Executive Briefings

POS Data Is Used To Quickly Recalibrate Store Replenishments

Point of sale is "where the facts are" that help retailers readjust orders to accommodate sales peaks and valleys that may be predictable but aren't always intuitive.

Some retailers are beginning to routinely feed point-of-sale data back into demand-based planning, scheduling and replenishment systems in order to boost sales and increase profit. But this difficult exercise is itself not routine.

"A lot of people are talking. Fewer retailers are actually doing it," said Chris Verheuvel, vice president of retail for Manugistics of Rockville, Md.

And those retailers now using the POS data aren't fully leveraging it, said Kara Romanow, an analyst with AMR Research, a Boston consulting firm, who recently completed a study on the subject. "It is a daunting task for them," she said.

However, those pioneering retailers are forging ahead, turning their backs on conventional planning systems that rely strictly on historical sales data that merchants periodically tweak to reflect new, best-guess demand forecasts. Instead, they're more effectively adjusting replenishment plans based on analysis of updated POS data.

"POS is the source of all information within a retail enterprise," said Colin Hague, senior director of business development at Triversity Inc. of Toronto, a firm that helps retailers better organize and utilize such data.

"POS data is where the facts are," says Matt Johnson, chief technology officer at Waltham, Mass.-based Syncra Systems, a leading supplier of supply-chain planning and inventory replenishment software. Retailers about three years ago began using that data to drive more accurate replenishment systems, he said.

The goal, according to Johnson, is to translate unconstrained consumer demand into retail-store-ready/case-level orders or shipments that better reflect anticipated consumption, based on recent consumer behavior.

Concurrently, some consumer goods makers are using customer-provided POS data to build category-wide forecasts, which retailers use to maximize sales space. Manufacturers offer these forecasts gratis to retailers as a sales incentive to win new business.

A couple of grocery and drug stores are using POS data to "keep shoppers happy," Christine Overby, an analyst with Forrester Research, said in her 2003 year-end review report.

The value of both of these sales plans stems from the fact that they use POS data, the best building block for accurate forecasts at both the store and stocking-keeping unit or product level, said Ed Thompson, vice president of consumer goods and retail business optimization at i2 Technologies Inc. of Dallas. Stated another way, he said, "The most accurate representation of [planning] data is at the point of sales." i2 is a leading provider of constraint-based fulfillment and replenishment plans.

"A lot of people talk about using POS data in planning systems. Fewer retailers are actually doing it."
- Chris Verheuvel of Manugistics

Office equipment retailer OfficeMax, for example, feeds store-level POS data into Manugistics Networks Demand software to improve sales forecasts and establish optimum stocking/replenishment levels at each store. Number crunching is done at OfficeMax headquarters.

Retailers have several gauges to measure the effectiveness of these costly and time-consuming planning and replenishment optimization exercises. One frequent goal is to increase gross margin return on investment. The formula is gross margin/inventory value.

Stated another way, the retailer essentially seeks to maximize that ratio or increase the operating profit by selling more goods at higher prices, while holding less inventory.

Like in most major IT projects, maximizing usage of POS data comes down to clearly defining the scope of the project and providing adequate resources to handle it. To do so, i2 offers Master Data Manager.

This software, relying on customer-defined business rules, transforms raw historical POS data into a usable format and identifies gaps that will be addressed, such as unrecorded sales or an unaccounted-for decline due to extraordinary weather, when compiling a forecast. This is done so that other software modules subsequently can analyze the data in several different ways, or in multiple dimensions, Thompson said.

The goal is to increase the speed and efficiency of the subsequent number crunching, and rein in related processing costs.

Given those constraints, some retailers restrain the number of products being reviewed and whose data is properly formatted, and the time period for which sales data are studied. Being more selective limits the work of such i2 products as Demand Planner, a forecasting tool, and the Replenishment Planner. Both products can tap DB2 and Oracle databases.

CPG Manufacturers
However, consumer product goods manufacturers generally exercise less restraint, according to Syncra's Johnson. They generally request the complete data set for the category. "It's easier to end everything than be selective," he said.

To build demand for its expensive software, especially from medium-sized businesses, i2 offers a risk-sharing-based fee arrangement that's tied to increases in either gross margins or sales.

Smaller retailers often opt for less ambitious, less expensive packages like that provided by Demand Works Co. of West Chester, Pa. It targets mid-market CPG and industrial companies for its demand planning software.

The company's internet-enabled Demand Works DP, for example, uses demand history, current POS-based orders, promotions, events, and user judgment to determine an optimal estimate of future demand and required safety stocks.

Manufacturers building category-wide product management tools meld customer-provided POS data with generic category forecasts, based largely on data collected from such independent third-parties as IRI, NPD and Nielson.

Analysis packages offered by vendors such as JDA Software Group Inc. of Scottsdale, Ariz., for example, combine this sales information with the manufacturer's category-specific forecasts to produce customized and optimized sales allocation plans.

And, given the maturity of many of these consumer products, applying the software's recommendations often translates into a retailer stealing market share from a direct competitor.

Retailers use i2's Demand Planner to build a seasonal, baseline forecast. This tool allows retailers to make modifications to account for lost or unreported sales, a process called data smoothing.

Typically, the software meshes historical data, the retailer's view of the competitive environment and the current economic climate to build a forecast--"a demand-centric view of the world," says Thompson.

The replenishment tool gives retailers the opportunity to make quick, on-the-fly adjustments to sales floor inventory that reflect modifications to the original forecast. The software calculates the optimal stocking needs of the individual store or group of stores.

Retailers generally make weekly adjustments after incorporating the results of POS data analysis into a rolling eight-week forecast, Thompson said.

The software's inventory adjustment options and recommendations result from it comparing actual sales, as reflected in the most recent POS data, against the plan, and offering scenarios to better align the two.

Such modifications are designed to allow the retailer to quickly react to changes in the demand, usually by readjusting the number and types of items delivered each week from its distribution center or its suppliers. In turn, distribution centers usually adjust their inventory once each month so they're better equipped to handle changing store needs.

To better align that POS data so that its usage is transparent to the retail customer Yantra Corp. of Tewksbury, Mass., which provides supply-chain execution/warehouse management software is working closely with Triversity. The companies signed a non-exclusive partnership in January.

Single Point of Contact
Triversity ensures that the customer that generates the POS data has a single point of contact with a retailer that sources the product via several distribution channels. Basically, Triversity works to coordinate distribution efforts so that the customer doesn't have to negotiate corporate boundaries delineated by different distribution channels. That coordination provides the retailer operating efficiencies created by Yantra, the supply-chain execution partner.

As an example, Triversity essentially works to ensure that a customer doesn't have to deal with different requirements when seeking to receive an order, whether it's placed at a retail location, through mail order, or via a web site.

Triversity coordinates and aligns that POS data to consolidate or bundle inbound shipments from suppliers and outbound orders from the retailer's warehouse so that it enjoys cheaper, more efficient transportation services.

Applying similar discipline to the outbound leg, customers ideally can receive consolidated orders, regardless of where they were placed, at a single location. Essentially that allows the customer to pick up single-location orders - whether placed via the web site, at a retail location, or through a catalogue - so long as adequate storage space is available. Channel conflict, in terms of receipt, ideally becomes a non-issue.

JDA Software Group Inc.'s Efficient Assortment tool is designed to boost sales, says Graham Lewis, product manager. Results of the tool are shared with a retailer's buying or merchandising arm.

Essentially, consumer goods makers combine a targeted retailer's POS data with internally generated market data and internal analysis that forecasts the performance of a specific product line in that geographic area. The goal is to derive the optimal assortment among all products or "the best mix and variety," Lewis said.

Armed with this data, consumer product manufacturers have the requisite ammunition to win customer goodwill and, invariably, new business. Retailers generally appreciate suppliers taking the time and absorbing the expense to evaluate the product category. It's like the work-up that a stockbroker does on a prospective client's portfolio in order to win his confidence and ideally gain new business.

Primarily, frequent users of Efficient Assortment software are suppliers of fast-moving consumer goods. Generally they offer quarterly re-analysis of the customer's assortments.

Shelf Assortment applies the purchasing recommendations spelled out in Efficient Assortment, its sister tool, to the retailer's shelf. This software, of primary interest to store managers, helps the retailer to plan or determine optimum assortment on the shelf. The two products are used both discretely and together.

Category Captains
It should be no surprise that leading consumer product goods manufacturers also are the primary users of POS-friendly demand planning and sales optimization software, Thompson says. They purchase it to bolster sales, hoping to emerge as the retailer's preferred provider or the category captain, Thompson said.

And in the most advanced vendor-managed inventory (VMI) relationships, the supplier is responsible for managing shelf inventory and ensuring against stock-outs.

However, these suppliers rely on software, rather than route managers, to provide necessary coverage.

Syncra Software's Collaboration Suite and Syncra Xt, the replenishment module, rely on statistical forecasting algorithms that mesh a manufacturer's internal shipments data with customer-supplied data to design an optimal replenishment plan for a specific period, usually a week. Customer provided data used includes historic point-of-sales data dating back a year or two, information about current inventory positions, planned promotional events and internal forecasts.

The software then translates demand for that product back into the supply chain, ideally feeding an ERP system, said Johnson. It aggregates replenishment demand back to the supply point, usually a distribution center, building in lead time. It also pushes requirements further back into the manufacturing scheduling software, ideally taking into account economic production requirements.

Retailers like VMI programs because they relieve them of much of the responsibility, and related time and expense, of monitoring sales of individual items and handling re-supply. In exchange for providing that premium service, the vendor often emerges as the top supplier of that product, thereby boosting sales. The stronger vendor/supplier relationship also generally makes it more difficult for the incumbent supplier to be displaced by a competitor, new entrant or discounter.

In turn, the VMI supplier partner can use that retail POS sales data to better plan and manage its internal supply chain, manufacturing and logistics operations, thereby cutting cost and improving its operating profit. Generally it allows manufacturers to make goods more efficiently because they enjoy longer, higher-yield production runs.

Also, the supplier may be able to exact a slight premium for taking on the additional inventory management responsibilities.

However, such arrangements cut both ways, says Art Mesher, executive vice president of Descartes Systems of Waterloo, Ont., a transportation software vendor. Should a big retailer demand both closer collaboration and rock bottom prices, the supplier has a severe disadvantage, he said. That's because the low-margin order that absorbs much of the supplier's capacity also prevents it from seeking shorter-run, more opportunistic business, according to Mesher.

Compare Forecasts
Should the inventory re-supply responsibility be a shared one, this less frequently employed effort is called Collaborative Planning, Forecasting and Replenishment. In this scenario, the retailer offers the supplier its product POS data. Then supplier and retailers each generally create and compare individual forecasts that become the basis of discussion aimed at reaching agreement on the appropriate re-supply strategy or schedule.

"CPFR forces the supplier to think like a retailer," said i2's Thompson.

However, CPFR efforts are being undertaken by only a very limited number of retailers, who generally apply the discipline to less than a dozen products each, at most, said AMR's Romanow. Her recently completed report said most of these efforts are isolated and lessons learned somewhat limited because much of the data is stored and viewed in information "silos."

Profitability Based on Good POS Tools is No Pipedream
Using JDA Software Group Inc.'s Efficient Items Assortment and Space Planning POS-data products is a win-win proposition. So says, Joe Teller, the marketing manager of non-cigarette-related products for the American subsidiary of Swedish Match, based in Richmond, Va.

The large maker of a full line of tobacco products, other than cigarettes, credits JDA's software for allowing customers "to make the category (covered) bigger and more profitable," Teller says. In fact, Swedish Match is the largest single domestic supplier of this range of items.

It's not surprising then that JDA Software usage has also driven the significant market share gains that Swedish Match has registered among convenience stores, a major outlet for its products.

In fact, these tobacco products are the eighth-biggest category, in terms of sales, for convenience stores, excluding gasoline, according to Teller. Such second-tier status makes these products ideally suited for the free, value-add service that Swedish Match provides using JDA's two POS-based tools.

Essentially Swedish Match relies on the tools to win retailers' confidence, boost the satisfaction they provide shoppers and ultimately generate new or incremental business. This is how it works.

Swedish Match asks the retailer for complete POS data, dating back a year or two, for all the items in the "other than cigarette" category, including its own. They include: White Owl brand cigars; Timberwolf brand snuff; chewing, pipe and rolling tobaccos, and related accessories such as lighters and matches.

Ideally, that data addresses every item, for every store for every week, according to Teller.

Then the fun begins. Swedish Match loads retailer-specific POS data, cluster-specific third-party data it purchases and internal knowledge about the specific market into the JDA software to determine the optimum assortment for that location and similar ones.

The software, which now takes into account the limited retail floor space, ultimately will address shelf assortment should Swedish Match buy that JDA module, as planned, Teller says.

Invariably the analysis pinpoints premium items that a retailer lacks in its assortment. Or else the software determines which potentially big sellers that the vendor isn't giving proportionate space.

The simple solution for the retailer: change in the mix and test the results.

In every instance, Swedish Match has more than covered the cost of the exercise through increased sales. It also generates invaluable goodwill among customers.

Consider that the company ran information for 70 clusters last year and hopes to do almost double that amount in 2004, according to Teller.

Contributing to the success of the exercise is the fact that Swedish Match respects the confidentiality of the data it receives from retailers, many of them direct competitors, Teller says.

Some retailers are beginning to routinely feed point-of-sale data back into demand-based planning, scheduling and replenishment systems in order to boost sales and increase profit. But this difficult exercise is itself not routine.

"A lot of people are talking. Fewer retailers are actually doing it," said Chris Verheuvel, vice president of retail for Manugistics of Rockville, Md.

And those retailers now using the POS data aren't fully leveraging it, said Kara Romanow, an analyst with AMR Research, a Boston consulting firm, who recently completed a study on the subject. "It is a daunting task for them," she said.

However, those pioneering retailers are forging ahead, turning their backs on conventional planning systems that rely strictly on historical sales data that merchants periodically tweak to reflect new, best-guess demand forecasts. Instead, they're more effectively adjusting replenishment plans based on analysis of updated POS data.

"POS is the source of all information within a retail enterprise," said Colin Hague, senior director of business development at Triversity Inc. of Toronto, a firm that helps retailers better organize and utilize such data.

"POS data is where the facts are," says Matt Johnson, chief technology officer at Waltham, Mass.-based Syncra Systems, a leading supplier of supply-chain planning and inventory replenishment software. Retailers about three years ago began using that data to drive more accurate replenishment systems, he said.

The goal, according to Johnson, is to translate unconstrained consumer demand into retail-store-ready/case-level orders or shipments that better reflect anticipated consumption, based on recent consumer behavior.

Concurrently, some consumer goods makers are using customer-provided POS data to build category-wide forecasts, which retailers use to maximize sales space. Manufacturers offer these forecasts gratis to retailers as a sales incentive to win new business.

A couple of grocery and drug stores are using POS data to "keep shoppers happy," Christine Overby, an analyst with Forrester Research, said in her 2003 year-end review report.

The value of both of these sales plans stems from the fact that they use POS data, the best building block for accurate forecasts at both the store and stocking-keeping unit or product level, said Ed Thompson, vice president of consumer goods and retail business optimization at i2 Technologies Inc. of Dallas. Stated another way, he said, "The most accurate representation of [planning] data is at the point of sales." i2 is a leading provider of constraint-based fulfillment and replenishment plans.

"A lot of people talk about using POS data in planning systems. Fewer retailers are actually doing it."
- Chris Verheuvel of Manugistics

Office equipment retailer OfficeMax, for example, feeds store-level POS data into Manugistics Networks Demand software to improve sales forecasts and establish optimum stocking/replenishment levels at each store. Number crunching is done at OfficeMax headquarters.

Retailers have several gauges to measure the effectiveness of these costly and time-consuming planning and replenishment optimization exercises. One frequent goal is to increase gross margin return on investment. The formula is gross margin/inventory value.

Stated another way, the retailer essentially seeks to maximize that ratio or increase the operating profit by selling more goods at higher prices, while holding less inventory.

Like in most major IT projects, maximizing usage of POS data comes down to clearly defining the scope of the project and providing adequate resources to handle it. To do so, i2 offers Master Data Manager.

This software, relying on customer-defined business rules, transforms raw historical POS data into a usable format and identifies gaps that will be addressed, such as unrecorded sales or an unaccounted-for decline due to extraordinary weather, when compiling a forecast. This is done so that other software modules subsequently can analyze the data in several different ways, or in multiple dimensions, Thompson said.

The goal is to increase the speed and efficiency of the subsequent number crunching, and rein in related processing costs.

Given those constraints, some retailers restrain the number of products being reviewed and whose data is properly formatted, and the time period for which sales data are studied. Being more selective limits the work of such i2 products as Demand Planner, a forecasting tool, and the Replenishment Planner. Both products can tap DB2 and Oracle databases.

CPG Manufacturers
However, consumer product goods manufacturers generally exercise less restraint, according to Syncra's Johnson. They generally request the complete data set for the category. "It's easier to end everything than be selective," he said.

To build demand for its expensive software, especially from medium-sized businesses, i2 offers a risk-sharing-based fee arrangement that's tied to increases in either gross margins or sales.

Smaller retailers often opt for less ambitious, less expensive packages like that provided by Demand Works Co. of West Chester, Pa. It targets mid-market CPG and industrial companies for its demand planning software.

The company's internet-enabled Demand Works DP, for example, uses demand history, current POS-based orders, promotions, events, and user judgment to determine an optimal estimate of future demand and required safety stocks.

Manufacturers building category-wide product management tools meld customer-provided POS data with generic category forecasts, based largely on data collected from such independent third-parties as IRI, NPD and Nielson.

Analysis packages offered by vendors such as JDA Software Group Inc. of Scottsdale, Ariz., for example, combine this sales information with the manufacturer's category-specific forecasts to produce customized and optimized sales allocation plans.

And, given the maturity of many of these consumer products, applying the software's recommendations often translates into a retailer stealing market share from a direct competitor.

Retailers use i2's Demand Planner to build a seasonal, baseline forecast. This tool allows retailers to make modifications to account for lost or unreported sales, a process called data smoothing.

Typically, the software meshes historical data, the retailer's view of the competitive environment and the current economic climate to build a forecast--"a demand-centric view of the world," says Thompson.

The replenishment tool gives retailers the opportunity to make quick, on-the-fly adjustments to sales floor inventory that reflect modifications to the original forecast. The software calculates the optimal stocking needs of the individual store or group of stores.

Retailers generally make weekly adjustments after incorporating the results of POS data analysis into a rolling eight-week forecast, Thompson said.

The software's inventory adjustment options and recommendations result from it comparing actual sales, as reflected in the most recent POS data, against the plan, and offering scenarios to better align the two.

Such modifications are designed to allow the retailer to quickly react to changes in the demand, usually by readjusting the number and types of items delivered each week from its distribution center or its suppliers. In turn, distribution centers usually adjust their inventory once each month so they're better equipped to handle changing store needs.

To better align that POS data so that its usage is transparent to the retail customer Yantra Corp. of Tewksbury, Mass., which provides supply-chain execution/warehouse management software is working closely with Triversity. The companies signed a non-exclusive partnership in January.

Single Point of Contact
Triversity ensures that the customer that generates the POS data has a single point of contact with a retailer that sources the product via several distribution channels. Basically, Triversity works to coordinate distribution efforts so that the customer doesn't have to negotiate corporate boundaries delineated by different distribution channels. That coordination provides the retailer operating efficiencies created by Yantra, the supply-chain execution partner.

As an example, Triversity essentially works to ensure that a customer doesn't have to deal with different requirements when seeking to receive an order, whether it's placed at a retail location, through mail order, or via a web site.

Triversity coordinates and aligns that POS data to consolidate or bundle inbound shipments from suppliers and outbound orders from the retailer's warehouse so that it enjoys cheaper, more efficient transportation services.

Applying similar discipline to the outbound leg, customers ideally can receive consolidated orders, regardless of where they were placed, at a single location. Essentially that allows the customer to pick up single-location orders - whether placed via the web site, at a retail location, or through a catalogue - so long as adequate storage space is available. Channel conflict, in terms of receipt, ideally becomes a non-issue.

JDA Software Group Inc.'s Efficient Assortment tool is designed to boost sales, says Graham Lewis, product manager. Results of the tool are shared with a retailer's buying or merchandising arm.

Essentially, consumer goods makers combine a targeted retailer's POS data with internally generated market data and internal analysis that forecasts the performance of a specific product line in that geographic area. The goal is to derive the optimal assortment among all products or "the best mix and variety," Lewis said.

Armed with this data, consumer product manufacturers have the requisite ammunition to win customer goodwill and, invariably, new business. Retailers generally appreciate suppliers taking the time and absorbing the expense to evaluate the product category. It's like the work-up that a stockbroker does on a prospective client's portfolio in order to win his confidence and ideally gain new business.

Primarily, frequent users of Efficient Assortment software are suppliers of fast-moving consumer goods. Generally they offer quarterly re-analysis of the customer's assortments.

Shelf Assortment applies the purchasing recommendations spelled out in Efficient Assortment, its sister tool, to the retailer's shelf. This software, of primary interest to store managers, helps the retailer to plan or determine optimum assortment on the shelf. The two products are used both discretely and together.

Category Captains
It should be no surprise that leading consumer product goods manufacturers also are the primary users of POS-friendly demand planning and sales optimization software, Thompson says. They purchase it to bolster sales, hoping to emerge as the retailer's preferred provider or the category captain, Thompson said.

And in the most advanced vendor-managed inventory (VMI) relationships, the supplier is responsible for managing shelf inventory and ensuring against stock-outs.

However, these suppliers rely on software, rather than route managers, to provide necessary coverage.

Syncra Software's Collaboration Suite and Syncra Xt, the replenishment module, rely on statistical forecasting algorithms that mesh a manufacturer's internal shipments data with customer-supplied data to design an optimal replenishment plan for a specific period, usually a week. Customer provided data used includes historic point-of-sales data dating back a year or two, information about current inventory positions, planned promotional events and internal forecasts.

The software then translates demand for that product back into the supply chain, ideally feeding an ERP system, said Johnson. It aggregates replenishment demand back to the supply point, usually a distribution center, building in lead time. It also pushes requirements further back into the manufacturing scheduling software, ideally taking into account economic production requirements.

Retailers like VMI programs because they relieve them of much of the responsibility, and related time and expense, of monitoring sales of individual items and handling re-supply. In exchange for providing that premium service, the vendor often emerges as the top supplier of that product, thereby boosting sales. The stronger vendor/supplier relationship also generally makes it more difficult for the incumbent supplier to be displaced by a competitor, new entrant or discounter.

In turn, the VMI supplier partner can use that retail POS sales data to better plan and manage its internal supply chain, manufacturing and logistics operations, thereby cutting cost and improving its operating profit. Generally it allows manufacturers to make goods more efficiently because they enjoy longer, higher-yield production runs.

Also, the supplier may be able to exact a slight premium for taking on the additional inventory management responsibilities.

However, such arrangements cut both ways, says Art Mesher, executive vice president of Descartes Systems of Waterloo, Ont., a transportation software vendor. Should a big retailer demand both closer collaboration and rock bottom prices, the supplier has a severe disadvantage, he said. That's because the low-margin order that absorbs much of the supplier's capacity also prevents it from seeking shorter-run, more opportunistic business, according to Mesher.

Compare Forecasts
Should the inventory re-supply responsibility be a shared one, this less frequently employed effort is called Collaborative Planning, Forecasting and Replenishment. In this scenario, the retailer offers the supplier its product POS data. Then supplier and retailers each generally create and compare individual forecasts that become the basis of discussion aimed at reaching agreement on the appropriate re-supply strategy or schedule.

"CPFR forces the supplier to think like a retailer," said i2's Thompson.

However, CPFR efforts are being undertaken by only a very limited number of retailers, who generally apply the discipline to less than a dozen products each, at most, said AMR's Romanow. Her recently completed report said most of these efforts are isolated and lessons learned somewhat limited because much of the data is stored and viewed in information "silos."

Profitability Based on Good POS Tools is No Pipedream
Using JDA Software Group Inc.'s Efficient Items Assortment and Space Planning POS-data products is a win-win proposition. So says, Joe Teller, the marketing manager of non-cigarette-related products for the American subsidiary of Swedish Match, based in Richmond, Va.

The large maker of a full line of tobacco products, other than cigarettes, credits JDA's software for allowing customers "to make the category (covered) bigger and more profitable," Teller says. In fact, Swedish Match is the largest single domestic supplier of this range of items.

It's not surprising then that JDA Software usage has also driven the significant market share gains that Swedish Match has registered among convenience stores, a major outlet for its products.

In fact, these tobacco products are the eighth-biggest category, in terms of sales, for convenience stores, excluding gasoline, according to Teller. Such second-tier status makes these products ideally suited for the free, value-add service that Swedish Match provides using JDA's two POS-based tools.

Essentially Swedish Match relies on the tools to win retailers' confidence, boost the satisfaction they provide shoppers and ultimately generate new or incremental business. This is how it works.

Swedish Match asks the retailer for complete POS data, dating back a year or two, for all the items in the "other than cigarette" category, including its own. They include: White Owl brand cigars; Timberwolf brand snuff; chewing, pipe and rolling tobaccos, and related accessories such as lighters and matches.

Ideally, that data addresses every item, for every store for every week, according to Teller.

Then the fun begins. Swedish Match loads retailer-specific POS data, cluster-specific third-party data it purchases and internal knowledge about the specific market into the JDA software to determine the optimum assortment for that location and similar ones.

The software, which now takes into account the limited retail floor space, ultimately will address shelf assortment should Swedish Match buy that JDA module, as planned, Teller says.

Invariably the analysis pinpoints premium items that a retailer lacks in its assortment. Or else the software determines which potentially big sellers that the vendor isn't giving proportionate space.

The simple solution for the retailer: change in the mix and test the results.

In every instance, Swedish Match has more than covered the cost of the exercise through increased sales. It also generates invaluable goodwill among customers.

Consider that the company ran information for 70 clusters last year and hopes to do almost double that amount in 2004, according to Teller.

Contributing to the success of the exercise is the fact that Swedish Match respects the confidentiality of the data it receives from retailers, many of them direct competitors, Teller says.