| Our
Approach to Acquiring New Customers |
| |
| New marketing science based
approaches have been developed to improve targeting accuracy. We will
discuss here two such approaches; value-based customization of advertising
messages, and targeting of high likelihood prospects using advanced
data mining techniques. |
| |
| Value-Based Customization
of Advertising Messages: The holy
grail of advertising has always been to discover and leverage the
"hot button" of each individual prospect. In this way, we
do not waste advertising resources by selling our prospects an aspect
of our product or service that does not meet their current need. Until
recently this "segment of one" sales concept has only been
a pipedream. After all, how is it possible to "know" the
most important feature of a product or service for each individual
prospect? Now however, new advances in measuring values and correctly
predicting those values for unsurveyed prospects have transformed
this "pipe-dream" into a reality. The steps in value-based
customization are as follows: |
| |
- Develop a values survey for prospects that
differs by product and industry.
|
- Use a survey methodology that precisely measures
the importance of key drivers of the product or service in question.
The survey approach we use to precisely measure these values is
a proprietary method of conjoint analysis (MOPT). See our ValueQuest™
tool for more information.
|
- Uncover the "hot buttons" of prospects
by finding clusters of prospects who value similar things. To
find the purest clusters of prospects we use our advanced multi-algorithmic
cluster analysis decision support tool, SEGMENTSOLVE™.
The clusters developed through this DSS tool are more defined
and fully characterized against demographics and habit and practices
data. We therefore term the resulting clusters "buyer value
segments."
|
- Build a simulation model which presents different
product concepts to the virtual customers built from each survey.
The advanced simulation system we use is our conjoint-based, simulation
modeling tool, PREFSOLVE™.
This simulator discovers how to optimally customize the product
or service to those "buyer value segments."
|
- Develop a predictive multivariate model which
predicts the buyer value clusters and individual values for customers
that did not take the survey. (See
the AIM™ protocol.)
|
- The end result of the predictive model is
an equation that calculates for any new prospect which "message"
or advertising "pitch" to use for each prospect.
|
| |
| |
| When the cross-hairs of
strategy focus on prospecting, the insights of the knowledge base
worker and her datamart or datastore are often left out of the loop.
This is not surprising. Not too long ago in its evolutionary past,
datamarts were based centrally on CIF's (Customer Information Files).
By definition they contained no prospects. Yet, even absent of prospects,
prospecting can now be done by creating functional proxies for prospects.
The key here is in the definition of a customer. In essence, for the
multi-product company, a customer can be both a prospect and can stand
in for a customer. The optimal method to do this, varies by industry,
but the shared goal is to emulate prospects from sub-segments of the
customer base captured in the datastore. |
| |
| The real benefit of leveraging
the datamart to develop prospecting strategies is not merely convenience.
New protocols such as Ultra-High Performance Data Mining (UHP-DM™)
now can leverage distributed, automated data mining tools. These tools
dramatically improve the accuracy of predictive data mining. By using
these tools, prospecting proxies can far outperform classical "if-then-else"
prospecting "rules" produced by conventional methods. |
