Consumer packaged goods are small, non-durable items that are generally bought at the grocery store. These might include soda, breakfast cereal and laundry detergent. Like most retailers and manufacturers of consumers’ goods of any type, the consumer packaged goods industry is beginning to use micro-marketing techniques (Buxton 1993). Through using a sophisticated combination of databases and manipulation techniques, product marketing programmes can be developed at the retail chain level. In the US, almost all grocery stores use check-out scanners. These scanners read a bar-code on each product and automatically provide the price of the product. This is convenient for stores because they can change the price of the product (such as for a sale) without having to re-tag each item. Not only is this convenient for stores, but because of this process, a very rich database of what brand’s products are purchased at what stores is generated. These data are aggregated into approximately 50 “scanner markets” that cover various portions of the US. By combing these data with the demographics that tend to drive product demand, such as age and income, a buying power index (BPI) can be modelled for stores. The BPI indicates how much of various products could be sold at that store. By knowing a store’s BPI, the retailer can improve the product mix (micro-marketing) to “push through more product” – to use the industry jargon.