The Client
One of the largest credit card processors in the world needed to identify specific
target merchants for launch of a new transaction processing product for merchants. The client understood that "bigger is not
always better" when evaluating merchants to seed a new program. The client also needed to specify targets, by name, as part
of a joint venture agreement prior to the product launch.
The Challenge
As with any product launch, the client had a standard set of questions that needed
to be answered. What merchants are most likely to buy the new product? What merchants are capable of using it? Which merchants
would I really want to use the new product (e.g., as a showcase or "bell cow" account)? However, a key limitation of the launch
was that the client could only name a specified number of merchants to reserve or set aside from the joint venture, which
offered a competitive solution.
How We Helped
The client asked The Ashtree Group to develop a methodology to evaluate and determine
the target merchants. We started with the most basic question: What would be considered a successful product launch? Through
many discussions, it was all boiled down to "X number of merchants processing Y transactions per month". Ultimately, transaction
volume was the key, but that would come as a product of merchants and size. Not surprisingly, the client wanted to target
the largest merchants that would buy the product.
This led to the development of a new business indicator for merchants. The measurement
was comprised of two dimensions: transaction volume and likelihood. Each merchant was plotted on a quadrant using the new
indicator. The upper right quadrant held the highest quality targets and the lower left quadrant held the lowest quality targets.
This proven evaluation matrix technique was simple in concept, but certainly more difficult in implementation.
Beginning with this concept, we then set about defining the specific attributes
of a merchant, how those attributes applied to a scoring methodology for plotting on the matrix, and where we could find supporting
data. We quickly found that much of the data for existing merchant customers was available, but was scattered about different
(and huge) databases. For existing customers, we developed extracts from multiple databases to feed a central model.
For those merchants that were not existing customers, we developed techniques for converting publicly available information
into data that we could pump through the scoring methodology. This allowed for an "apples to apples" comparison of volume.
The more difficult aspect of the project was defining the likelihood of purchase
dimension. We defined and rated several "soft" criteria such as use of existing client products, use of competitor products,
relationships with banking partners, etc. Codifying folklore, anecdotes, and soft business intelligence resulted in a hard
score that could be used to compare merchants on the "likelihood" dimension.
We built a database, model, and "dashboard" to collect the extracts from larger
data stores, and to add the results of the "soft" evaluations for likelihood. The database (in MS Access) performed all of
the number crunching, sorting, and ranking, and ultimately generated the "Volume" score and "Likelihood" score which was used
to plot the merchant in the quadrants.
As with any evaluation model, the inputs change over time. So, the database allowed
for routine input of fresh data to constantly re-evaluate merchants to ensure highest quality targetings (and annual re-negotiation
of the joint venture agreement). We also developed a sales activity tracking portion of the database to allow the product
executive to track her sales force and sales "funnel".
Success
The results of the segmentation delivered a list of merchants, with supporting
data, for targeting the new product launch. These merchants were carved out of the joint venture agreement and reserved for
the client's own sales force to pursue. The sponsoring executive stated: "This [model] is the only place in our company where
we can find out everything we need to know about a merchant - including those that aren't our customers". Furthermore, the
segmentation provided quick traction for the product launch, delivering the early success required of the program.
