Friday 21 February 2014

Why Lean is so inefficient

It's no good. I simply can't resist it. The old monsters at management consultants McKinsey & Co are just too tempting a target, and not just because they are so responsible for the shape of the modern world and its organisations - but because they are so slow to grasp why those organisations are stuck.

I subscribe to their online newsletter and it is always well-written and informative, and sometimes so wonderfully old-fashioned that I find it irresistable.

The elite consultancy, named after its founder James Oscar McKinsey, the first management consultant, still lives and dies by the highly misleading maxim “everything can be measured and what can be measured can managed”.   And therein lies the problem.

What actually inspired me to write this post was their new hymn to the idea of Lean, which as anyone who works in big organisations knows was attributed to Taichi Ohno at Toyota, and is now being rolled out everywhere from the NHS to the way car insurance works.

This is what they predict:

"An unprecedented amount of product-performance data is now available through machine telematics. These small data sensors monitor installed equipment in the field and give companies insights into how and where products are used, how they perform, the conditions they experience, and how and why they break down. A number of aerospace and industrial-equipment companies are starting to tap into this information. They are learning—directly from customer experience with their products—about issues such as the reliability of giant marine engines and mining equipment or the fuel efficiency of highway trucks in different types of weather."

Now, bear with me a minute.  Although Lean emerged from assembly lines, Ohno was not actually very impressed when he visited Ford immediately after the Second World War, as I describe in my book The Human Element.

There was too much extra work from bad workmanship, and there was far too much extra stock lying around taking up space. On the same visit, he and his team happened to shop in a supermarket called Piggly-Wiggly and then they got excited. The supermarket had a system that only ordered enough stock as it was bought by customers. 

It was the beginnings of the Just-In-Time delivery systems that so many companies use today, and the basis of Toyota’s success.

Toyota’s ‘lean’ systems are all the rage in the UK now, and it makes a lot of sense.  It tries to bring jobs back together again rather than splitting them all up into little bits. 

The problem with Lean outside a factory is that it keeps the Toyota version of the assembly line at its heart. It requires standardised systems and processes. It means measuring the speed that everyone does their jobs and holding them to the new system. 

 It is therefore a prime candidate for putting the processes into inflexible concrete, and Lean IT is the latest twist in what is on offer from IT consultants. 

The trouble is that Lean assumes everywhere is a factory, a little like Toyota. The central purpose may be to improve the flow by bundling jobs back together again, but it isn’t necessarily about getting people back face-to face – and it certainly doesn’t mean letting staff get on with their jobs in the way they know best. Quite the reverse.

It is the obsession with data that is the clue. Of course these tweaks are important, and they do improve services. I know government functions that have been transformed just by moving the photocopier a bit nearer the front desk.

But here is the problem.  Lean assumes that the basic system will always stay the same.  It wires in data sensors and processes, and software systems, that make sure it does.  It allows no leaps of imagination - no sudden realisations that, actually, what you need is a different kind of institution altogether.

This is McKinsey's basic weakness, and it is a big one - as one of their biggest critics, the systems thinker John Seddon has explained.  This kind of Lean approach tends to lock organisations into their own administrative and control systems, even when they are doing the wrong thing, desperately improving the efficiency of it - and wondering why no major leaps forward are possible.

This is the problem with targets too and with the New Labour approach to services, which still clings on to public services in so many ways.

It allows you to improve the throughput of the NHS, as if it was an assembly line.  It allows to to improve the efficiency of the existing system until it can improve no more.

What it can't do is tell you, given that you want to improve the broad health of the population, whether you might need a different kind of organisation altogether.

Subscribe to this blog on email; send me a message with the word subscribe to When you want to stop, you can email me the word unsubscribe.

No comments: