SCANNING AHEAD

MARC MILLSTEIN

Scan data is the driving force behind the emergence of a new age of promotional analysis and efficiency.

The promise of retailer-generated point-of-sale scan data is that brand marketers will be able to use increasingly specific sales information, on a monthly, daily or even more frequent basis, to build volume and boost promotional effectiveness.

Until now, the value of scan data has been limited, in part, by the ability to acquire clean and accurate enough information and analyze it. Common retailer practices, such as using a multiple key for some items at the checkout instead of scanning each item individually, have resulted in some data that failed to meet the information needs of brand marketers.

But the situation is changing. More manufacturers are seeking -- and are willing to pay for -- the high-quality scan data necessary to develop more sophisticated marketing programs. Retailers, in turn, are cooperating more fully on eliminating practices that harm data quality.

Also of critical importance, leading firms such as Nielsen Marketing Research, Information Resources Inc. and Efficient Market Services, along with a growing number of cooperative joint ventures, are aggressively stepping up efforts to provide brand marketers with precisely the scan data quality and detailed analysis they need.

The surging importance of high-quality scan data for implementing more sophisticated and efficient marketing practices was cited by several brand marketers and industry observers interviewed by Brand Marketing.

"The more information-based selling we can do, the better we can inform our customers and set better priorities. The better we understand our customers and provide data for them in a good format, the more successful we are going to be. So we are being very proactive in this field," said Mark Hannay, vice president of trade relations and national accounts at Stouffer Foods Corp., Solon, Ohio, a division of Glendale, Calif.-based Nestle U.S.A.

"The biggest challenge today is that everybody already has a large amount of data. But it is not all in a form or of a quality that is usable. The question is, how can we put this information in a practical, simple format and use it to set priorities? I would be kidding you if I told you we were all the way there yet, but we are making some good inroads," Hannay said.

"Better-quality scan data is a critical issue," said Jeff Hill, managing director of Meridian Consulting Group, a Westport, Conn.-based category management and trade marketing consulting firm.

"Most of our [packaged goods manufacturer] clients right now are going through the evaluation process on data quality. They are beginning to view data and data analysis as an increasing competitive advantage for them," he said.

By far the key issue in how quickly and how much brand marketers are able to make use of the critical POS information available from retailers today involves questions of heightened quality and enhanced in-depth analysis.

"There are several reasons why scan data today can be dirty. One is the checkout scanning procedures themselves; for instance, if the checkout clerk uses a multiple key for some items. Then there are also things that can happen internally within the system that cause problems," said Danny Moore, president of Efficient Market Services, Deerfield, Ill.

The situation, though, is improving, even in terms of checkout practices and the equipment itself, Moore said. "The retailer POS equipment is getting better, so it can handle and process the scan data a little better than it did in the past."

Michael Scroggie, president of Catalina Information Resources, Anaheim, Calif., cited several problems with scan data quality that are impairing the manufacturer's ability to make full use of the critical information.

"One problem is how retailers do their accounting close on the POS system controller and the impact that has on the scan data quality. The whole area of retailer involvement in scan data accuracy is a crusade we are now on," Scroggie said.

"Anybody concerned about using scan data to run a business, whether a retailer or a manufacturer, has to get down to looking at the scanning procedures at the front end of the supermarket," he said.

But considerable improvement is taking place, he stressed. "Some of the chains are becoming quite progressive to the point where they have disabled the quantity key. Other chains have said that the quantity key can't be used on less than six items."

Retailers, though, need to pay even more attention to this issue. Having accurate, clean scan data is a critical element in the entire Efficient Consumer Response initiative to automate many current industry practices and drive $30 billion in costs out of the system, Scroggie said.

"One of the big potential gains in the industry is the ability to minimize labor in the supermarket and have more automated procedures to reorder product. Well, guess what. That only works and is only as good as the quality of that information coming across the scanner," he said.

Stouffer's Hannay also emphasized the growing importance of high-quality scan data and analysis for developing successful programs to boost sales and form enhanced partnerships with retailer customers. It is also used to analyze market conditions and beat back potential competition.

Hannay stressed that the ability to make full use of scan data in today's market often hinges, in large part, on the use of third-party market research and analysis firms to clean, interpret and supply the critical information.

"Third-party firms are involved in cleaning the data for us. Very much so. The data goes from the retailers to the market research company we are working with. They then clear the data and they send it to us," Hannay said.

Hannay said Stouffer's use of scan data and other specific market demographic information is growing increasingly critical to successfully implementing the company's short-term and long-term sales and partnership strategies.

This type of information, for instance, "gives us the ability to look at stock levels and selling rates by individual store, or by clusters or neighborhoods. It gives us the ability to recognize out-of-stocks in a particular marketplace and develop better response procedures. We've been effective in terms of trying to use technology and information-based selling," Hannay said.

"So, for example, if we run a major promotion, we can now better recognize which stores we need to get in to assist with in terms of amount of product. This information can help in taking costs out of the system by allowing us to do a better job in product distribution on a store-by-store basis," he said.

"It gives us the ability to be much closer to our business. With a typical retail buyer responsible for probably a couple of thousand items, monitoring our few hundred items is just one piece of their business. Scan data allows us to manage the business better. It's beneficial to us and beneficial to them," he added.

Although everyone agrees on the importance of using cleaner scan data and analyzing it to implement better promotion and marketing programs, questions still remain about how quickly the industry can move forward in this area.

Meridian's Hill said many of the companies he's worked with are moving aggressively to explore the possibilities involved in acquiring and using scan data.

"They are beginning to exhaustively evaluate alternatives in the marketplace. They are looking for new data bases and application software packages to drive that data. The manufacturing community is going through a very considered analysis of the data bases themselves or going to an outside firm to gain information about what is available," Hill said.

"Probably the single biggest issue they are looking for is the ability to increase [scan data] accuracy at the store level," he said.

CIR's Scroggie said the next phase in the use of scan data information will involve both analysis and frequency. "The next phase, we feel, is going to be store-by-store census tracking and getting down to week-to-week or day-by-day analysis.

"Probably the next generation, looking out 10 years from now, might even be the ability to look at scan data on an hour-by-hour, lane-by-lane or even transaction-by-transaction basis," he said.

The use of the most sophisticated scan data applications "is in the evolutionary stage now. It is not just one or two manufacturers using the data now; it is in the 50 to 100 range now. But it is not in the thousands yet," Scroggie said.

"I think it will happen on a widescale basis, though, within the next two to five years. I would say that 1994 is going to be the pivotal year, and 1995 and beyond is where you will really begin to see mass or greater rollouts of this technology," he said.

Stouffer's Hannay said his company is now already involved in tests in which it is acquiring scan data information daily.

"In two markets, we are actually now working with daily scan data. But in our business, because we are not on a direct-store-delivery basis, it isn't the benefit of daily information as much as it is the ability to bring the data in and look at it based on merchandising, or advertising, weeks," Hannay said.

"We are able to look at exactly when and where a promotion ran, whereas before we would have to be looking at calendar weeks and time periods not exactly tied to when the performance occurred," he said.

Hannay, though, said making the promise of scan data for perfecting product sales and promotions a reality still requires considerably more effort and commitment on the part of retailers and manufacturers.

"There is already a commitment from the manufacturer and the retailer. But it is now going to grow as fast as the retailers want it to grow. If the retailers are not interested in working with the manufacturer to drive category sales and customer counts, then it doesn't make sense for a manufacturer to spend money and resources and time on this type of information to try to make it work," Hannay said.

"What is really needed is a commitment from both parties to try to drive that forward. You can't force it. If I have great data and the customer is not interested in hearing about it, then we are going to stop buying it. There's no point in it," he added.

"We buy the data to work with customers who are very interested in better promotions, better inventory control, better marketing and putting the right products in the right place at the right time."