Bread -> Butter and Beyond – Data Mining in Supermarkets
Have you ever visited a nearby Supermarket ? Whats next item placed near bread, its butter/milk…but why next item ? why it cannot be placed to somewhere else ? The reason is intuition, right, as soon as you pull out any item, our eyes looks to closest items nearby, and even if you don’t have butter in your buying list, you end up buying it, just because it can be used later with bread.
How data mining techniques workout here ?
The Supermarket owner has purchase records of all the transactions. He checks out what were the most frequent bought items ? He decides to put frequently bought items to be placed adjacent to each other so that customer doesn’t have to go looking for it. There are two terms which help decide which items should be placed nearer, and they are Support and Confidence. The first term tells how many transactions contain both bread and butter while the later one tells how often the Butter appears in transaction given bread. Observing these values helps owner understand customer purchase behavior.
This scenario is termed as Market Basket Analysis and it is achieved using a popular data mining technique called Association Rule Mining.
What are advantages of doing it ? Well of course your customers are happy as we have implemented customer focused strategy and maximized the profits.