked exactly as designed. It produced more submissions. The problem was that management wanted improvement, not paper moving from one side of a desk to the other.
Eventually, a senior manager stopped the incentive. Bodek speculated that it might have been Fujio Cho, but he wasn't certain. Bodek thought that response risked suppressing participation along with the dysfunctional payment. His proposed fix was more targeted:
“I'd only give $20 if it's worth $20.”
If the idea didn't justify the payment, he suggested thanking the employee and entering the idea into a monthly drawing for a small prize.
Bodek later said he thought this happened at Toyota's plants in Cambridge, Ontario, and Georgetown, Kentucky, but he wasn't certain. I haven't independently verified the location, how broadly the policy was used, or whether his proposed solution was adopted.
I'm treating it as his account rather than settled Toyota history. But the mistake he described is familiar to anyone who has managed an improvement program.
But for context, the Toyota Times reported in 2024 that the company's Creative Idea Suggestion System still offered a standard 500-yen reward for an improvement. The article does not mention or corroborate the $20 North American policy Bodek described. Because Toyota Times is published by Toyota, it is useful evidence of the company's more recent practice, not independent confirmation of the older story. So it seems Toyota is paying for improvements, but to the equivalent of about $3 US.
Paying by the Idea Pays for the Wrong Behavior
A flat payment per suggestion defines the submission as the unit of value.
The organization asks for ideas, counts ideas, and pays for ideas. It should not be surprised when people try to maximize the number they submit.
That isn't evidence that employees are greedy or dishonest. It shows that they understood the system management created.
The same problem can occur when organizations promise employees a percentage of estimated savings. I've long thought that's a bad idea.
People naturally focus on ideas that can be converted into dollars. Improvements involving safety, quality, patient experience, or reduced frustration become harder to justify because their value might not fit neatly into a spreadsheet.
Then come the arguments about who originated the idea, how much it supposedly saved, whether the savings are “hard” or “soft,” and who deserves what percentage.
The incentive system begins consuming attention that could have gone toward improving the work.
Small Recognition Is Different from Buying Ideas
In Healthcare Kaizen, Joe Swartz and I described a more carefully designed approach at Franciscan St. Francis Health.
Franciscan did not pay employees for merely submitting suggestions. Employees received 200 “VIP Points” after a Kaizen had been completed, documented, and approved by their supervisor. The points could be exchanged for merchandise or gift cards, and everyone listed as participating received them.
As we wrote:
“They have no direct cash payouts for ideas, and they reward for the process of doing a Kaizen.”
Franciscan also treated every completed Kaizen equally. A major cost-saving improvement received the same points as a small change involving the location of a stapler.
Some might find that unfair, but that was intentional. A small improvement might prevent an error, make work safer, reduce frustration, save a few minutes, or improve care for one patient. The benefit can be real even when nobody can calculate a credible ROI.
Other rewards were similarly modest: a $5 cafeteria or coffee-shop card, a Post-it pad, or a small notebook. At the beginning of the program, the entire incentive system cost about $1,000 a year. Four years later, it was approaching $4,000 annually–not much for an organization completing thousands of Kaizens.
Healthcare Kaizen describes Toyota's incentive system at the time as quite different from the $20-per-submission policy in Bodek's story. Supervisors could approve small payments for implemented ideas, while larger rewards required approval at progressively higher levels. Small improvements did not require an ROI calculation because benefits involving safety or quality could be valuable even when their financial impact was difficult to quantify.
The practical distinction is not cash versus no cash. Toyota's 500-yen reward and Franciscan's VIP points illustrate a different design: a small token tied to an improvement, rather than a flat payment large enough to make submitting ideas an end in itself.
Reward Implementation–and Learning
In Healthcare Kaizen, we offered three guidelines:
- Keep incentives relatively small.
- Base them on implementation, not mere submission.
- Consider recognizing a serious implementation attempt even when the proposed change is not ultimately adopted.
That third point matters. If improvement follows Plan-Do-Study-Act, not every test should be expected to succeed. An unsuccessful test can still produce useful learning.
The book also warns that monetary rewards can distort motivation when people begin focusing on the payout instead of the improvement. Some organizations had paid for mere suggestions rather than changes that were actually implemented.
The reward might increase activity while weakening the purpose behind it.
Fix the System, Not the Employee
The person who found the $20 loophole did not design the incentive.
Management decided that every submission was worth the same amount. Employees responded rationally.
A flat per-idea payment is essentially a piece-rate scheme. It says the unit of value is the submission, not the result. The organization gets exactly what it asked for: more submissions.
This fits Dr. Deming's broader criticism of quotas and pay-for-performance systems. Once management attaches a reward to a count, people will pay attention to the count–sometimes at the expense of the purpose behind it.
That is not a character flaw. It is feedback about the system.
Whether anyone at Toyota called it PDSA or not, management introduced a policy, observed consequences it had not intended, and changed course. The original plan was flawed. But they learned from it and adjusted.
The important question was whether management learned that employees could not be trusted–or that the incentive had made low-value submissions rational.
Bodek argued for the second interpretation. The incentive was broken. Asking employees to improve their work was not.
People Already Want the Work to Go Better
In healthcare–and in most workplaces–people already have reasons to want their work to improve.
They want fewer delays. Less rework. Safer conditions. Fewer hassles. Better service. More time to do meaningful work.
They usually do not need $20 to notice a problem.
They need a supervisor who listens, a practical way to test a small change, and help remove barriers. They need recognition and follow-through. They also need confidence that identifying a problem will not be treated as complaining, disloyalty, or an admission that they caused it.
Bodek didn't use the term psychological safety in that interview twenty years ago. But a sustainable improvement system depends on people believing two things: that it is safe to speak up and that speaking up will lead to something useful.
A $20 payment guarantees neither.
The lesson from the swimming pool isn't that organizations should never recognize or reward improvement. It's that leaders need to be precise about what they are encouraging.
Pay for submissions, and you will get submissions.
Recognize implementation, learning, and collaborative problem-solving, and you have a better chance of getting the improvement you actually wanted.
