STORAGE AND RETRIEVAL

To a large extent, data warehousing holds the key to retailers' and wholesalers' ability to develop powerful data-base marketing and decision-support tools capable of leading the industry to future success.

mination to make use of emerging technology in this area is just now moving to the forefront of retailers' and wholesalers' strategic plans.

In the next few years, data warehousing will be a critical factor in determining which companies succeed in moving forward with a wide variety of initiatives, from comprehensive frequent shopper programs to complex executive information systems tools.

At issue is how to collect, store and make use of the huge amount of shopper transaction data now being collected daily at the point of sale. Questions such as which types of data warehousing systems work best, what role they can play in the future and what their potential is for serving state-of-the-art, information-driven systems must still be answered. But clearly the industry is now engaged in rolling out these systems and testing their capabilities. Here's how SN's discussion with five key executives went on the subject:

SN: What is driving the need for data warehousing systems today?

PETER ROLANDELLI: What we're talking about with data warehousing is the growing demand for information in all areas of customer data, by category and by item. For example, a company that is planning to drop a particular item may want to see which customers are purchasing it. If they find out their best customers are buying it, they may decide to keep the item. But doing that depends on having the ability to store a huge amount of data and analyze it quickly. Data warehousing is a technical innovation that will allow us to get a large amount of information, whatever we are looking for, easier.

PATRICK STEELE: What is driving data warehousing is the need for information to make decisions -- whether that means looking at product movement for category management or capturing sales trends and individual item transactions for data-base marketing.

All of a sudden the need to store vast amounts of data on a weekly, monthly, quarterly or yearly basis in such a way that people can have easy access to it is becoming extremely important. Companies want to be able to mine all the data they have to react quickly to business needs and improve sales.

BILL MAY: This idea of data warehousing is tied closely to the issue of decision-support systems. If an organization wants to be successful, it must be able to capture the right data from all its systems and use it for making decisions. The goal is to have a single data base for all customer transactions, so that as I upgrade a data base it will also be upgraded for everyone else in the organization.

DAVID HAYES: Data warehousing is essentially a place where end users can go to tap into decision support systems and get answers. It's a condensed user-driven data base. The use of data warehousing is becoming much more important.

The data we all are now collecting has to be combined to analyze things such as total cost and profit of delivering products to a customer. To do that requires consolidating different data bases, a task that traditionally was performed on a mainframe. But it can now be done using a data warehouse to which end users have direct access.

RAY HAMILTON: Data warehousing allows a company to pull information from a single data base that contains all the critical information collected over time. The concept of data warehousing is based on the ability to store huge amounts of data inexpensively with parallel computing. But data warehousing is a huge project and has to be done very, very well to meet expectations.

SN: Where is the industry today in terms of making use of data warehousing technology?

MAY: For us, it means extracting daily data from on-line transaction processing systems, storing it and making it available to a wide variety of associates and employees for decision-making. The concept is simple, but implementing it is extremely difficult. The data warehousing tools now available are still new and unproven. We are trying to figure out which ones make sense and which don't.

STEELE: The demand for using large amounts of data has created a heavy emphasis on developing tools for storing information. Conventional data-base management systems used for on-line transaction processes don't adequately support data warehousing needs. If a company uses a large data base for both

on-line processing and data warehousing, performance suffers. Data warehousing has emerged as a way to satisfy these needs.

It's scary how big these data bases are becoming. There is a lot of work being done today to develop solutions for handling vast, and I mean vast, amounts of data on a very timely basis to get answers to questions. The technology is getting better and better. Two or three years ago, only a very limited number of solutions existed for successfully storing large amounts of data. Today, there is a wide range of solutions and the technology continues to improve every day.

HAYES: Data warehousing is pushing the level of data access down to the end user, and that's where it has to be. The information doesn't do anyone any good if the end user can't get at it. As more users need more information, this whole issue is going to become more important. SN: Is creating a successful data warehouse something that can be accomplished in-house or are companies turning to third-party firms?

STEELE: I think much of this is being done in-house. Probably the leader in using information from a data warehouse standpoint has been Wal-Mart, which developed its own in-house proprietary system. There are some good solutions available that companies have been able to buy and modify to satisfy their needs. But much of this is being done in-house.