| Oracle® Retail Advanced Science Engine Implementation Guide Release 14.1 E59126-02 |
|
 Previous |
 Next |
| Oracle® Retail Advanced Science Engine Implementation Guide Release 14.1 E59126-02 |
|
Previous |
Next |
Oracle Retail Advanced Science Engine (ORASE) is the centralized science engine that supports retail business processes by driving the analytics for both the Oracle Retail Advanced Science Engine (ORASE) and for the Oracle Retail Assortment and Space Optimization (ASO).
ORASE performs data mining and develops analytical parameters to support business processes in Oracle Retail Category Management (CM), Oracle Retail Demand Forecasting (RDF), and Oracle Retail Analytics (RA). It is comprised of the following modules:
Customer Decision Tree (CDT)
Demand Transference (DT)
Advanced Clustering (AC)
Market Basket Analysis (MBA)
ASO provides a way for planners to inform their decisions about assortment rationalization and to perform Micro Space Optimization.
This implementation guide addresses the implementation of ORASE and ASO.
Implementers should be familiar with application servers, the installation process, Oracle databases, system and IT administration, RPAS applications, UNIX commands, including shell configurations and scripts, UNIX directory operations, and symlinks.
ORASE is architected in a modular fashion to serve as a centralized science engine supporting multiple solutions. Figure 1-1 and the discussion that follows illustrate the interaction among the various components that comprise ORASE. The four applications shown in the figure within CM are:
Category Planning (CP)
Assortment Planning and Optimization (APO)
ASO
Market Basket Analysis
ORASE provides science-based functional extensions to CM, RDF, and RA. These three applications provide data inputs to ORASE and ASO and receive data outputs (Clusters, CDTs, and DT models) from ORASE. ASO also generates optimized assortments and planograms (POGs) and replenishment updates. ASO results are not passed as output, but may be extracted from the ORASE database, if desired.
CM is a category planning, assortment planning, and assortment rationalization tool. CM uses Clusters, CDTs, and DT models from ORASE as well as space-optimized assortments from ASO to determine which categories should be carried in a store, how much space should be allocated to a category, which items should be in a category assortment, and how much space should be allocated to an item.
RDF is a statistical and promotional forecasting solution. With the introduction of DT models from ORASE, RDF is able to forecast demand change when an item is either removed or added to an assortment.
RA is a suite of retail-enterprise-level fully integrated Oracle BI applications. ORASE provides RA clusters to analyze sales, inventory levels, promotions and customer data, and market basket analysis to calculate product affinity relationships.
ORASE has its own intra-ETL (extraction, transform, and load) ability that reads from RA or the RA Data Model (RADM). ORASE leverages data from RADM via direct schema reads and must be co-deployed on the same database. Note that ORASE does not require the RA BI product, only RADM. In addition, ORASE has its own schema that has been optimized for the analytical processing required for its science modules.
The Oracle Retail Merchandising System (RMS)/RA ETL is available to ORASE retailers, so that they can load ORASE data from RMS into RADM and the ORASE schema. See the RA and RMS documentation sets for more information on the ETL associated with RMS and RA.
The ORASE solutions have a similar workflow and user interface (UI). The workflow lets users implement new science modules using similar techniques. For example, a retailer who uses Demand Transference and the Customer Decision Tree may then be able to more easily learn and use Advanced Clustering and other aspects of demand modeling. This approach lowers the future total cost of implementing various science modules.
The Oracle Retail Advanced Science Engine User Guide and the Oracle Retail Assortment and Space Optimization User Guide provide details about using each of these applications.
This section provides an overview of each of the ORASE modules.
Customer Decision Trees (CDT), with their dynamic hierarchical structure, help retailers gain insights into the customer decision process. CDT results illustrate a prioritization or importance of specific product attributes that determined the customer's purchase.
The product includes new science that mines retailer data to understand customer behavior and preferences across multiple channels to develop CDTs. The solution provides insights into what attributes are driving customer purchases. This CDT generation process can be further informed by retailer business insights around attribute prioritization or supplier CDTs.
Demand Transference (DT) refers to the shifting or transfer of demand among items within an assortment, as items are added to or deleted from the assortment.
Demand Transference science mines retailer data to identify demand transference effects, which are then used within CM and RDF to drive plans and forecasts informed by planned assortment changes. ASO uses the results from DT to predict the effects on demand of similar SKUs as SKUs are dropped or added to an assortment.
Advanced Clustering (AC) builds store clusters with similar consumer demand patterns and integrates those clusters into solutions such as assortment planning, category management, pricing, promotion, allocation, and the supply chain.
AC also groups like stores, items, and entities, based on sales volume, profit margin, store format, customer type, demand profile, and promotional effectiveness.
ASO provides a way for planners to make decisions about optimized assortments. It takes as its input the collection of planograms and the assortments that are mapped to the planograms across a set of stores. A planogram is a collection of fixtures (shelves, pegboards, freezer cabinets) of various lengths. Stores may be grouped together into clusters that share some user-defined characteristics. ASO provides the user with the means to optimize the assortment and the space allocated to it to meet a variety of business goals.
