|
 |  |
 |
 Overview Success Stories
|  |
Chordiant Decision Management—Turnover Prediction A major oil company operating service stations turned to Chordiant Decision Management for reliable indications of expected turnover for specific sites. The ChallengeNewly formulated environmental guidelines in the Benelux forced our customer to invest substantially in existing sites. Investing in new sites is constantly being considered as well. In order to reach informed decisions about which existing sites to invest in and which new sites to develop, the oil company needed a reliable indication of the expected turnover for a site. The Solution Consultants using Chordiant Decision Management developed a set of models that predicted a site's turnover based on information regarding the surroundings of a location, available facilities and so on. The first models were specifically developed for the Belgian market, so data concerning some 300 gas stations in Belgium were gathered. Because of the diverse nature of available information and significant differences in predictive performance, it was decided to develop a number of models that each assessed a site from a particular viewpoint. Four models were developed. The first model assessed a site on the basis of its surroundings using commercially available sociodemographic data, data about nearby competitors and specially collected data like traffic counts. The second model based its prediction on a site's facilities like the number of tanking positions or the availability of a car wash. The third model considered information about the operation of the station, like opening hours and level of service. The fourth model served to combine the predictions of the other three into an overall assessment of a site. The Outcome 
The oil company uses a geographic information
system (GIS) to access and store data about
their gas station sites. The resulting models
were integrated with this system to allow for
interactive inspection of the models'
predictions. The result of this integration can
be seen in Figure 1. By selecting a site on the
map, the user can analyze the site's particulars
and investigate what-if scenarios using the
models' predictions.
The ResultThese what-if scenarios provide our customer insight in the impact of changing a site and the likely turnover for a newly planned site. They can now interactively assess the feasibility of investing in existing sites, distinguish sites where upgrading or extending the gas station will yield significant benefits and discover new profitable sites. |  | |
 |
| © 2000-2008, Chordiant Software, Inc. All rights reserved. Legal/Disclaimer | Privacy Policy | Site Map | |  |
| Solutions | Retail Banking | Card Services | Lending | Insurance | Telecommunications | News & Events | About Chordiant | |
