Equivalizing Size and UOM

Size and unit of measure (UOM) information from the data provider cannot always be used for the modeling process.

Size and unit of measure (UOM) information from the data provider cannot always be used for the modeling process. An understanding of consumer behavior is required to apply the correct size and unit of measure for modeling. For example, data on candy bars comes as follows:

• Hersheys almond bar 6 oz
• Milky Way bar 5.3 oz
• Three Musketeers bar 6.2 oz

The consumer is not paying for each candy bar by the ounce. Instead, the consumer is choosing a one count bar of candy. As such, the size should be 1, and the UOM should be count (CT).

In another example, data on laundry detergent comes as follows:

• Tide 128 ounce (32 loads), $9.99
• Tide Ultra 64 ounce (32 loads), $9.99
• Tide 64 ounce (16 loads), $5.99

The consumer is not paying for the ounces in the package. Instead, the consumer is paying for the number of loads completed by each product. By changing the size and unit of measure to the number of loads, the products can be equivalized, and the size/UOM relative to the price makes more sense.

Two UOMs within a Demand Group is not preferred, but is sometimes necessary. For example, nail polish remover has both ounce and count UOMs. If the distribution of products in each UOM is equitable, then this is not necessarily a problem. The modeling process will assign a unique equivalent value to each UOM. If one UOM value does not have a sufficient sampling, however, the median price of that UOM might be skewed.