Target Setting*

[David Kerridge <dfkerridge@mac.com>]

We must bear in mind that the objection is not to targets as such, but
to *arbitrary numerical* targets.

"We must get this roof repaired before the leak causes more damage" is
not arbitrary, but a recognition of reality.

"We must cut costs by 10% or we will be out of business" is (if true)
just a recognition of the facts of life.

In both cases the target supports the overall aim of the system. For
example, if we go out of business we cannot achieve the aim.


But when we set a target, however good the reasons, we must recognise
that if the system does not meet it already, we can only meet it by
changing the system. That usually means taking resources from other
activities. 

This may be the right thing to do, but we should never do it without
being aware of the consequences. Presumably the reason that the system
could not meet the objective without change is that the other activities
are normally more important. So the system will have to be changed back.

Equally we should keep careful note of how often such emergency action
is required. If such response becomes a part of the system, rather than
a response to a special cause, something is badly wrong. Short-term
thinking can easily become addictive. And a change to the system,
however minor and temporary, introduces chaos with invisible and
unpredictable costs.

David Kerridge