dancing martial arts masters of the bicycle factory

when the system turns on its finest components

A year or so ago, I wrote a book with Will Butler-Adams, the CEO of Brompton Bicycles.  One of my favourite bits of it, which has stuck in my mind ever since, was introductory to a chapter about one of the hardest things he ever did – reorganising the factory onto a production line system.

Leaving most of the details to the book, I’ll summarise it as follows.  Assembling a folding bicycle requires somewhat more than a thousand components to be put together.  A lot of this is “pre-assembly” (screwing together the different bits of a pedal or a handlebar, and so on), which can be done in bulk.  The final assembly, on the other hand, is trickier. Brompton bikes are mainly built to order.  There might be five options at each of twenty stages of assembly, and five to the power twenty is more than nine trillion.  So assembly was a skilled job, done by specialist bike builders.

The assemblers were paid on piecework, and this meant (as piecework schemes usually do) that some of them got very good indeed at what they did.  The standard time taken to assemble a bike was about an hour, meaning that the rated output was seven or eight bicycles per assembler per shift.  But some of them could do thirty.  According to people I interviewed at the factory, it was like watching a dancer, the speed they moved at, always picking the right component first time and moving tools from one hand to the other.

However, even if everyone could move like that (and very few could), it still wasn’t a system which would be able to handle the kind of growth that Brompton’s order book was seeing. (At this point in the story, waiting lists had got out of control, a regular theme in the company history).  So, in order to expand, Brompton had to reorganise entirely; to a production line system, in which every stage of assembly was carried out one at a time at a different station, with part-completed bikes moving around the factory floor. 

This meant that you could hire many more workers, and you could match employment and output to seasonal demand – a new employee can’t master nine trillion combinations, but they can handle five or ten, which allows them to work one or two stations.

Unfortunately, in a production line system, someone who works really fast is just piling up half-completed bicycles in front of the next station.  They are still a good employee, who can work wherever they’re needed, but there just isn’t any way for someone in a production line to stand out by a factor of five.  You don’t have piecework in an assembly line.  The dancing martial arts masters’ human capital was gone – their specialised skills were adapted to a different system.

This was hugely culturally disruptive, and the complete story is in the book.  Brompton tried to smooth the financial hit to the top-earning assemblers, but there was really no way round it.  And of course, previously to the reorganisation, they had been the elite of the elite – the culture carriers, the people who everyone else looked up to.

I’ve noted on a few occasions in the recent past of this blog (and will continue to do so!) that one of the key criteria for a viable system is the ability to restructure itself to cope with new challenges and conditions which were not anticipated at the time it was first designed.  But it’s worth emphasising that “reorganisation” is not by any means cost free.  It has a very significant cost in terms of the identity and organisational coherence of the system; during the transition from piecework to the assembly line, a company can end up devoting extraordinary amounts of its energy into internal in-fighting and away from productive activity. 

A key function of the highest level of management in Stafford Beer’s “viable system” model is the balancing of the variability of the environment with the capacity of the operations to withstand change.