they perceive as being most important. Others may prefer the sales offering that
achieves some acceptable assessment score across each and every attribute desired by
the buyer. However, research into how purchase decisions are made suggests that
most buyers use a compensatory, multiattribute model incorporating weighted
averages across desired characteristics.3 These weighted averages incorporate
(1) assessments of how well the product or supplier performs in meeting each of the
specified characteristics and (2) the relative importance of each specified characteristic.
Assessment of Product or Supplier Performance
The first step in applying the multiattribute model is to objectively rate how well each
characteristic of the competing products or suppliers meets the buyers’ needs. Let us
use the example of General Motors (GM) evaluating adhesives for use in manufacturing.
The buyers have narrowed the alternatives to products proposed by three suppliers:
BondIt #302, AdCo #45, and StikFast #217. As illustrated in Exhibit 3.3,
the GM buying team has assessed the competitive products according to how well
they perform on certain important attributes. These assessments are converted to
scores as depicted in Exhibit 3.4, with scores ranging from 1 (very poor performance)
to 10 (excellent performance).
As illustrated, no single product is consistently outstanding across each of the
eight identified characteristics. Although BondIt #302 is easy to apply and uses
the buyer’s current equipment, it is also more expensive and has the shortest durability
time in the field. StikFast #217 also scores well for ease of application, and it
has superior durability. However, it has the longest bonding time and could negatively
influence production time.
Accounting for Relative Importance of Each Characteristic
To properly compare these performance differences, each score must be weighted by
the characteristic’s perceived importance. In the adhesive example, importance
weights are assigned on a scale of 1 (relatively unimportant) to 10 (very important).