Tuesday, April 9, 2013

What is going on in the service industry?

When I think of service industry, I think certainly not one-size-fits-all. I do not think that the human factor is a small and subordinated to the whole. I do not think service industry even slightly resembles production from the 80s in the last millennium. Am I wrong? 

Many in the service industry, from healthcare to ICT, are busy embracing principles that were seen as a solution for production from the distant past. A time of high volumes, few changes, thinking that there is infinite capacity, labor marginal factor. Why?

6 Sigma and the service sector

Let me start with 6 Sigma. Allow me to remove the Lean portion. Lean I do think can add value to the service industry. This is different from 6 Sigma.

The basis of 6 Sigma is implied in the name itself: 6 Sigma. Sigma is a term from statistics. The goal is to really minimize the variation in the output. This is only possible by a high degree of standardization of both the product and the production process. Of course, this only works with high volumes of products that hardly change in specifications. And even then, only if the production process is organized and monitored very tightly.

6 Sigma is a single loop learning with but one purpose: to reduce the variation. The assumption is that this is good for the organization, customer, supplier, employee. And yes, in some manufacturing companies, this is indeed the case: large steel companies, (petro) chemical companies. Companies where 'process operators' are running around to monitor and adjust.

How many service companies will recognize themselves in the above? In my view, the core of service company is flexibility. The demand is rarely the same, the circumstances in which the service is to be delivered varies significantly. It depends highly on the human factor. The 'production' process is primarily about people.

Even companies that belong to the 'big production companies' have the experience what it means when 6 Sigma is also applied in other areas. For example 3M (Six Sigma 'killed' innovation in 3M). Still 6 Sigma insist on the fact that it can. Dave Snowden now refers to it as Sick Stigma.

Kanban and s/w development
I was told that 'Kanban' must be written with a capital K. This in contrast to 'kanban' which is used in production. Maybe I'm too stupid, but I cannot reconcile with kanban s/w development.

Kanban is a production control system that 'pulls' work. It is a very beautiful and elegant production control system. I remember the time I saw kanban in practice. The first thing everybody noticed was: some people sit idle doing nothing. The second thing that struck me was that everything was very standardized: workplace, the work, the products, etc.

The idle time of people, that was a tough one. But the director explained. They produced to stock, but the stock had a maximum (I think up till what they sold in 3 months). The demand fluctuated and thus the demand for capacity fluctuates. If the demand for capacity fluctuates then sometimes people sit idle. And for the 'flexible shell' lovers, the director was not crazy ... but not stupid either, he was above all pragmatic. Idle time was a price he was more than happy to pay for bottom-line it had given the best bang for the buck.

Anyway, idle time in s/w development? That would be a first. High degree of standardization, everything specified down to the smallest detail? I know quite a few s/w companies, but none of them come near.

It's probably just me. But a visual planning board with sticky notes on it, what Kanban seems to be, is not something that even resembles kanban. The only 'pull' that I see in Kanban is the removal of the post-its.

Do not get me wrong, visual planning systems with something as simple as cards or post-its works beautifully in many organizations with excellent results. Better than any s/w planning whatsoever. Even very large complex projects can be managed much better by sticking post-its on a big piece of paper.

Optimal Batch Size and s / w development

Apparently there are people walking around with the idea user stories or whatever can be seen as a batch that can or even must be optimized. There are even some games developed in order to make this clear. Batch sizing stems from the fact that set-ups consume a relatively large part of the capacity. Again, high volume, low variety environments.

A breakthrough came when Eli Goldratt realized that a production batch does not have to be the same as a transfer batch. Huge improvement was achieved in lead times, capital tight up, obsolescence, etc.

Beyond that, the beautiful picture on the right is in many cases not true. Holding cost often have a step function. Turnaround time is considered not that important. Flexibility is 'created' by keeping stock piling inventory. Sounds as far from the reality of s/w development.

Learning from other sectors is well
Adaptation from other sectors is good. There is much to learn with some surprisingly good results. It is important to know the background where it comes from. It is also important to remember where the challenges lie in your own environment. And, also try to get hold of the learning experiences of the area where the "new" idea comes from. There is usually a lot of dues paid and why would you want to do this as well?

No comments:

Post a Comment