Ray Morgan
is a remarkable person. The teams he has lead have created amazing examples of
engineering excellence and creativity, including the solar plane which holds
the altitude record to
this day. I once asked him how he did it. He replied simply “I made my company a
massively parallel computer.”
It took me a while, but I finally got his point.
Think of your organization like a computer. Like a
computer there are inputs, outputs, translators, memory and processors. Most
organizations are set up to ensure that one or two key individuals are the
processors: they take the information from the various input devices (their
advisors), call upon memory (in the form of regulations, financial information,
etc.) and then process the information into a decision (output) that they then
expect executed by the rest of the organization.
In a small organization this “computing architecture” may
be optimal. In fact, you see the “one main processor” design in many
early-stage start-ups. But the larger your organization becomes the more
inefficient this design becomes. Think of your computer at home: the more data
you put on the disk, the more programs you have open, the greater the load you
put on your processor. This causes your computer to slow down and eventually
crash. It’s the same thing for an organization: put any one processor in the
critical path, no matter how advanced, and your computer eventually stops
working.
Ray went on to explain that he felt his job
as a leader was to help everyone possible who worked with him to become better processors. As long as there was some ordering mechanism for outputs (i.e.
more processors didn’t just create more confusion) having as many processors as
possible helped his organization be more creative, solve problems faster and
reduce turnover. No matter how big it got and no matter how complex the tasks
before them, his division seemed to be able to work through it.
What Ray didn’t say (probably due to his
inherent humility) was that he is a world-class system architect. He didn’t
spend his time being the biggest, the fastest or the best processor in the
organization. He understood that as one person he could never be as efficient or effective as
many people aligned around working a problem. Instead, he spent his
time setting up information channels, processes and systems to ensure that each
team member (processor) could add as much value as possible to any problem they tackled.
Being a system architect is probably the most difficult
work that an individual or team can undertake. This explains why so few managers do it well. As individualists and capitalists we are raised to be the
best possible egocentric mega-processors rote education can produce. We are told that if you want to be a
leader you have to be assertive and controlling. This means that
when things go wrong most managers feel the undeniable urge to take control of
the situation. And since there is always something going wrong, it is easy to feel that we are delivering the most value for the organization when we are the biggest and best processor of risk and
cost information, parsing information and making decisions
that send other people marching to put out the myriad fires.
This is most likely why most organizations lack for a coherent
approach to resource allocation, process design and business model realization. This is also a key reason that people don’t usually reach the top when they play the role of the
thoughtful background players. We are not programmed for architecture,
especially when processing is so much more viscerally and socially appealing.
And yet it is lack of architectural competence that is the single most important reason for business disruption. It is not
the failure to make grand strategic pronouncements that imperil us, nor is it a
lack of people to take the lead and “make it happen.” It is the translators in
the middle, the people who look at where the organization wants to go and what
it has to work with and create the systems and processes and culture that
enable that vision to become reality.
If HR wants to be a “strategic player” it needs to be
doing two things well (aside from the day-to-day transactions): making as many people
in the organization into processors and creating a system architecture that can
harness the increase in organizational power that follows.
Thanks for the lesson Ray (the best father-in-law a guy could hope for).
In a small organization this “computing architecture” may be optimal. In fact, you see the “one main processor” design in many early-stage start-ups.
Posted by: silagra | February 08, 2010 at 02:16 PM