Skip to main content
The Cross-sell & Bundles report analyzes your order data to find products that customers frequently buy together. It uses product affinity metrics — co-occurrence, confidence, and lift — to surface the strongest cross-sell opportunities and help you build data-driven product bundles.
This report requires order data with multi-item orders. If most of your orders contain only one product, the data set for affinity analysis will be limited.

​
Key Metrics

  • Orders Analyzed — The total number of orders included in the affinity analysis for the selected period.
  • Multi-Item Orders — The number (and percentage) of orders containing two or more distinct products. This is the data set that powers affinity insights.
A low percentage of multi-item orders is not necessarily bad — it might mean your average product is a standalone purchase. But it does mean there is room to grow basket size through bundling and cross-sell recommendations.

​
Charts and Tables

​
Product Pair Affinity Table

The core of this report. A table listing product pairs that appear together in orders, with columns for:
  • Product A and Product B — The two products in the pair
  • Co-occurrence — How many times these products were bought together
  • Confidence — If a customer buys Product A, how likely are they to also buy Product B (expressed as a percentage)
  • Lift — How much more likely these products are to be bought together than you would expect by chance
Understanding these metrics in plain language:
  • Co-occurrence = 45 means these two products appeared in the same order 45 times.
  • Confidence = 32% means that 32% of customers who bought Product A also bought Product B.
  • Lift = 3.2 means this pair is bought together 3.2 times more often than random chance would predict. A lift above 1.0 indicates a meaningful affinity. The higher the lift, the stronger the connection.

​
What to Do with This Data

  • If a product pair has high lift (>2.0) and high co-occurrence — This is a strong bundle candidate. Create a product bundle or prominently feature the cross-sell on the product page.
  • If confidence is high in one direction but not the other — The relationship is one-directional. Show Product B as a recommendation on Product A’s page, but not necessarily the reverse.
  • If you see unexpected product pairs — Investigate why customers buy these together. You might discover a customer need you had not considered, which could inform new product development or merchandising.
  • If multi-item order percentage is low (<20%) — Implement cross-sell recommendations on product pages and in the cart to encourage larger baskets.

Product Performance

See individual product metrics to understand which items in a bundle drive the most value.

Revenue & Sales

Track how bundling strategies impact your overall AOV and revenue.

Category Insights

Explore cross-category affinities to find bundle opportunities across product lines.