Categorica™ is an application for building product structures to support advanced product management applications in the retail sector. It is applicable to both suppliers and distributors, and is particularly valuable when used to enable effective implementation of the Category Management business process, as defined in [ECR 1995] and [ECRE 1997].
The product structure is a model of the consumer's product selection process. Through advanced product management applications that support strategic planning based upon a consumer-oriented product structure, the business can expect to:
The result of these benefits is the development of a more competitive, profitable, and robust business.
The first step in an advanced product management business process is product set definition (called category definition in the category management process). The definition of a product set has two parts: the specification of the group of products that comprise the set, and the creation of a product structure, which gives a hierarchical organization to the set of products. The product structure breaks the larger group of products into successively smaller, more manageable groups of products.
The most common criteria used to determine which products should be in a product set are:
Figure 1 shows the top three levels of an example product structure for the carbonated soft drinks category. It is common to refer to the levels of a product structure as the category, subcategory, segment, subsegment, and microsegment of products. Of course, some product structures will have more than five levels, in which case more level names will be required.
Figure 1: A partial product structure for a carbonated soft drinks category, oriented so that moving "down" the hierarchy means following lines to the right. The carbonated soft drinks category is broken down into two subcategories by type: cola and non-cola. The cola subcategory is broken down into the two segments by package type: bottle and can. The non-cola subcategory is broken down into five segments by flavor: orange, root beer, lemon/lime, ginger ale, and other. The rule for determining the segment differs by subcategory (package type for the cola subcategory, and flavor for the non-cola subcategory).
The white rectangles of Figure 1 represent groups of products appearing in the product structure. The black rectangles represent the distinguishing characteristic used at each decision point in the product structure hierarchy. As this example illustrates, product structures can use more than one characteristic per level to determine the next level's sub-groups.
In the cola subcategory, a product's package type determines its segment. But, in the non-cola subcategory, a product's flavor determines its segment. Every group in the hierarchy could potentially have a different rule for further discriminating products into the next level of detail.
Product characteristics that have a greater influence on the consumer's product selection appear as discriminators higher in the hierarchy. Figure 1 indicates that within carbonated soft drinks, the most important characteristic is the type (cola vs. non-cola).
Different branches of a product structure may have different lengths. One branch may have individual products immediately below the segment level, while another branch may continue down to the micro-segment level, or further.
Each product in the product set "rolls up" exactly one path to the top level. Said another way, each discriminator (black box in Figure 1) forms a partition of the parent group, yielding the individual sub-groups at the next level "down" the hierarchy.
The product structure provides the reporting structure that is used throughout an advanced product management process, and is invaluable for ongoing performance monitoring. Business planning templates and monitoring reports have increased value when they are organized by the consumer-centric product structure.
For maximum benefit, business planning and reporting systems should capture as many product characteristics as possible. This includes:
By merging all these sources of product information into one coherent view of products, the business will increase its ability to analyze and optimize performance.
The product structure must be incorporated into the applications that support advanced product management processes. Usually, this means capturing it in a database-ready format and constructing automated database loading procedures to update the application databases.
Given a set of products and their characteristics, there must be a systematic way to place the products appropriately in the product structure because:
As new products appear, information about them must make its way into the business' supporting systems. Human intervention to add product structure information is tedious, time-consuming, costly, and error-prone. Therefore, it is vital that there be an automated system that examines the characteristics associated with a new product and applies a set of carefully constructed rules to place the product in the product structure.
Since advanced marketing applications require the use of syndicated market, consumer and competitor data, the set of products appearing in analyses will be much larger than the set that would be used for strictly internal reporting. This larger set of products must go through the same placement process as internal products. As before, manually adding product structure information to all products in the market is not feasible.
As the business gains more experience from operating in a consumer-focused mode, it is inevitable and in fact desired that modifications to product structures will become necessary. This activity requires the ability to examine the impact of the product structure changes on a variety of business performance metrics (such as gross sales, units sold, and gross profit). As products are moved within the structure, the values of these metrics will change, and the business user must be able to see a side-by-side comparison of different structures to perform the needed "what-if" analyses.
An easy-to-use interactive application must support this activity in order to maximize the productivity of the business users. It must be easy to create, explore, modify, and compare product structures, so that the overall cost of change is low. Only with such an application will the business be able to fluidly incorporate new knowledge about its consumers and marketplace into the product structure, and turn such knowledge into a competitive advantage.
There are two complementary aspects to product structure management. First, there must be a front-end application which empowers business users to construct their own product structures on an experimental basis, to determine the most effective organization of their business. Second, there must be a component that supports back-end processing for the population of data warehouses and other large systems with the information created through the front-end tool.
Categorica™ provides both of these components in an integrated product structure management application.
The Categorica™ engine is a hierarchical rules-based categorization engine that can be used for both interactive applications and automated data engineering systems.
The Categorica™ front-end interface is a workbench application with which business users can:
The Categorica™ back-end applies a set of product structure rules to a set of products, yielding an updated product set with the additional characteristics to support the product structure.
Any supplier or distributor in the retail sector that is implementing an advanced marketing business process benefits from the product structure management capabilities of Categorica™. Most of the tier-one retailers and suppliers are either implementing or planning to implement the category management business process, which is an excellent example of an advanced marketing business process with a consumer focus. For these companies, Categorica™ provides critical enabling functionality, whether the business process is being supported through a PC application, a data mart application, or a data warehouse application. Data warehouse applications benefit greatly from the Categorica™ engine's automation capabilities.
For more information on Categorica™, please contact Focus Research, Inc. at:
Focus Research, Inc.
P.O. Box 629 Cupertino, CA 95015
gregor@focusresearch.com
[ECR 1995] Category Management Report: Enhancing Consumer Value in the Grocery Industry, Joint Industry Project on Efficient Consumer Response, 1995.
[ECRE 1997] Category Management Best Practices Report, Efficient Consumer Response Europe, 1997.
1. A syndicator provides market, competitor, and consumer data. AC Nielsen and IRI are two major syndicated data providers.
Copyright © 2008 Gregor N. Purdy. All rights reserved.