📋 Overview
Your best-selling product may point to demand for a companion item, but sales totals alone cannot tell you what shoppers purchase together. Amazon’s Market Basket Analysis report in Brand Analytics helps eligible brand owners find those connections. This guide explains what its percentages mean, how to evaluate a product pairing, and how to turn a pattern into a controlled merchandising or assortment test.
🔎 What a Product Pairing Actually Tells You
Market Basket Analysis identifies products purchased together with your products. For a selected reporting period, it shows the top three purchased-with products for a reported ASIN, along with their combination percentages.
Your anchor ASIN is the product you are investigating; a companion ASIN is a product reported alongside it. These are working terms for your analysis, not additional Amazon report fields.
The denominator excludes single-item orders
The combination percentage measures orders containing both the anchor and companion divided by orders containing the anchor and at least one other different item, expressed as a percentage.
In other words, the denominator is the anchor’s qualifying multi-item orders, not every order containing the anchor. An order containing only the anchor does not enter that denominator.
This distinction changes your decision. You cannot read a combination percentage as “this share of all anchor buyers also bought the companion,” or multiply it by all anchor orders to estimate companion demand.
The percentage also does not tell you how often the anchor appears in multi-item orders overall. A strong pairing within qualifying baskets can coexist with many orders that contain only the anchor.
The relationship is directional
Reading product B alongside anchor A answers a different question from reading A alongside anchor B. Each anchor has its own qualifying-order denominator, so you should not assume the percentages are equal in both directions.
A widely purchased accessory may appear beside a specialized tool, while that tool represents only a small part of the accessory’s broader purchasing patterns. Treat each direction as a separate merchandising hypothesis.
A shared basket is evidence of association
The report establishes that products appeared in the same purchase, not that one caused the other to sell. Shoppers might be completing a project, buying gifts, replenishing unrelated supplies, or responding to a temporary promotion.
The companion can also belong to another brand. Use that result to understand the shopper’s task; it is not evidence that you should copy the product or that your own alternative will perform the same way.
🧭 Build a Shortlist in Seller Central
1. Open Market Basket Analysis in Brand Analytics
In Seller Central, open Brand Analytics and locate Market Basket Analysis. Access depends on your eligibility as a brand representative for a brand enrolled in Amazon Brand Registry, along with the appropriate account permissions.
If you cannot see the report, check your brand relationship and user permissions before assuming the report contains no data. Confirm availability for your marketplace in Seller Central’s help information.
A useful starting point is an anchor you already have a business reason to investigate: a product with an obvious accessory need, an item supporting a planned assortment expansion, or a product whose buyers often ask about compatibility.
2. Select a completed reporting period
In Market Basket Analysis, select an available completed period and find your anchor ASIN. Record the marketplace, period, and exact ASIN so you can repeat the comparison without mixing different scopes.
Choose a period that fits the question. For an evergreen accessory decision, avoid making a holiday-heavy period your only evidence. For a seasonal pairing, compare it with a relevant seasonal period rather than assuming it applies throughout the year.
Use the same reporting-period type for comparisons. A broader period can combine different shopping occasions and hide changes that matter to your decision.
3. Record the reported companions without filling in gaps
For each anchor in Market Basket Analysis, capture the displayed companion ASINs, their order in the report, and their combination percentages. Keep the values attached to their original anchor and period.
Build a simple working sheet with these columns:
- Anchor ASIN and product: The exact product you are evaluating.
- Companion ASIN and product: The reported pairing, identified at ASIN level.
- Reporting scope: Marketplace and selected period.
- Combination percentage: The displayed value, with a note that its denominator excludes single-item orders.
- Likely shopping task: Installation, storage, maintenance, gifting, or another plausible purpose.
- Decision to investigate: Merchandising, compatibility information, assortment research, or no action.
The report is a ranked shortlist, not a complete basket export. A product missing from the displayed companions is not proof that shoppers never buy it with your anchor.
4. Inspect each companion’s product detail page
Open the companion ASIN’s product detail page and identify what shoppers actually bought. Check its function, dimensions, included components, and compatibility rather than relying on a shortened product name.
Classify the relationship before choosing a tactic:
- Functional complement: Helps shoppers use, maintain, install, or store the anchor.
- Shared occasion: Serves the same event or project without directly interacting with the anchor.
- Alternative or comparison purchase: May reflect shoppers buying options to compare rather than needing both.
- Unclear connection: Does not yet support a specific action.
Do not infer compatibility from co-purchase alone. If fit depends on a connector, material, size, or model, verify it from reliable product specifications before making a claim.
5. Check whether the pairing persists
Return to Market Basket Analysis and repeat the review for comparable completed periods. Note whether the same companion remains visible, whether its position changes, and whether the percentage moves alongside known promotions or availability changes.
A stable connection gives you a better reason to investigate than an isolated appearance. It still does not establish incremental demand or provide a sales forecast.
Keep the shortlist small enough to investigate properly. Prioritize pairings with a clear shopping task, verified product fit, and an action you can actually execute.
⚖️ Match the Pairing to a Test You Can Measure
Use existing products before committing to new inventory
If you already sell a genuinely suitable companion, test presenting the products together in your Brand Store, using an appropriate page or section organized around the shopper’s task. Describe each product accurately and make clear what each purchase includes.
If the reported companion belongs to another brand, first investigate the need it fills. Review whether your existing catalog serves that need before treating the result as a sourcing instruction.
For a possible new product, write down the required compatibility, likely use case, expected costs, and operational constraints. The basket report supports the research question; it does not establish enough demand to justify a purchase order by itself.
Choose the success measure before changing merchandising
Define the specific behavior you want to change. For a Store pairing, that might be increased companion sales without weakening the anchor’s contribution margin or creating compatibility-related returns.
Record your baseline, the date of the change, and other events that could affect the result. Compare matched periods after the change, keeping price, promotions, and availability as consistent as practical.
Use subsequent basket reports as supporting evidence, not as your only success measure. The combination percentage can increase because the mix of qualifying orders changed, even when total companion sales did not increase.
Likewise, stronger companion sales do not prove your merchandising caused the increase. Without a controlled comparison, describe the result as directional evidence and check the economics before expanding the test.
Keep the offer separate from the observed relationship
A co-purchase pattern does not mean you have a bundle. If you decide to create one, check Amazon’s applicable bundling requirements separately rather than treating the report as approval.
When showing complementary products, do not imply that an accessory is included with the anchor if it is sold separately. Keep product detail page content factual and about the product itself, without seller-specific promotions or links.
Do not place orders or arrange purchases to manufacture a co-purchase pattern. Test through legitimate merchandising and genuine customer demand.
❓ FAQs
Does Market Basket Analysis show who bought the products?
No. It provides aggregated product-pairing information, not a customer contact list. You cannot use it to identify or message the shoppers behind a pairing.
Can I recreate the report from my own order exports?
Not completely. Your seller order data does not reveal all items purchased from other sellers in the same customer basket. You can study co-purchases within your own orders, but label that analysis separately and define its denominator.
Can I choose what appears in Frequently bought together?
Market Basket Analysis is a reporting tool, not a control for Amazon’s Frequently bought together recommendations. A reported pairing does not guarantee that Amazon will display those products together on a detail page.
Does the report show which product the shopper added first?
No. The reported pairing does not establish add-to-cart sequence. Do not use it to claim that the anchor initiated the shopping trip or that the companion was an impulse purchase.