DirectMail.com Receives Patent for GeoSelector - Business Geographics Made Easy
DirectMail.com recently received a patent
for the GeoSelector technology with which marketers may interactively
target a geographic region and purchase potential prospects for a mail
campaign, all via a mapping interface. Editor in Chief Joe Francica
interviewed Shawn Salta, a vice president at DirectMail.com, to determine
what aspects of the geocoding technology were patented and to discuss
the extent to which demographic variables can be used to do more
precise target marketing.
Directions Magazine (DM): What aspect of GeoSelector is patented?
Drawing a targeted area or a buffer zone does not seem like a function
that would be patented because it is used by many other mapping
Shawn Salta (SS): Our inventive concept of GeoSelector was to solve
a problem within the direct marketing industry. Prior to our invention,
the methods of selecting data from a database in order to build a
mailing list were time-consuming and lacked true targeting
abilities. Anybody wishing to target a precise audience, based
upon geography, was restricted to selection by radius, ZIP Code or some
other location-defining criterion such as street, city, state, etc.
Since GeoSelector's inception, we have continued to develop a growing
suite of high value applications for the direct marketing community. We
bring value to our customers by automating the previously inefficient
and costly tasks of precise targeting, market intelligence and data
visualization. We are not a mapping or GIS focused organization. How
our patent compares with the functions of mapping software solution
providers would have to be determined by individuals who have the
technical expertise to provide opinions on such. I have copied the
abstract of our invention below - you can view our patent in more
detail from the USPTO website - our patent number is 7,561,169.
"An embodiment relates generally a method of generating
user-specified information. The method includes selecting a
plurality of points on a map, where the plurality of points forms a
closed polygon. The method also includes determining a plurality of
coordinate pairs, where each coordinate pair is associated with a point
from the plurality of points. The method also includes retrieving user
specified information for an area enclosed by the plurality of
DM: Did you build the geocoder from scratch? Can you more fully explain the unique aspects of the geocoder?
SS: We really didn't build our own geocoder; we used a free,
open-source one because we couldn't afford to pay for a commercial one
at the time. The geocoder used for this early version came from http://www.geocoder.us/.
(The perl script version was placed in a .net >wrapper.) The
geocoder used data from TIGER/Line, downloaded from the Census
[Bureau]. It had the benefit of being free, and in many cases, though
not all, the data were at the household level. It was better than what
we were getting at the time since those data were not even coded at the
household level at all.
Today, our data providers use a commercially available geocoding solution based on Tele Atlas data.
DM: Can you elaborate on the pricing options?
SS: We provide a basic version of GeoSelector to the general public
on our website, where users are granted free access to the software and
the demographic reports without having to first register and without
having to purchase a single record. The "fee" for using that version of
GeoSelector is built into the cost of the list that's ultimately
purchased, and even with that, our prices are among the best available.
Our "Enterprise" version of GeoSelector provides for additional robust
features, the most important of which allows users to upload their own
database for overlays, appends, look-alike locators and in-depth
reporting. It is licensed on a very attractive fee structure, depending
upon the number of users and the number of records which we store for
DM: Will you host users' data for them?
SS: Yes, we can host users' data. As we've just mentioned, the
chief distinction between the basic version of GeoSelector and the
Enterprise version is the ability to use the application overtop of
your own database. The beauty of GeoSelector is its ability to run
overtop of any database, with few limits on the number of records or
the type of, or number of, selects. There's no software to purchase,
you can learn to use it very quickly, and it allows you to understand
your data visually, in ways that are simply impossible in a flat file
or data report.
DM: Explain the "look-alike" functionality. What criteria can the
user select in order to filter information on a precise industry
SS: When you refer to industry clusters, we're going to assume that you're talking about business lists.
For consumer lists, visitors to our site have the ability to refine
their lists, using hundreds of select criteria. Some of the most
popular are based on demographics by age, income, marital status,
occupation, etc. Additional filtering by lifestyles, homeowner and
mortgage data is also possible. Mosaic clustering and more are also
provided. Mosaic clusters place records into one of 60 groups,
identified by affinity or other shared characteristics that are also
crossed up with True Touch data, which further identifies the channel
dominance or preference by breaking the population into a dozen more
GeoSelector's Reporting functionality allows the user to determine
which Mosaic clusters are most prevalent in an uploaded data set. Using
this information, it allows the user to select prospect records which
have many of the same demographic characteristics as the source record
by selecting like clusters. In addition, we have technology in
development which will analyze the demographic characteristics of the
source record on an individual level to determine which demographic
variables have the most profound effect on response. We then allow the
user to target to precisely those values through our demographic
selection process. GeoSelector can be customized to support filtering
criteria based upon any values within a dataset.
DM: You mention "high-end" demographic marketing tools. Whose
demographic data are you using? Are you employing a psychographic
SS: Our licensed consumer data are from Experian - Mosaic
Clustering, True Touch Impacts and Behavior Bank Elements are available
within the system.
Mosaic is a household based segmentation system that classifies all
U.S. households and neighborhoods into 60 Mosaic types and 12 groupings
that share similar demographic and socio-economic characteristics.
True Touch is a multi-dimensional contact system that, when combined
with Mosaic, not only helps users identify their best prospects and
customers, but also tells them what type of messages and offers
motivate their purchases, the communication channels they prefer, and
the times they can be reached.
Behavior Bank Elements are household indicators of self-reported past transactional experiences.
Our business leads database is from InfoUSA.
DM: Do the modeling routines you have built into GeoSelector use a spatial interaction model such as a Huff or gravity model?
SS: Our modeling is based around an SPSS decision tree, which
analyzes the relationships between demographic variables on a file as
they relate to response. This feature is not available in the current
product release, but will be available in the near future.
DM: Is your routing and scheduling module available for a separate price or is it truly built into the product?
SS: I am not sure what a scheduling module is in the context of
GeoSelector. The routing used is provided by a publically accessible
third-party Web API. GeoSelector does not provide its own routing
engine; it simply buffers around existing polyline routes up to a