Accurate scan data isn't at the top of most retailers' wish lists.
But generating much cleaner and more accurate scan data at the retail level could be the key to many of the industry's ambitious Efficient Consumer Response initiatives, including efficient promotions, efficient product introductions and continuous replenishment programs.
Until now, most retailers have been reluctant to take the steps necessary to improve scan data accuracy, such as cutting down on multiple key use at the front end, monitoring scanning practices closely and educating employees about the importance of clean data.
But the growing urgency to streamline the distribution system and work closely with
manufacturer and wholesaler partners is beginning to prompt retailers to place a greater emphasis on improved scanning practices.
And it couldn't come a moment too soon, industry sources said.
"Only about 85% of UPCs [universal product codes] are correctly read on the first pass over a scanner. That is the same as in the mid-1970s when point-of-sale scanning was introduced, even though there have been considerable technological advances in the scanners' ability to read product codes," said Sidney Pasoff, corporate vice president for management information systems at Oshawa Group, Etobicoke, Ontario.
"It is amazing to me that we haven't improved. What's going on here? Most technologies don't even last 20 years, and here we are still trying to figure out how to get this one to work," Pasoff said.
The need for retailers to be more aggressive about cleaning up their scan data, especially in the age of ECR partnering, was stressed by Paul Nicholson, vice president of finance and MIS at Pay Less Supermarkets, Anderson, Ind.
"ECR is still very much in its infancy. There's a lot of potential. But for the system to evolve in the long term, the scan data at the front end must be accurate enough to be passed through the distribution center to manufacturers to allow product to be pulled rather than pushed through the system," Nicholson said.
Clean scan data is crucial "to allow manufacturers to produce product based on what's selling, to ship product to distribution centers based on what's selling, and to allow distribution centers to ship product to stores based on what's selling as opposed to the current system," Nicholson said.
"Will anything cause the situation to improve?" Pasoff added. "The answer is yes. ECR is going to be the driver and enabler to cause us to change. Because one thing seems to be obvious: Highly accurate and complete scan data is necessary if ECR is going to succeed," he said.
Nicholson and Pasoff spoke on the topic of scan data accuracy at the Food Marketing Institute's MarkeTechnics convention in Tampa, Fla., last month. Along with Carlene Thissen, president of Retail Systems Consulting, Chicago, they led a workshop titled "Improve the Quality of Scan Data and Build a Cornerstone for ECR."
Thissen outlined several key areas involved in the process of improving scan data accuracy. "There are four basic elements when we are talking about scan data. One is the UPC code, or PLU, price look up. Another is individual product movement. Next is price, and then there is the total dollar amount sold for each item in the store. All are important when we talk about scan data," she said.
Nicholson of Pay Less, for his part, outlined his company's rationale for deciding to take dramatic steps to bolster, and profit from, scan data accuracy over the past several years.
According to Nicholson, accurate scan data is the key to building bottom-line profitability in today's increasingly automated and technology-driven retailing environment. Pay Less operates nine supermarkets in the range of 30,000 to 75,000 square feet. "In late 1970s and early 1980s, we experienced a series of productivity gains that were primarily a result of new technology, scanning, direct store delivery, new front-end systems and other things like that," Nicholson said. "But in the late 1980s those gains basically ended, and it was a struggle just to keep productivity flat, much less make gains," he said.
That flattening out of productivity, "combined with low inflation and a lot of sales pressure from alternative formats, such as the clubs, discount stores and superstores, made maintaining profitability a real struggle," Nicholson said.
"With that background, our company began in the early 1990s to take a serious look at using the scan data. Up to that point we had had lots of data, but both in our company and in the industry, the potential of that data had not been realized. People are still struggling with just getting started using the data," he said.
A close look at the state of Pay Less' scanning accuracy, though, quickly revealed major flaws in the data's accuracy, Nicholson said.
"We began to see we had an incredible amount of inaccuracy in the data we were collecting. Quite frankly, we had thought we were pretty good. We had thought the data was very accurate and that we were in good shape.
"But when we got into using it, we learned otherwise. We continue to be amazed at the difficulty of keeping accurate scan data. It is a constant struggle, but it is well worth it," he said. But the chain identified scanning practices that required improvement and has since seen major gains in both the quality and practical use of the data for improving a wide range of business practices.
Specifically, Nicholson cited 10 areas in which Pay Less Supermarkets is already using accurate scan data to implement sophisticated ECR programs today, or is planning to do so in the near future.
Among those areas are:
Computer-assisted ordering: "All our general merchandise and health and beauty care items are now on a computer-assisted ordering program. This requires the utmost accuracy of scan data, though," Nicholson said.
For the program to work, a perpetual inventory log must be kept at the front end in the computer system. The computer then puts together a suggested order, thus cutting dramatically the amount of time that buyers have to spend on each item, he said.
"We now only have to spend time on the very few items that for whatever reason might need to be looked at separately. It has reduced down to 5% to 10% the number of items we look at manually, rather than the 100% before."
Promotion management: "For a couple of years now we have been tracking item movement separately for all items on sale. Any movement report that we print shows regular movement and promotion movement of each item," Nicholson said.
"We calculate all markdowns for scanned items, and then look at a sale item early in the week to see what kind of movement we are having. Typically, the movement on Monday is a fixed percentage of the weekly movement. So with an item on sale, we can see on Monday evening what has sold and get a pretty good fix on how big the display should be the rest of the week. But that requires highly accurate scan data," he said.
Shelf and space management: "We installed a system a few years ago and were really struggling with the amount of data and maintenance required. We are just getting into using it now. But it is a good tool and we are going to be gaining quite a bit of benefit from it in the future. It also depends heavily on good accurate scan and movement data," Nicholson said.
Coupon scanning: "We've been scanning coupons in all our stores for a year now, including family code validating. You might ask what coupon scanning has to do with scan data accuracy," he said.
The answer, Nicholson said, is that when the family codes are being validated, and the coupon is scanned, the system will flag that coupon if the item was not purchased.
"If the cashier uses the multiple key or scans one item a bunch of times instead of each item individually, and then a coupon comes up for one of those items purchased but not scanned, the system will say the customer didn't buy it. So it is extremely important that the cashier discipline be there," he said.
Price integrity: "We work very hard on that, but once again the accuracy of your scanning data is of utmost importance," Nicholson said.
Automated checkout: "About three years ago we installed automated checkouts, which continue to work for us in most respects. But there is another whole level of data maintenance that goes along with automated checkout from the security file perspective," he said.
Pay Less also plans to use its scan data to implement several new programs before long. One of those areas involves new-item evaluation.
"We're working toward that. We're working with our wholesaler. We hope to get several major manufacturers involved to the point where they can retrieve movement data on new items and very quickly evaluate whether it is one of the majority of items that is not going to make it and therefore weed it out of the warehouse," Nicholson said.
"Or whether it is one of the minority of items that is going to make it and therefore get it permanently slotted in to support the retailers," he added.