Enterprise Predictive Modeling Platform


The Enterprise Predictive Modeling Platform is a world-class hardware and software solution that centralizes and standardizes all predictive modeling and reporting across the enterprise. Among other things, companies use modeling to predict future customer behavior such as likelihood to purchase a specific product, or to leave for a competitor. Modeling enables great improvements in targeting accuracy and efficiency that help focus marketing and retention efforts on key customers with the greatest impact on profit.

A new modeling project can regularly take from several weeks, to a few months to complete primarily because of the preparatory work that proceeds model building. The search for, identification, acquisition, understanding, exploration, handling (merging, duplication), and cleaning of data from many different internal and external sources limits the number of models that can be built by a single resource. Consequently, the few models that are built cannot be rapidly rebuilt to accommodate rapid shifts in the competitive landscape occur. Enterprise Predictive Modeling Platform was created to remove many of these short comings.

Data from automated feeds from legacy systems and other disparate sources (including BI and CRM applications) are directed into the platform’s data mart and subsequently cleansed, pre-processed, and standardized by other automated processes for subsequent modeling and reporting. All data handling processes (scripts) are automated to ensure that periodic (daily, weekly, monthly, etc.) captures are timely and complete. Historical data are maintained for trending and other modeling purposed, and at definable intervals they are archived for possible future use.

Consistent, standardized data are now available from a centralized location for use by analysts throughout the enterprise for modeling, forecasting, data mining, metrics creation, process management, reporting and other key activities. Reports and models developed by one department are comparable and consistent with those from other departments due to the use of the same starting data. Models are more rapidly developed. Collaboration with other internal groups seeking automated or ad hoc reporting and a source of data for metrics etc. are enhanced. "Piggy-backing" opportunities for other important analyses – e.g. customer life-time value calculation, business life-cycle assessments, credit worthiness, price-value assessments, line of credit optimization, price-elasticity analysis, etc. become possible.

An invaluable feature that increases modeling productivity is a customized scoring module that is built on top and integrated into the platform’s data mart. After a model is built, all customers in the database are scored with the model’s output, and the model is archived. Within a short period of time new customers are added to the database, but they have yet to be scored. You may choose to update the customer’s scores on a monthly basis to ensure that all new customers receive a score. The scoring module automatically performs this function to ensure that no customer falls through the cracks, and that all scores are current. All scores are stored in the modeling data mart and are instantly available for access by list management, sales, marketing, advertising, product development, and other groups.

As the marketplace changes over time most models become less accurate in their ability to predict. Modeling metrics can be leveraged to identify which models need to be rebuilt and when. A model can be rebuilt and scores deployed in a matter of hours. This ensures a high degree of modeling accuracy and consistency over time, through changing market conditions, and across the enterprise.

Inteletix’s Enterprise Predictive Modeling Platform is a powerful profit and productivity enhancement system for your business. Some clients reported that the Enterprise Predictive Modeling System:

  • Accommodates at least 115 new models per year (more can be added as they are conceived)
  • Reduced model development and deployment time from months to days.
  • Requires only a single resource to administer and maintain
  • Increased outbound sales by 2-20 fold in terms of sales conversions (closed sales) depending on the specific product
  • Enabled monetized metrics to be produced that facilitated socializing the benefits of models to C-level executives, as well as customer-facing employees, thus gaining more rapid buy-in and critical support for daily use of models in strategy formation, product management, and facilities deployment
  • Generated the first churn model that subsequently supported the formation of new retention strategies. Over 52% of highly vulnerable customers were retained by use of a term contract. Customer erosion was reversed and stabilized over the next 18 months, saving $41.4M.
  • Includes propensity to buy (specific product) models, customer growth potential index, churn models, Key performance indicators for various products, modeling tracking metrics
  • Platform partially responsible for a 10% increase in revenue growth in one year.

learn more

E-mail: info@inteletix.com