A hot topic of discussion lately on one of the forums I frequent has been JoAnn’s processing of buy one get one free items. It seems they have recently implemented a new computer system where all the lowest priced items are free in a buy one get one scenario. At first glance, this seems like a reasonable process. But, a closer look and one quickly realizes the flaw. If there are only have two items in the transaction, the lower priced item should be the free one. If there are multiple items, the distribution of discounts needs to change. Huh? Here’s an example.

Ann has filled her basket with buttons, which are part of the buy one get one free sale. Gold buttons are $10. Red buttons are $4. Blue buttons are $3. Yellow buttons are $2. Ann picks up two of each, expecting her total to be $19, as follows:
Gold buttons $10
Gold buttons $0
Red buttons $4
Red buttons $0
Blue buttons $3
Blue buttons $0
Yellow buttons $2
Yellow buttons $0

Total —–> $10+$4+$3+$2 = $19

Instead Ann is charged $28 when it is calculated this way:
Gold buttons $10
Gold buttons $10
Red buttons $4
Red buttons $4
Blue buttons $0
Blue buttons $0
Yellow buttons $0
Yellow buttons $0

Total —–> $10+$10+$4+$4 = $28

Ann questions the total and the cashier explains that is how the computer does it, but if she rings each set of buttons separately, Ann’s total will be $19. Of course Ann would prefer to use her $9 to get lunch on the way home from shopping, so the transaction is canceled and rung as four separate transactions. Ann leaves the store, having paid what she expected but is frustrated by her shopping experience.

There are several things wrong with this. (Or, this is why JoAnn should care about this problem.)

1) It’s a terrible customer experience.
You either pay more than you expected or suffer through multiple transactions.

2) It’s an inefficient use of store labor.
Cashiers are either ringing multiple transactions for one customer or ringing returns and new transactions after the original.

3) It’s more expensive to process all those transactions on several levels.
a) Ringing multiple credit card transactions is more expensive for the merchant than single transactions.
b) There is a higher amount of data needing to be transferred/kept on the servers. Even though they are smaller transactions, there are still many more of them and each one has transactional data attached to it.
c) Customers given 5 receipts for 10 items are more likely to lose the receipt they need (one of the five, undoubtedly) and need assistance.
d) see above re: store labor cost inefficiency

4) It distorts the store’s sales numbers. By having so many more transactions, it skews numbers computed about the average per transaction. The average transaction’s sales dollar total goes down, for example. The attachment rate of other products goes down – if Sue buys 10 b-g items and 3 associated items, the cashier rings 5 transactions. The first 4 transactions are only b-g items. The fifth is b-g and 3 other things. So, in quick data analysis, it looks like 4 people came into the store and only bought b-g items.