Principle 6: Variability (Item 1)
All systems and processes exhibit variability, which impacts on
predictability and performance.
Variation means lack of consistency. Lack of consistency means additional
cost and rework. Rework is a total waste of your resources. In most
companies, rework is the largest consumer of managers' and employees'
time. Reducing your variation usually means reducing your costs. People
who understand variation are more likely to understand the concepts
of Process Capability that we explored in Principle 4 (`To Improve
the Outcome, Improve the System'). Consistency means that your products
will more often do what they are supposed to and your customers will
be more satisfied.
Managers who lack an understanding of variation
will probably not practice according to a systems view of the company.
This results in consistent under- or over-reaction. Product and service
quality (and morale) often deteriorates as a result.
Research indicates that Principle 6 has the second strongest correlation
with KPI improvement (after Principle 10 `Value for All Stakeholders').
Unfortunately, Principle 6 is also equal last (together with Principle
5 `Improved Decisions') in how well companies are at implementing
it. Assume for a moment that it is a strong driver of success. If this
is so, there is a lot of unused potential. If companies were to do better
at implementing Principle 6, they would achieve significantly better
business results – i.e. they would be more successful.
Reducing variation can give considerable competitive advantage. Reducing
variation gives more stable products and services (and hence more satisfied
customers) and less wasted effort on internal furfies.
Variation affects us in five ways.
- It can lead us to react inappropriately to random events –
this uses our time and energy unnecessarily and we can make the situation
worse by our well-meaning interference
- Errors (variation) accumulate from one step to the next in our processes
- In fact, there is a funnel effect in complex process with errors
from many different sources accumulating. This means that the final
product can (and will) have errors from many different sources.
- Variation means rework and rework is a waste of time, energy and
resources. Having less rework means less employee expense and that
you can move on to things that are more productive. You can increase
your throughput.
- You can reduce your rework throughout your process by constantly
working to achieve smaller and smaller variation. This also means
you can provide products and services that are of more value to your
customers, because your products will more often do what they are
supposed to and your customers will have less rework to do to be able
to use them. The result is increases customer satisfaction.
If you think any of these, you don't
understand variation
- "We can learn a lot by studying
the cause of each thing gone wrong"
- "Whenever things get better
or worse, we can learn why"
- "Each problem or mistake has
a human cause"
|
Everyone intuitively understands variation and variability. We all
know that the weather is not exactly the same every day, or that
when we commute to work it does not take exactly the same time
every day, or that no two people are exactly the same, or that
no two concert performances are exactly the same (even by the
same group).
Maybe you think we are exaggerating the "exactly".
That is the point. We intuitively know that no two things or events
are exactly the same. They vary.
Variation is something normal that you see every day. You see it as
a range of results around an average. These examples are adapted from
training material prepared and used by the Australian Quality Council.
The waiting time at the counter averaged 10 minutes; the shortest
was 3 minutes and the longest 25 minutes.
The average height of students in this class is 1535 mm. The tallest
child is 1610 mm and the shortest is 1390 mm.
Sometimes we may be only interested in the range.
Temperatures vary over a 24 hour period and the weather report usually
gives the range by quoting the highest and lowest temperatures: "Today's
high was 25C and the overnight low was 12C".
Outside of work, most people see variation as the norm and people are
tolerant of expected variations. Our friends are different ages, shapes,
heights, skin colors; daily temperatures go up and down; we don't expect
the forecasters to get it exactly right; the newspaper is delivered
around dawn; our train is often a few minutes (or seconds) late; we
finish work at different times every day.
We expect that things will vary over an implicit range and we feel
comfortable about that. However, we do get concerned when we think or
feel that the variation is abnormal:
Over the past few months, my train has been more than five minutes
late twice. We were then told that it had been canceled. Paula is doing
well at school and her reports have been satisfactory. However, with
her latest report, we have been asked to come to the school to have
a discussion with her teacher. My car has been going well since I bought
it. But, just in the last week, it has been difficult to start so I
may be heading for some problems.
We innately sense, from the accumulation of our life experiences, that
events like these are in some way special. We are able intuitively to
separate them from the common, normal events.
Knowledge about variation can lead
to
- Understanding the causes of variation
and how to respond appropriately
- Understanding tampering: how leaders,
despite good intentions, can increase variation
|
In fact, variation in the workplace is no different to that which is
routinely encountered and coped with in the community.
The difference occurs in the workplace when people in positions of
authority in management do not understand the basics about variation
and deliberately intervene in normal variation situations. It is like
demanding a report on why the weather is different today.
Unfortunately, many of the management models currently being promoted
by consultants and used by companies do not take variation (or its accumulating
effect) into account. These deterministic models are often built by
accountants or engineers who may see the world in such a linear fashion.
They fail to understand that the output of a process is not the
sum of the average outputs from the parts of the process.
In deterministic models, if machines (or people) A, B, C and D (each
dependent on the one before) can produce 10 parts per hour on average,
then the output of the process should be 40 parts per hour. All the
accountant need do is measure the `variance' from that 40. In a deterministic
world, any deviation from 40 must be due to slackness.
As we shall see, the output from our four-step process depends on the
variation of each step, and interdependency between steps, as well as
their average output.
Deterministic models can be useful in helping to understand the links
in the process. However, to be useful for prediction, they need to be
made into statistical models by including estimates of variation.
|