How does the CMA generate Calculated Price per Unit, and how does that calculation differ from Average Price per Unit?
In Circana data, 1 “Unit” = one scan for whatever the UPC represents. So, if a UPC represents a 6pk, every time that 6pk scans it adds one “Unit” to the data.
Calculated Price per Unit finds the sum of total dollars for a UPC in whichever Geography is selected. Then it divides this number by the sum of total units sold. For example, say we’re looking at the numbers in one geography for a single UPC “X” within a three week span:
- In Week 1, UPC “X” had $75 in dollar sales and 16 units sold
- In Week 2 UPC “X” had $50 in dollar sales and 11 units sold
- In Week 3 UPC “X” had $25 in dollar sales and 6 units sold
Calculated Price per Unit then divides $150 total dollar sales by 33 total units sold, equaling $4.55
In contrast, Average Price per Unit first divides total dollars by total units for each week. For example:
- For Week 1, $75 would be divided by 16, equaling $4.69
- For Week 2, $50 would be divided by 11, equaling $4.55
- For Week 3, $25 would be divided by 6, equaling $4.17
Then, the sum of $4.69 + $4.55 + $4.17 ($13.41) is divided by the total number of weeks (3) equaling an Average Price per Unit of $4.47.
In summary, Calculated Price per Unit is calculating an average price, but only after adding up the numbers from all rows of data. This is achieved by collecting the sum of dollar sales from all rows of data and dividing this by the sum of units from all rows of data. In contrast, Average Price per Unit first calculates a price per unit for each row of data and then averages those numbers. This is achieved by dividing the sum of each row’s average price by the number of rows. These two measures only show slight differences, because they’re calculating along such similar lines.