Dear Gemba Coach,
What is the worst mistake you’ve made on the gemba?
Yikes. Worst mistake? Tough question. I guess the obvious answer is being too tough at times and not enough at others. Getting people to change is difficult because in order to really change, they’ve got to go through a period of discomfort as they try to adopt new behaviors and abandon old ones. Sometimes, you’ve got to push them hard into the pool – others you’ve got to stop and listen and hold hands. There’s never a good balance, or at least, I have never found it. On the one hand what you tolerate will continue, on the other hand when some people get really pissed off they can hold long grudges which come back and bite you. But I suppose this doesn’t count as a mistake as such, more something I’m not very good at and a constant topic for kaizen – the balance between strong mind and warm heart.
Actual mistakes that have slowed the pace of progress are harder to spot – it’s difficult to see yourself work – but, upon reflection, I can guess at two mistakes I’ve repeatedly made: one technical, one political.
My greatest technical mistake is not spending enough time on the product/process matrix. In most situations I’ve seen, from production to service (it’s actually much worse in service), the product/process balance is deeply built-into the organizational structure. When I first started lean, sensei were adamant on flow when you can and pull when you can’t. That meant we created cells wherever the cycle time was not too different from takt, and then dealt with capacity equipment when cycle time was at least an order of magnitude faster than takt.
Our main tools to get started were the spaghetti chart and the flow chart. Over time these were replaced with value-stream mapping, which shows another vision of the plant but tends to hide the diversity of products. When you strive to create continuous flow cells, you actually have to go through the hardship of clarifying the product/process matrix. There are four keys to understanding mura in the process:
- Complex flows
- Large batch sizes
- Not enough will to adjust production rhythm to sales rhythm
- Poor supply logistics
By tackling the complexity of flows and stabilizing which product is made on which equipment we usually find a gemba key to address the other issues. On the other hand, without tackling this upfront, even if we make progress on the other subject, the people on the gemba can’t find their autonomy in handling products and so the instability remains as spot decisions hinder as much as they help.
In many high-mix low-volume environments it’s easy not to devote enough attention in stabilizing product groups, and in service environments in many cases the question is not asked at all. Working on the product /process matrix long enough to sort it out is key – but hard, and as I look back I’ve definitely missed a trick or two there, at a much larger costs in terms of performance improvement overall. In terms of inventory turns, the larger benefits come from component inventory reduction through milk runs and so on, but this, in turn requires a detailed understanding of fraction-mix decisions at the production scheduling stage.
Don’t Ignore Naysayers
My worst political mistake is rather different in nature, but probably more costly in the long run. In any lean initiative, at the management level there are those who help and those who hinder – it tends to be as simple as that. Help with what? On the gemba, getting closer to monozukuri through hitozukuri:
- Solving product/production problems for specific customers
- Through developing people’s skills in an atmosphere of mutual trust.
You get there through kaizen, but kaizen requires the proper questions asked, which involves quality, just-in-time and jidoka. In organizations, this bothers many people who have far different agendas. Typically, the purchasing manager is largely evaluated on the price reduction he or she gets from the supply base, and arguing that suppliers should deliver good parts on time makes his or her job harder, not easier. In fact, if the purchasing manager takes on a cheaper supplier with lower quality, they get the bonus and the plant manager gets the hassle. I’m not saying all purchasing managers think that way, and indeed, I’ve worked with some excellent ones who “get” the quality and lead-time issues. Al I’m saying is that you’ll always find some people in the organization who simply don’t see it the lean way because it doesn’t suit either their goals or their style.
The mistake is ignoring them. When lean works, first it delivers results in terms of cash and performance and second it keeps you rather busy. The tendency is to focus on the people who are willing to teamwork and kaizen and ignore the rest. This works great because by creating a team of people committed to improvement, you get spectacular results, but it also fails because the others can gang up on you – I’ve seen this happen time and time again, and, at certain points in decision-making have enough influence to kill the improvement effort. To be fair, their argument is that their way to improve is more effective and we should stop this people-coaching stuff and do more systems-implementation stuff.
The hard point is that it’s very hard to get teamwork from someone not doing kaizen on their own, because every weakness you’ll expose will be used against you. I’ve just lived through this as the commercial director of a company refused to see the problems he was creating by inflated pricing and crazy 80% discounts through promotions and such, which completely screwed up the demand signal, and kept arguing that what destroyed the OTD was none other than the lean effort. Ouch. Every time we tried to get this guy onboard, he used the information shared about what production did not know how to do (and needed sales help) against the COO. Not fun.
The temptation then is to let well alone and stop talking. I’ll get on with my stuff, you’ll get on with yours. This rarely ends well, because at some point some real crisis will come up, and if the next level up in the food chain doesn’t deeply understand lean and realize what is going on, the naysayer can sometimes win the day through sheer Machiavellian politics. We don’t like to consider this, but some people, particularly at corporate do make a career of blame politics, and lean guys are simply not equipped to deal with that as their attitude is to be open about one’s problems. The tough challenge thus is to keep talking to theses guys, no matter how irritating, aggravating and occasionally dangerous it is. Stopping communication gives you a respite for a while, but I’ve often found it kills you in the end.
Continuous Challenges
Product/process understanding on the technical front and not dealing well with adversaries on the political front would definitely be on the top of my personal list of mistakes. In the end it comes down to pretty much the same thing. By the very nature of continuous improvement, lean continuously challenges the status quo, and sooner or later is going to step on the toes of someone of the big-system-solution persuasion rather than kaizen with all people all the time. Results help, of course, but it’s surprising to see how little results count as an argument in a purely political food fight. I’ve seen very senior executives throw away millions in cash (literally, millions) to implement their favored form of centralized supply chain, centralized IT system, financial control, purchasing strategy and so on. Their arguments are always the same: no pain no gain, we’ll get the returns later on and better a tough solution now than this endless small-step continuous challenge nonsense.
The conundrum remains: we need results now to be convincing, but this means pushing people hard and making enemies. We need to work with people and connect to their interests and let them find their own footing, but this slows down getting results and opens flanks to the real enemies (which we often don’t see clearly because we’re not talking to them, but they talk to the CEO). No easy answers, but that certainly keeps the job interesting!
