Monday, December 8, 2014

Basket Analysis I: What is Basket Analysis

This is planned to be the first blog in a series on Basket Analysis.  In this blog, I will try and answer the question, What is Basket Analysis?

Basket Analysis technology offers an exciting and innovative way of looking at retail data. Simple business intelligence will provide retailers with all sorts of analysis by individual products, including sales by cashiers, customers, profit contribution, waste, discounts, labour, time of day, weather,etc. All this analysis is essential for most retailers in today's market. However, more and more retailers are using the behaviour of shoppers to look for opportunities to improve their operation.

Basket Analysis is about analysing which sales items are bought together (in the one transaction, or different transactions for the same customer).   A Basket Analysis solution will enable you to query and report on which products are bought together and what the profit contribution of all the "baskets" (transactions) that involve a combination of products that are in, or out of, the transaction.

Basket Analysis is not a set of metrics or KPIs from which you can run your business. It is a data solution that allows you to ask questions about the product mix of transactions. Perhaps the most simple question is "Tell me information on the total transaction value by individual product items".   For every sales item, show me the
  • Item Average Price
  • Item Sales Quantity
  • Item Sales Value
  • Item Cost of Goods
  • Item Gross Profit %
  • Item Average Discount
  • Transaction Sales (total value of transactions including this item)
  • Transaction Count (number of transactions including this item)
  • Transaction Average Value
  • Transaction Average Quantity (average items/transactions)
  • Transaction Average Discount
  • Transaction Cost of Goods
  • Transaction Gross Profit
  • Transaction Gross Profit Percent
  • Transaction Value Multiple (total transaction sales/item sales)
Here is an illustration report of that very simple question that might be asked of a Basket Analysis solution.  It is simple, because the analysis is by individual items.  A more sophisticated query would be on particular combinations (existence, or non-existence) of items.  That is for subsequent blogs in this series.


From this simple report there are already many interesting insights available.  For example
  • Items that have a low sales item value but high transaction value (Transaction Value Multiple) are items that whilst they contribute very little profit themselves maybe bringing in customers who purchase more profitable items.
  • Items with a very low Transaction Value Multiplier, close to 1, are bought alone and probably do not induce customers to buy other products.
  • This simple report for one sales item, such as Cheese Cake, by waiter/cashier can give you insights as to which staff/outlets are good at on selling.
  • Sales items with the highest Transaction Gross Profit are ones that are likely to be important to your business, even though they may not represent well in top individual sales items.
For the next Basket Analysis post see Basket Analysis II.


No comments: