|
July 1998 Volume 2.10
The Right Customer Value Road
It seems simple doesn't it. The way to work out what customer's
value the most would be to ask them. Unfortunately when you ask
this type of question people tend to tell you what is important
rather than which purchase criteria are the most relevant. While
it is prudent to know what customers feel is important in a product
or service, it is critical to know what criteria they value when
making relative purchasing decisions.
Asking people directly determines what is known as the 'stated
importance'. To determine stated importance, people are asked to
indicate the importance of each attribute in a set of attributes
using one of several methods. Typical methods include: 1) Interval
scaling - the person reads a statement indicating its importance
on a 7 or 10 point scale 2) rank ordering - the person ranks the
order from a list of attributes 3) forced allocation - respondents
assign a specific number of points normally 10 to 100 across categories
of attributes. Using one or more of these stated importance techniques,
the researcher then creates a ranking and weighting of the relative
importance of the various attributes. In our experience the forced
allocation method provides the best information.
Each of the above approaches provides useful information provided
the company understands the differences in the output. For example,
with interval scaling, all of the items can be rated as very important
to the customer. With rank ordering, the technique does not provide
any information about the magnitude of difference between the items.
Just because one particular item is ranked in the middle of the
list doesn't mean it is much less important than the item ranked
at the top of the list.
One often cited example is that of the importance of airline safety.
It emerges as a high-stated importance attribute, but it is not
a significant predictor of airline choice. On time arrival of people
with their luggage is a better predictor of customer choice of an
airline. Safety is more of an expected or minimum requirement; something
that is expected, but does not contribute significantly to the prediction
of choice. However it could easily become a significant predictor
of choice given a poor record of safety violations and problems
(e.g., ValueJet in the United States). Well, if stated importance
doesn't indicate what customers value, is there a better way? Yes,
the derived approach has been found in practical applications in
many markets and countries to provide a strong indication of what
customers value when they make purchasing decisions. With this approach
econometric models are constructed using multiple regression techniques
to determine impact weights rather than importance weights.
Consider a study of billing satisfaction. Courtesy was one of six
stated attributes of overall billing service quality. But, using
the derived technique, courtesy had such a low relative derived
impact weight that it was statistically insignificant. Intuitively
this is not plausible since lack of courtesy to a customer inquiring
about his or her bill certainly does influence perceptions of billing
service quality.
Digging deeper into the data showed that customers almost always
scored courtesy as excellent, with very little variation. A variable
that exhibits little or no variation cannot explain variation in
another variable and will exhibit low or no impact (or derived impact).
This in no way says courtesy is unimportant; rather it says that
given the current consistent high level of service provider performance,
courtesy is not an issue. But the onset of discourtesy would raise
it to issue status. It would then start to impact customer purchasing
behaviour and the models would begin to show an increasing impact
weight.
Another example comes from American Express Cards (pages 234-235,
Customer Satisfaction Measurement and Management, Naumann and Giel):
Customers stated that their most important need was immediate replacement
of their card if it was lost. In actual fact less than 3% of customers
ever lose a credit card, and these customers were happy with the
existing replacement procedure. From regression analysis, American
Express learned that what customers really valued was the card being
accepted widely, the billing being correct, and the card being perceived
as a prestige item. From this work American Express focused their
process improvement work on the key purchase criteria indicated
by the derived impact weights, and focused marketing efforts on
the replacement card issue where customers were already happy with
the service level.
It is not always possible to use the derived impact weight approach
due to factors such as a limited budget or a small number of respondents.
When it is possible to undertake modelling, the derived approach
indicates the right customer value road that managers need to follow
on to improve sales.
By Ray Kordupleski, President Customer Value Management Inc.
Pancakes for the General
The mist slowly lifts from Lake Shongun, New Jersey. Is that General
Washington's boat appearing through the mist? Has he heard about
the pancakes that our host is cooking for brunch? This is the area
where Washington wintered his army and many of the victories over
the British were won. While 200 years ago it might have been the
General dropping in, today it is one of our Customer Value associates.
Ray Kordupleski, President of Customer Value Management Inc., is
a master pancake maker as well as being the expert on applying customer
value techniques. Ray tells us that his pancake recipe has evolved
through experimentation and a little theory, in the same way that
the Customer Value Management approach has developed. The pancakes
are simply marvellous - crisp on the outside but soft and tasty
inside, served with maple syrup, complemented by a fresh fruit salad
based around Kiwifruit.
Regards,
Rodger Gallagher
Yogurt Pancakes
1 cup flour, 1 teaspoon baking powder, ½ teaspoon baking soda,
¼ teaspoon salt, 1 tablespoon sugar, a dash of nutmeg, a dash of
cinnamon, a dash of cardamon (optional), 1 cup vanilla yogurt, 2
eggs separated, ¼ cup vegetable oil, and milk as needed for consistency.
Mix dry ingredients. Make a well in center. Mix together yogurt,
egg yolks, and oil. Put yogurt mixture into well. Slowly work flour
in. Whip egg whites. Stir into mixture. Add enough milk for a pourable
consistency. Get griddle hot. Pour batter to make pancake. Cook
on one side till it bubbles. Flip over. Let cook.
These pancakes are quite successful at our brunches and at both
the Indian Guide and Indian Princess campouts.
Ray Kordupleski
|
