A letter on good and bad Kaizen practices

I recently visited a contract electronics manufacturer with a striking capacity for kaizen – the steady improvement of every step along its key value streams. Dozens of kaizen events were being performed across the company to eliminate wasted steps and to remedy quality, availability, adequacy and flexibility problems in each value stream. At the same time, kaizen teams were trying to speed continuous flow and to perfect pull systems when flow was not possible.

The managers were pleased with their work and I had to admire both their technical skills and their enthusiasm for rapid improvement involving the employees touching each value stream. However, I noted that most of the value streams being improved were for products that had been launched recently. I wondered why so much kaizen was necessary.

Indeed, I pondered – as I often do these days – whether the kaizen effort was analogous to old-fashioned end-of-the line quality inspection in mass production organizations. Value streams for new products were being put in place without much thought to lean principles or much rigor in thinking through the details of every step and action. Kaizen teams were then inspecting the processes once in operation, finding them far from lean, and launching waves of corrective action.

Given that many bad practices had been built into the value streams, these kaizen efforts were necessary and highly productive. But why wasn’t the organization performing lean process design as an integral part of the development process? And was the organization’s skill in after-the-fact kaizen – that is, its talent for process rework – actually reducing the pressure for the hard conversations about lean process development that ought to be taking place during product development instead?

As I’ve reflected on this situation, I’ve wondered if the practices of Toyota and other lean pioneers have been misunderstood. Kaizen is an important activity at Toyota and involves all employees. But new processes launched at Toyota are usually extraordinarily lean to begin with and post-launch kaizen is a small part of Toyota’s competitive advantage.

The secret lies in Toyota’s product/process development system that focuses on creating “profitable operational value streams” – to use a favorite phrase of the late Allen Ward. These streams have been thoroughly “pre-kaizened” by examining every step in the proposed production and fulfillment process long before launch.

The first step is to make sure someone is responsible for thinking about the whole process needed to bring a new product from order to delivery. By thinking about the production process at the same time the product design is being evaluated, it’s possible to optimize both.

The second step is to lay out the process on paper and consider the different ways that it might be conducted. For new types of products requiring new processes it is particularly important to consider a number of different ways the whole process and each step might be conducted and to conduct simple experiments to see which way works best. (This is the process development analogue of the Set-Based Concurrent Engineering methods used to evaluate different approaches to the design of the product. It’s also a key element in the 3P Production Preparation Process now conducted by advanced lean organizations.)

The third step is to test any new ways of conducting process steps with simple prototypes – even cardboard mock-ups – to learn how well they actually work (another element of 3P). The knowledge gained from these experiments then needs to be written down and turned into the experience curves of the sort Toyota develops from experiments with simple prototypes of new products.

(In fact, this knowledge is Toyota’s great advantage in concurrently and rapidly developing new products and processes: At this point, most of Toyota’s production processes are highly standardized and fully documented. Most new product designs only need to comply with well-understood process requirements to launch as smoothing flowing streams. By contrast, most organizations I visit have poorly documented processes with weak standards and little real knowledge of trade-offs in designing a process one way versus another. They will need a lean leap in consciousness and practice in order to catch up.)

Once the best process is determined, which may result in changes to the product design as well, the next step is to finalize equipment designs and information management systems.

Finally, it’s time to develop standard work for every step in the value stream and standard management for the whole value stream. This includes a training plan for every employee, a plan for every part and a maintenance plan for every piece of equipment.

If all of these actions have been completed by the start of production, the value stream should be very lean from the first item delivered. Kaizen will still be important, based on hansei (or reflection) about the performance of the process once operating, but it can start from a higher level in a more stable process so that additional rapid improvement is actually easier.

It’s my feeling that many organizations are now ready to elevate their level of play. As I hope I’ve made clear, this is not by de-emphasizing the idea of kaizen, but by performing the PDCA process that is at the heart of kaizen inside the development process. This will insure that every new value stream for every new product commences its productive life as a very lean stream.

Given the steady reduction in the length of product lives, I believe that it will become ever more important to achieve “process quality at the source”. Otherwise, the product may be ready to go out of production before process problems are ever addressed through kaizen as rework.