CASE STUDY: Lever Europe - ITS Monitor

Tools: Oracle OLAP
Industry: Fast Moving Consumer Goods
Application: Decision Support System

Overview

Lever Europe operates in 16 Markets and has a dozen factories. These are linked through an Internal Trading System (ITS). To examine the relationships between the factories and national markets a decision support system called ITS Monitor was created. This provided quantitative analysis of the internal supply chain.

A set of models was developed and displayed using Oracle OLAP. Results were made available from a central server to sites across Europe.

Some details in this Case Study have been changed to protect client confidentiality.

Return on Investment

Systems of this scale represent a significant investment. However the cost is small compared to the value of the transactions examined. Any improvements in factory services improve the whole supply chain.

The solid design and build of the system allowed the majority of the programming effort to continue to be used despite large changes in the systems technical environment. The system began with DOS client (the software on the PC) and a VAX server (The number crunching software on a central computer) using data from a proprietary logistics system. This system progressively moved to a Windows NT client running on UNIX server taking data from a bespoke Oracle system. The majority of the program code stayed the same throughout these changes protecting the client's investment.

Company Background

Lever Europe was formed by the federation of Unilever's European detergents assets. This European wide perspective allowed the factory network to specialise into a number of production units that served the whole of Europe.

Situation

Previously the supply chain was managed by the production, logistics and sales functions within a country. In the last 10 years increasingly sophisticated transaction systems have managed a rise in cross border supply. The growing importance of cross border supply made key performance indicator necessary. These monitor the internal supply chain relationships and allow setting of objectives and development of service level agreements.

Solution

A decision support system discussed here was developed. Later Lever Brothers also developed in house a management information system giving summaries of transactions, forecasts and stock levels.

The ITS Monitor used data from the ITS transaction system and added models of the business supply chain. Data was loaded on a weekly basis into a data-warehouse. The business models were then applied to this data to give a number of key performance indicators. These examined flexibility, reliability and variability of supply, presenting to users as a series of drillable datacubes.

Village maintained the system during most of its life. Weekly operations and monitoring of the system was undertaken. The system was upgraded into Windows 3.1, Windows NT and web versions of client software. Programming changes were made to reflect changes in business rules and underlying transactions system. New measures were added to the system.

Technology

The server is based on Oracle OLAP (also known as Express) running on a UNIX server. Various Oracle OLAP clients have been developed. Base data was imported weekly from a relational database system.

OLAP technologies (On Line Analytical Processing) are different from relational database technologies in their approach data management. Use of this technology would only be appropriate in projects over £100,000. For smaller projects and quantities of data a similar system could be developed using relational databases and suitable client components such as Borland's data-cubes.

General Conclusions

The related areas of Decision Support and Management Information System are critical in competitive environments. If an organisation has good transaction processing systems valuable data can be made available to help management.

It is relatively cheap to produce a simple management information system to present managers with summary data across their operation. The cost of this might start at about £5,000 for a single site system drawing data from one or two transaction systems. Simple multi-site systems costs might start from £10,000.

Decision support systems (DSS) such as the ITS Monitor are complicated to design, build and maintain. DSS's contain business models that are applied to data to provide for example key performance indicators. A simple DSS might cost £15,000. A system of the complexity of the ITS Monitor would cost more than £100,00. One off decision support requirements are often managed at a lower cost using spreadsheets.

We believe the long survival of the ITS Monitor is a tribute to the correct early design decisions.