colleagues white-boarding

Market Basket Analysis

Connect sales and customer data to measure the likelihood that your products are purchased together.

Understanding the buying patterns of existing customers is just as important as landing new ones. Market basket analysis explores additional customer needs to measure the likelihood that customers will buy certain products together.
image

Top-Line Growth

Automate product recommendation systems based on sales trends
Automation_icon

Bottom-Line Returns

Better meet demand for products based on products purchased
Image

Customer Experience

Improve buying experience for customers by showing products they’re likely to be interested in

Business Problem

Market basket analysis reveals which items buyers purchase together. Retailers use market basket analysis to understand the best way to co-locate products in both physical and digital stores. It shows them how to cross-sell and up-sell items that customers often put into their shopping carts at the same time.

It’s not difficult to infer relationships between a small number of products (itemset) and a small number of transactions. But as the quantity of products increases, it becomes necessary to study a quickly growing number of transactions to arrive at meaningful, statistically supported conclusions. And as promotions and recommendations come into play, data points from other sources become important. Spreadsheet models don’t suffice for high-volume retailers trying to build associations among hundreds or thousands of products.



Alteryx Solution

Market basket analysis is an analytic approach that includes variants like cross-selling and next-product analysis. It aims to describe the relationship between an “if” (antecedent) item and a “then” (consequent) item.

This analysis measures the frequency of the itemsets identified among all transactions, as well as the strength of the associations among those items. Analysts can apply metrics like support (proportion of transactions showing association), confidence (probability of purchasing the consequent, given the antecedent), and lift (strength of association between antecedent and consequent). With the Apriori and Eclat algorithms, they can apply association rule mining and pruning to use computational power efficiently and calculate metrics for many potential itemsets.

Alteryx Designer includes tools purpose-built for market basket analysis, including options for calculating support, confidence, and lift. All that’s needed is a dataset of transactions with the items included in each. With just two tools, analysts can generate association rules and view visualizations of those associations, including a network diagram displaying their relationships. Once the market basket analysis is complete, the results can be viewed in dashboarding tools such as Tableau.

Market Basket Analysis Workflow

1 - Data Access

Connect data sources containing sales records

2 - Data Prep

Apply rules algorithm to create product associations

3 - Automated Results

View results and export into Tableau

Additional Resources

electronic dashbaord

Starter Kit for Tableau

Alteryx intelligence suite tool icons

Starter Kit for Intelligence Suite

colleagues white-boarding

Customer Journey Analytics

Pie graph on phone and computer

Customer Segmentation

Transform Your Analytics

Get ready to unlock hidden insights in your data

Free Trial