JUST SAY THE WORD – Mendomi teaches that trust, not just competence, is what makes teams and organizations (and AI adoption) truly work, explains the author as it dives into another term from his lean glossary.

Words: Michael Ballé

Roberto Priolo: Mendomi is an interesting concept. It may not be listed in traditional lean lexicons, but it has a clear connection to the idea of respect for people. If it's about providing the right conditions for people to do their job as well as they can, the question that comes to mind is: what does mendomi mean in the age of AI?

Michael Ballé: As far as I understand, mendomi is a Japanese term that refers to taking care of others, showing concern for their well-being, and providing guidance or support when needed. In Japanese workplaces, mendomi is often associated with experienced employees and managers who help junior colleagues learn, solve problems, and adapt to the organization. It reflects the importance of long-term relationships, mutual responsibility, and group harmony we’d expect from Japanese organizational culture. Although practices vary across companies, the idea of supporting and developing others remains a recognized aspect of management and teamwork in many Japanese organizations.

Westerners often have a rude awakening when working for a Japanese company in Japan, as taking care of you might not be what you expected. In different cultures, taking care of someone can be interpreted differently. In the US, for instance, it will mean providing a fair compensation and recognition, while in France, it means protecting your rights. There are vastly diverging worldviews here. I wasn’t particularly aware of “mendomi” as a thing until I encountered it on a Toyota shop floor. There, it meant making an effort to make work easier. I then found out that it had a more general sense of looking after employees beyond their work station, such as checking up on people and families when catastrophes strike, like COVID or the floods in Durban, mentoring programs in the plant, etc. In general, it’s an injunction for leaders to stay attentive to personal circumstances that might affect someone’s work.

Very concretely, you’ve standardized your line, but today’s operator is left-handed. Change the set-up. You’ve made sure all parts are within reach of the operator for easy placement, but today’s person is unusually short, or tall. Figure out a way to help them. Mendomi in practice, is how much of an effort you’re going to make to make that person’s job easier. I’ve seen so many managers noticing something off and then sitting on their hands and waiting for the problem to go away or the person to deal with it. Mendomi is about trying harder to make each person’s work smooth and seamless. Yes, we are looking to standardize the work. But more than that we are looking to achieve standardized work: adding value in the best possible way within takt time. According to the reality of the person doing the work, this means scratching our heads to adjust the work environment.

This is particularly important for engineers to understand. Once we’ve made an engineering change, in the product or in the process, the work is not finished until assembly is as painless as it was before the change. We can look at it as a unit, first engineering change, then engineering leadership of kaizen to iron out all the kinks the change will have induced. Mendomi: show you care, make an extra effort.

In the context of work, AI has two simple goals: 1) make each person more autonomous to turn them into super-employees, augmented workers with their AI agents that can do the work of three or four other people, so that, 2) you can reduce the number of people doing the work, which has been the ultimate goal of industrial societies all along: divide, separate, automate, reduce the number of people it takes to do the job.

In the broader sense, AI creates a new challenge for the Japanese idea of mendomi. If a company sees itself as responsible for taking care of its employees, then replacing them with AI is not only a business decision but also a moral one. It raises a difficult question: how can a company claim to take care of its people if it no longer employs them? The issue is not whether AI should be adopted, but how organizations can balance technological progress with their responsibility to support, retrain, or redeploy the people whose jobs are affected.

But at a direct level, AI is here to stay, so we can ask ourselves the real questions: How do I mentor someone to use AI to improve their job? What difficulties does AI create and how can we work on that? How can I make the extra effort to help people dealing with this new tool?

AI is typically an engineering change: when the dust settles, we will know better what it does and what it does not. So far, we can see it increases individual productivity but not necessarily overall project productivity, or certainly not as fast or by as much. Coordination issues are becoming overwhelming and getting AI to deliver value is still very tricky. We hear about companies firing staff, having quality problems, and having to re-hire to get it all under control. But we’ll learn. The mendomi point, to my mind, is more individual. AI changes nothing; it’s just another tool. Just like we can either walk past the person having to pick up boxes on the floor and stressing their back or stopping, recognizing, acknowledging and finding what to do about it (such as build with them a karakuri), the same is true with AI. We can ask ourselves what new struggles the tools will create and how we can help people with these difficulties.

People are already complaining about normal difficulties with every-day AI use. They report loss of control and agency. Because AI can complete tasks independently, users often do not fully understand how decisions were made or whether the results can be trusted. Instead of doing the work themselves, they spend more time supervising, verifying, and correcting the AI, creating a new cognitive burden as their attention shifts from performing tasks to constantly monitoring them. Many also worry that relying too heavily on AI will reduce their own skills and judgment over time, making them feel less capable and less connected to their work, and so on. So what can we do to fix that? What extra effort can we think of to alleviate these concerns? Let’s keep thinking.

RP: It's easy to dismiss the idea of mendomi, and indeed that of respect for people, as kumbaya-thinking, but the implications of "caring for people" (or of failing to do so) have big repercussion on them and on the organization. Can you explain why mendomi actually encompasses a much bigger idea than just "giving people standards to follow"?

MB: Research on trust generally identifies three core dimensions. First, competence refers to the belief that a person or system has the ability and expertise to perform a task effectively. Second, honesty or integrity reflects the expectation that it will act consistently, fairly, and according to accepted principles. Third, benevolence refers to the belief that it genuinely cares about the interests and well-being of others rather than acting only in its own self-interest. In other words, trust is not only about believing that a solution can do the job, but also believing that it will act honestly and take your interests into account. To simplify, we willingly follow people we find both competent (they understand the problem that needs to be solved, and know how to go about it) and trustworthy (they are honest and benevolent—towards us, at least).

We hate people we see as competent and untrustworthy because we feel they’ll credibly (competently) find a way to screw us over. We despise people who are neither competent nor trustworthy (how we feel about most politicians) and we pity trustworthy people that we think are incompetent. I’ve had the immense privilege of working with some people I trust completely (if something goes wrong, I’m definite it’s unintended). Work progresses in bursts: we all do our thing, then communicate quicky, discuss the plan, the event, have a laugh or a quick joke to release the tension, adjust, figure out the next step, and then go back to doing our thing. It’s a flow experience of getting things done, in a team. There is ongoing communication and correction, but it’s fluid and easy; and we achieve greater things than we thought we could do. It’s an amazing thing.

With people you don’t trust fully, either because you’re not sure of their competence or of their trustworthiness, immediately things are more heavy-going. First, you reduce your goals to what you think you can control. Then you’ll spend a lot more time in communication and making sure that the objective is defined, that the way forward is clearly laid out, and that the boundaries of what is acceptable and not acceptable are understood. We know these people well. They will show up, show up, and then unexplainably let you down with something you didn’t expect. Things get done, in the end, but with more caution, coordination and control (and they certainly aren’t as much fun). These are the people you need to check in on when you haven’t heard from them for a while, and go through goals and ground rules and incentives with them again, and so on. It works, but it’s slow and rather dull.

Then, of course, there are people you can’t trust, period. Either because they’re incompetent, and whatever they have to do, they’ll mess up. Or because they’re so out for number one that they’ll look and find the opportunity to use whatever is happening to further their own agenda and leave you hanging. Not all of these untrustworthy types are unpleasant. Many get away with it by being super nice and fun to be around, and interesting (they’re always in the middle of some kind of drama you get sucked in). We are a hyper-collaborative species, so we easily get taken in by untrustworthy people because we simply don’t have the cold logic to count the times they take advantage of you and never actually help. The difficulty is that your organization can impose you to work with them, or you like them and think that this time they’ll be a good friend, and so on. But whenever they’re around, things fail and blame is the game.

I really, really believe that trust is the secret ingredient that makes things run. Trust in yourself is the necessary ingredient to learn because learning means living with the frustration of not getting it right until you do, so you’ve got to keep going—think of how many people you know who use “I feel like an impostor” as an excuse not to do the hard work of learning something new. Trust in others because that is how coordination costs go down and how work becomes fun, to be fair. The working in bursts of get ahead – get together – get ahead – get together is such a profound and exhilarating experience!

Mendomi is the key to building trust in any relationship (or discovering there is no point in trying), in two ways. The first way is the “Coach Lasso” way: small, personal attention that show that you care about who this person is and what they say they like or dislike. A simple exercise is looking at what someone is wearing and wondering what item they want you to notice, and making a note of it, connecting with it. A further step would be looking at what someone is doing and finding a way to help, in a “pass the salt” way, when someone is struggling with too many things at once. This is very concrete and something we can all work on: don’t let the old lady struggle with her bags as she’s trying to climb the steps. For us Westerners, learning to see mendomi opportunities is a very real thing, and almost a spiritual exercise. It starts with saying thank you, and then figuring out of ways to be helpful, in the small things.

Caring about what happens to people is clearly the bigger thing (though not necessarily to them) and tricky because then your expectations of what should happen to them are often out of line with theirs. I am not proud of it, but I can’t count the time I’ve gotten into a heated argument with someone I care about because they were doing or missing something that I through was really important to their development. That’s… really tricky, but on the other hand surrounding yourself with people you have helped pass difficult times or go through blockages is how you create this trust environment where they will help you through your own struggles.

So no, absolutely not, mendomi is not a kumbaya thing. The belief that everyone can come together, get along, and solve problems simply through goodwill and cooperation is simply naïve and unrealistic. However, to succeed, we need people to come together, get along (somewhat) and get stuff done. It will not happen on its own—we need to work at it, and mendomi is a key to do so.

We spend an enormous amount of time and effort evaluating competence and remarkably little time evaluating trustworthiness. We hire people because they're smart, promote them because they deliver results, and build AI systems because they're increasingly capable. Competence alone has never been enough. Competence tells us whether someone can solve the problem. Trustworthiness tells us whether they'll solve it in a way that also works for us. And that second question is usually the one that determines whether things actually work. High-trust teams move faster not because they are more competent, but because they don't waste time protecting themselves from one another. They communicate less, coordinate less, and achieve more. As AI becomes more capable, this distinction only becomes more important. The real challenge is no longer building systems that can do things; it is building systems, organizations, and cultures that people believe are acting with them rather than merely for them. Mendomi as a personal practice and organizational norm is the key to this.

_THE AUTHOR