SUPPORT SYSTEMS 1998-09-14

Decision-support systems, which allow retailers to pluck information they can act upon from oceans of data, are becoming a business necessity in the supermarket industry, say participants in SN's Executive Roundtable.

port is tempered by the demands such systems place on their own operations. These range from making sure the data going in are "clean" to ensuring the right people have access to the right decision-support tools -- and that they use them correctly.

SN: What are some of the current key benefits of decision-support systems, and what would you like to see them do in the future?

JACK SCOTT: All the major players are looking at decision-support systems, because they're asking how they can take scan data and market basket analysis, which comprise extremely large volumes of information, and effectively integrate that into their operations. They've been compelled by the competitive nature of companies like Wal-Mart, which has really taken decision support to an operational level, and used it to improve their in-store mix, pricing and service for the community.

The grocery industry has been a little slow in doing that compared to other retail markets, but I think we're catching up in a big way.

RON WALDBILLIG: As we get more involved in decision support we will be able to drill down further into our information. Today we can know what our customers buy and when they buy it. The next step is to find out why they wanted to purchase it. Were they happy with the shopping experience? And what did they miss during the shopping experience?

We need to create an "Issues of the Day" list that will take all of our stored information and condense it down to critical issues the store director needs to be aware of. This could include out-of-stocks, customer complaints, a bottleneck in receiving departments.

MIKE HUBERT: The industry is using decision support to monitor electronic item files and maintain gross margins. I also see it having a role in moving toward computer-assisted ordering, and people are starting to do this today. This has been a dream since scanning evolved.

SN: Are the tools that are needed to do decision support available at a cost-effective price?

JOHN GRANGER: I think the solutions are there. While the ability to store massive amounts of data is still expensive, it is less than it used to be, though at times it can still seem astronomical.

BOB SCHOENING: The technology is there, and with a reasonable effort on the part of an IT shop putting it together and testing it, they can get it to a point of providing a usable product.

Our concerns about hardware and software vendors, however, are that they are not doing an adequate job of testing their products in the many different environments that the supermarket industry would choose to run them. This leads to a lot of internal concerns about keeping the systems up, operating and available for use.

SN: What other pressures do the growth of decision-support systems put on supermarket IT departments?

HUBERT: People have come to expect immediate access to information, whereas in the past it was not unheard of to make requests for reports and not receive them for a couple of weeks. We expect data at our fingertips, and in today's world it should be there. The other thing is that there are more user-definable reports rather than one for all. Many companies are trying to find a more streamlined approach to generating those reports. GRANGER: Data protection and recovery ability, meaning the ability to get data back in the event of a problem is an issue. This can also refer to the risk of data being misinterpreted through the decision-support tools.

SCHOENING: The key issue is the data itself -- is it of a quality to render you able to make good decisions about your businesses? This applies to all the different sources of data. Most of us started by pulling in our point-of-sale data, and there are certainly some concerns with that.

Some supermarkets have excellent POS data, and most of that has to do with the discipline of the scanning function at the store level, along with the extent to which the items available in the supermarket also exist in the POS system's item file.

Other major data sources are a company's financial systems as well as various third-party data suppliers. Supermarkets also have to make sure that data is of high quality.

SCOTT: A broader issue is that wholesalers and retailer-owned cooperatives like Certified, along with independent retailers, traditionally have had the attitude 'You're my wholesaler, I'm the retailer, and I'm not going to share all this information with you.' But to really stay competitive and stay ahead, [both partners] need to get past the customer-vendor relationship and into partnership-alliance relationships, which allows information to be in a centralized hub.