Agenda

Review homework

Chapter 3 - 11, 36(c)

Case Study: How We

Slashed Response Time

Lecture/discussion

Chapter 4: Statistics

Metrics
Measurement
Statistics

Week 7

Week 8 assignment

Homework

Problems chapter 4
2, 3

Quality Metrics
Chapter Four
We best manage what we can
measure

Quality Metrics

Metric

A metric is a verifiable measure stated in

either quantitative or qualitative terms.

95 percent inventory accuracy

as evaluated by our customers, we are
providing above-average service

Metric
A metric is a verifiable measure
that

captures performance in terms of how

something is being done relative to a
standard,
allows and encourages comparison,
supports business strategy.

Customer quality measures

Customers typically relate quality to:

Feature based measures; have or have

not - determined by design

Performance measures - range of values conformance to design or ideal value

True versus substitute

performance measures

Customers - use true performance measures.

Producers - use substitute performance measures

example: a true measure of a car door may be easy to close.

true performance measures typically vary by each individual
customer.
Unfortunately, producers cannot measure performance as each
individual customer does.
these measures are quantifiable (measurable units).
Substitute measure for a car door: door closing effort (footpounds).

Other example: light bulb

true performance measure -- brightens the room

substitute performance measure wattage or lumens

Educating Consumers

Sometimes, producers educate consumers on

their substitute performance measures.
What are substitute performance measures for
the following customer desires:

Good Gas Mileage

Powerful Computer

What is the effect of educating consumers on

performance measures?

What is a metric?

Another term for a substitute performance

measure is a metric.

Metric is a standard of measurement.

In quality management, we use metrics to translate
customer needs into producer performance
measures.

Internal quality metrics

scrap and rework

process capability (Cp or Cpk)
first time through quality (FTTQ)

Identifying effective metrics

Effective metrics satisfy the following

conditions:

performance is clearly defined in a measurable

entity (quantifiable).
a capable system exists to measure the entity
(e.g., a gage).

Effective metrics allow for actionable

responses if the performance is unacceptable.

There is little value in a metric which identifies

nonperformance if nothing can or will be done to
remedy it.
Example: Is net sales a good metric to measure the
performance of a manufacturing department?

Use of quality metrics

Quality metric data may be used to:

spot trends in performance.

compare alternatives.
predict performance.

However, organizations should consider the

costs and benefits of collecting information for
a particular quality metric.

collecting data will not necessarily result in higher

performance levels.
higher quality companies often use fewer metrics
than their competitors.

Acceptable ranges

In practice, identifying effective metrics is often

difficult.

Main reason: non-performance of a metric does not always lead

to customer dissatisfaction.

Consider the car door example again, if door closing

effort is the metric, will a customer be dissatisfied if
the actual effort is 50 foot-pounds versus 55 footpounds.

Producers typically identify ranges of acceptable

performance for a metric.

(a) For services, ranges often referred to as break points.

(b) In manufacturing, these ranges are known as targets,
tolerances, or specifications.

Break points

Break points are levels where improved performance

will likely change customer behavior.
Example: waiting in line

Suppose the average customer will only wait for 5 minutes

Wait longer than 5 minutes -- customer is

dissatisfied.
1-5 minutes -- customer is satisfied.
less than 1 minute -- customer is extremely satisfied

Should a company try to reduce average wait time

from 4 to 2 minutes.?

Targets, tolerances and

specifications

Target (nominal) - desired value of a

characteristic.
A tolerance specifies an allowable deviation
from a target value where a characteristic is
still acceptable.

Lower specification
limit (LSL)

Upper specification
limit (USL)

TARGET
-1

+1