Principle 4 - To improve the outcome, improve the system (Item 8)
In order to improve the outcome; improve the system and its associated
processes.
Corollary: All people work in a system: outcomes are improved
when people work on improving the system
One way of reducing variation in processes so that their output is
consistent is to document them. Then people working in the process
can follow the documented procedures. This means that everyone in
the process does not have to discover and invent it for his or herself
every time. This lets you be consistent in your output and reduces
variation in what you deliver. Several international standards give
good guidance with the process of standardizing your processes.
Standards based control schemes use documentation
to clearly describe all the process and product/service control criteria
required for maintaining stable and capable performance of the business
process at each point. Process maps, process descriptions and checklists
are all useful in documenting processes. They all have different uses.
By following the documentation, people can be following a previously
identified `best' or standardize' practice.
Most companies that document their processes
find enormous value in just carrying out the documentation. The documentation
process causes discussion about the work-process and uncovering of
its problems, difficulties and inconsistencies.
As we described previously, the people carrying
out the tasks and using the schemes have expert local knowledge about
how the process actually works and the nature of problems in the process.
Their involvement in both design and improvement is a basic requirement
for being able to use the control scheme as intended. If the people
who should use them did not design them or do not fully agree with
them, or if they think they are impractical, they will not use them.
It is usually a responsibility for operational
management to work with their people to design and improve local control
schemes.
We see four types of failure with the use of
control schemes. They have many things in common.
- The company insists that the control process
be followed always and will not allow it to be modified or improved
— on pain of dismissal. This makes for very rigid processes
that resist improvement even if they need it. The output of the
process may be constant, but it may be junk. Not a desirable outcome.
- People give only lip service to the control
process because they know it leads to nonsense results. This occurs
most often when the process workers are excluded from the design
of the control scheme, and when process workers are not allowed
to modify the formal process. In these cases, if they can or if
the process is important enough, the process workers usually abandon
the documented process and use one that really works. This has the
undesirable effect that there is in fact no standard process and
that process variability is not controlled.
- "Our processes are extremely good.
If only our people would follow them". This occurs because
of the failure to recognize that people are part of the system and
contribute to system variation through their interactions within
the control schemes. If the process descriptions really are
good and are practical, then people would use them. If employees
will not use them, there is probably something wrong with the processes,
not the people. The old thinking is that everything would perfect
if it were not for those `useless employees'. The `if only' syndrome.
- The processes have become so documented
that they are boring. People do not enjoy working with them. There
is no variety. (This is discussed again in Principle 8 `Innovation'.)
Checklists are often of the most practical
value for people doing routine work or precision work carried out
infrequently but for which the omission of small details can be critical.
Checklists are widely used by military to ensure control of processes.
Check lists are an important
control. Many businesses find that providing staff with checklists
and process maps is an important step to providing consistent responses.
Military organizations have found checklists particularly useful in
tension filled situations that have been thought through in advance
but which the individual on duty at that time may not have experienced.
Traditional quality control or quality assurance
designs are strongly focused on `after the event' methods such as
inspection, testing, sampling and sorting. Because of their design,
these often give information too late to be of much use in preventing
costly failures. You need proactive or `before the event' controls.
There are three types of control design:
- Feedback
- Feedforward
- Cognitive
Feedback is a reactive `after the event'
control. For example, "the clothes are wet, it must have
rained". It relies on some form of output – often variation
outside of specification — to trigger a reaction. By then, it
is all over. However, you can stop making the same mistake again and
(if manufacturing), you can prevent the faulty part going to a bottleneck
(see later).
Feedback from the problem back to the cause
is usually slow. Meanwhile, problems continue and get worse. The cost
of running the process in between the problem occurring and you finding
it can be very high — a combination of process and product costs.
In addition, some faulty products get through your screening and become
complaints. The cost increases.
However, inspection can stop you making the
situation worse. Goldratt points out the cost if you put rubbish through
a bottleneck.
"We rely on final inspection to sort
out the rejects from the prime product. There's a delay of about two
hours between the production operation and the inspect process. We
make about 4000 widgets per hour; these are valued at $100 each. Last
night, we had a deep scratch on the product, rendering it useless.
Cost us 8000 units before it was discovered and corrected. It was
at a bottleneck resource. $800,000 down the drain!!!"
Inspection does not guarantee
no errors – it is too late. They have already happened. You have
already made the rubbish.
Feedforward is a proactive `during the
event' control that is based on an input measure. For example, `the
clothes are wet because it is raining now'. It requires a sound diagnostic
knowledge of the process to quantitatively `model' its action triggers.
Cognitive is a predictive `before the
event' control that is based on some form of intelligent prediction.
For example, "clouds brewing and winds are getting stronger,
it will rain soon and wet the clothes". It is complementary to
feedforward. You need sound diagnosis combined with broad knowledge
and input from all your senses to provide flexible and adaptive prediction.
There are many parallels between weather and
business eg winter is coming, brass monkey will shrink, days get shorter,
soup will be in demand etc etc. Even squirrels know that there are
lead indicators. Another example of a lead indicator is the Internet.
It is changing the way business is done, how information is sent and
consumed. We know it will cause change but how, where, what and why.
Modern approaches to control require an in-depth
understanding of the complex interactions, process flows and interdependencies
within your business system and how they determine your outputs. Such
sophisticated understanding is the basis for feedforward and cognitive
control designs.
In terms of effectiveness, the preferred sequence
for control design is:
Cognitive is better than Feedforward is better than Feedback
This does not mean that feedback controls are
not required. It does mean that predictive and preventative control
design is more effective than reactive control design. Squirrels die
if they do not prepare for winter.
Service present an even worse case. Services
are not amenable to feedback control design. Customer complaints and
surveys are useless for day-to-day control. Poor service simply passes
through and out of the system. You end up with annoyed customers.
For every one that does complain, 10 are equally annoyed but cannot
be bothered to tell you.
There are few quality assurance nets for services
and usually the service part of the business has the biggest impact
on customers. Poor service costs you because annoyed customers do
not come back. This is bad news because it costs more to get new customers
than it does to retain your existing customers – which is why
organizations put effort into building customer loyalty.
Many people think they have `done quality' when they get their five
ticks from ISO. As you can see, the Business Excellence Principles
are very different from ISO checklists. There is a strong role for
a Quality Assurance process.
- Quality Assurance (QA) lets you standardize the way you produce
so that what you produce is consistent.
- QA gives you the chock that stops your processes falling back
into the chaos they once were – back to the old way of doing
things. Once you have spent the time and effort to improve your
processes, you want people to use them.
In the past, ISO has concentrated mainly on the standardization phpect
of Principle 6 (`Variability'). Because ISO is about `standards'
and standardization', it will always retain that emphasis. Currently,
ISO is trying to move to include more of the Business Excellence Principles.
ISO 9004 is an attempt to recognize the importance of the Business
Excellence Principles in all organizations.
Your documented standards must not restrict your agility. If your
documentation entrenches the way you do things to the point where
you cannot respond to changes and new situations, your business can
become stuck.
Agility has become increasingly important. Agility
is the ability to adapt quickly and effectively to opportunities, changes
in operating environment and changing stakeholder needs. Agility might
mean rapid changeover from one product to another, rapid product development,
rapid response to changing demands or the ability to produce a wide
range of services. Agility might demand special strategies such as modular
designs, sharing components, sharing manufacturing lines or providing
specialized training. Agility increasingly involves outsourcing, alliance
agreements with important suppliers, developing new competencies and
novel partnering arrangements. Agility requires overcoming inertia and
maintaining a balance between responding to stimuli and keeping your
alignment and constancy of purpose. Documentation can be a solid form
of inertia.
|