Amazon brands often approach product bundling by asking a simple question: Which products would make sense together? That may seem like a reasonable starting point, but it leaves out the most important piece of information: what customers are already doing. The best Amazon bundle ideas often aren’t created in a brainstorming session. They’re hiding in existing purchase behavior.
Market basket analysis reveals which products customers frequently purchase together, helping brands identify natural product relationships and turn them into intentional Amazon bundling strategies. Instead of guessing which combinations shoppers might want, brands can use actual customer data to determine which virtual bundles, multipacks, gift sets, and cross-selling opportunities have the strongest potential.
For brands looking to increase Amazon average order value (AOV), basket size, and overall sales, market basket data can become one of the most useful inputs in bundle development.
What Is Market Basket Analysis on Amazon?
Market basket analysis examines the products customers purchase together within the same order.
For example, suppose a brand sells several varieties of tea. Market basket data might reveal that customers who purchase one particular black tea frequently purchase a complementary herbal tea at the same time. A skincare company might discover that customers purchasing a cleanser often add a specific moisturizer. A supplement company could find that two products designed for different needs nevertheless appear together in a meaningful percentage of customer orders.
These purchasing relationships matter because they show demonstrated customer demand. Rather than assuming Product A and Product B belong together because they are part of the same product line, market basket analysis can reveal whether customers actually behave that way. That distinction should influence how brands build their Amazon bundling strategies.
Why Guessing at Amazon Bundles Is Risky
There are plenty of reasons two products might appear to make a good bundle. They may have complementary uses. They may look good together. The brand may want to increase sales of one of the products. They might fall within the same category or target the same customer.
But none of those factors guarantees customers want to purchase them together. Creating a permanent multipack or physical bundle can involve inventory planning, packaging, fulfillment, forecasting, and other operational costs. If the bundle doesn’t resonate with shoppers, the brand may be left supporting an offer that generates little incremental value. Market basket data gives brands a stronger starting point. It allows them to ask:
- Which products are already being purchased together?
- How frequently does that combination occur?
- Which product tends to lead the purchase?
- Are customers combining complementary products or buying multiple varieties?
- Which combinations produce larger baskets?
- Are certain product pairings especially common during specific seasons or shopping events?
- Could an existing purchasing pattern support a dedicated bundle?
The answers can transform bundling from a merchandising hunch into a data-driven growth strategy.
Start With What Customers Already Put in Their Baskets
One of the most valuable uses of Amazon market basket analysis is identifying organic product affinities. If shoppers repeatedly purchase two ASINs together without being explicitly encouraged to do so, that behavior is a meaningful signal. The brand isn’t trying to manufacture a relationship between the products. Customers have already established one. Those combinations can become strong candidates for Amazon virtual bundles or future physical multipacks.
For example, market basket data might uncover purchasing patterns such as:
- Complementary products: Customers purchase two products that serve different but related purposes.
- Flavor or variety combinations: Shoppers frequently buy multiple flavors, scents, colors, or formulations together.
- Stock-up behavior: Customers purchase multiple related products because they want a larger supply.
- Routine-based combinations: Several products are commonly used as part of the same routine.
- Gift-oriented combinations: Products are frequently purchased together during holidays or other gifting periods.
Each behavior suggests a different potential bundling strategy.
The goal is not simply to find two products that appear in the same basket. Brands should determine why customers are buying them together and then design an offer that makes that behavior easier, more attractive, or more valuable.
Market Basket Data Can Reveal Bundles You Wouldn’t Expect
Some of the most valuable findings may be combinations the brand would never have chosen internally. Merchandising teams naturally think about products according to catalog structure: product line, flavor family, category, price tier, or intended use. Customers don’t necessarily shop according to those structures. Their baskets reflect actual needs.
That is why market basket analysis can uncover relationships that aren’t immediately obvious from the catalog itself. A lower-volume SKU, for instance, may appear disproportionately often alongside one of the brand’s best sellers. On its own, the product might not look like a priority. As part of a bundle, however, it could contribute to a larger and more valuable customer order.
This is an important shift in thinking.
The question isn’t just “Which products sell the most?“
It is also “Which products help customers buy more?“
Use Market Basket Data to Prioritize Virtual Bundle Testing
Finding a promising product relationship doesn’t necessarily mean a brand should immediately manufacture a new multipack. Amazon virtual bundles can provide an opportunity to test the concept first.
Virtual bundles allow eligible brands to merchandise complementary products together without physically packaging them as a new combined SKU. That makes them particularly useful for testing whether a market basket insight can be turned into an intentional offer.
If customers already purchase Product A and Product B together, a brand can test a virtual bundle featuring those products and evaluate how shoppers respond.
That can help answer questions such as:
- Does making the combination more visible increase adoption?
- Does the bundle increase units per transaction?
- Does it improve average order value?
- What discount, if any, encourages incremental purchases?
- Does the bundle attract new behavior or simply discount purchases customers would have made anyway?
- Does demand remain strong enough to justify a permanent multipack?
This creates a more disciplined path from data → hypothesis → test → investment. Instead of committing inventory and packaging resources based on an assumption, brands can validate the opportunity first.
Don’t Confuse Correlation With a Good Bundle
Market basket data is powerful, but brands still need to interpret it carefully. Two products appearing together doesn’t automatically mean they should become a bundle. The pairing should make sense from the customer’s perspective, and brands should consider the economics behind the combination.
A strong bundle candidate should generally satisfy several criteria: meaningful purchase affinity, logical customer value, sufficient demand, workable margins, operational feasibility, and the potential to increase the total value of the order.
Brands should also look at the size of the opportunity. A combination that appears in a high percentage of orders but involves very low-volume products may have limited impact. Conversely, even a modest affinity involving a high-volume hero ASIN could represent a substantial revenue opportunity.
Market basket data therefore shouldn’t be viewed in isolation. It should be considered alongside sales volume, profitability, advertising performance, seasonality, inventory availability, and broader customer behavior.
Look Beyond the Most Obvious Two-Product Combination
Market basket analysis can also help brands think beyond simple Product A + Product B bundles. Patterns across the catalog may reveal opportunities for:
Variety Packs
If customers regularly purchase several flavors or variations together, a curated variety pack may reduce friction and make product discovery easier.
Routine-Based Bundles
Products purchased sequentially or as part of the same routine can be grouped around a specific customer need rather than a product category.
Premium Bundles
If high-value customers frequently combine multiple premium products, the brand may have an opportunity to create a higher-priced offer designed specifically to increase AOV.
Seasonal and Gift Bundles
Market basket patterns around Prime Day, the holidays, Mother’s Day, back-to-school season, or other periods can reveal combinations that may not perform the same way year-round.
Multipacks
When customers repeatedly buy several related products together, that behavior may support the development of a permanent multipack.
The data can therefore influence not only which products belong together, but also what type of bundle the brand should create.
How Palais des Thés Used Market Basket Data to Guide Amazon Bundling
A strong example comes from WBX’s work with Palais des Thés, a French tea house specializing in high-end, sustainably sourced tea bags, loose-leaf teas, and gift sets.
The brand came to WBX with two important Amazon growth goals: increase Amazon channel sales and improve total basket size and average order value.
Rather than simply creating bundles based on assumptions about which teas might complement one another, WBX used the WBX Glass Market Basket report to determine the initial bundle recommendations.
During the July 2024 Prime Day event, the team launched five virtual bundles designed to increase shopper basket size and ultimately grow Amazon sales. The initial results provided a reason to expand the test.
Following that success, WBX explored additional bundle opportunities and discount levels. Fifteen more bundles were launched with a deeper discount during the 2024 Fall Prime Event and in the period leading up to Thanksgiving.
The process didn’t stop after a single successful test. WBX continued reviewing purchasing data, evaluating new bundle opportunities, and testing discount rates to determine which combinations were best suited for Amazon-exclusive launches. The results demonstrate why market basket data can be so valuable to a bundling strategy.
Palais des Thés increased total sales by 40% while achieving a 12% lift in average order value and a 3.5% year-over-year lift in basket size. The lesson isn’t simply that bundles work.
It’s that data should determine which bundles are worth testing, and testing should determine which bundles are worth scaling.
Discounts Should Be Tested, Not Assumed
Product selection is only one component of an effective Amazon bundle strategy. Pricing matters too.
Brands sometimes assume a bundle needs a substantial discount to attract customers. But an unnecessarily deep discount can increase unit volume while sacrificing profitability and may subsidize behavior that would have occurred anyway. Instead, brands should test different discount levels alongside bundle configurations.
The Palais des Thés strategy illustrates this iterative approach. After the initial virtual bundles generated positive results, WBX expanded the number of bundles and tested deeper discounts during subsequent high-traffic shopping periods. This allows brands to evaluate the relationship between:
Bundle discount → conversion → basket size → AOV → profitability
The goal should not be to create the cheapest possible bundle. It should be to determine the offer structure that generates meaningful incremental value for both the customer and the brand.
Your Best-Selling Product Isn’t Always Your Best Bundle Anchor
Another important implication of market basket analysis is that brands shouldn’t automatically build every bundle around their top-selling SKU.
A hero product may generate substantial standalone sales but have relatively weak cross-purchase behavior. Another product might generate fewer individual sales while consistently leading customers into larger baskets.
That second product can be extremely valuable.
Brands should therefore evaluate not only SKU-level revenue but also each product’s relationship to the rest of the catalog.
- Which products act as gateways to additional purchases?
- Which ASINs frequently appear in larger baskets?
- Which products introduce customers to additional categories?
- Which combinations lead to higher-value transactions?
Understanding these relationships can influence bundling, cross-selling, merchandising, and even Amazon advertising strategy.
Use Advertising to Support Proven Bundle Behavior
Market basket insights can also inform how brands promote their bundles.
Once a combination has demonstrated strong customer demand, advertising can help scale it. Brands might promote bundles around relevant category searches, retarget shoppers who have shown interest in one of the component products, or coordinate bundle promotions with Prime Day and other major Amazon shopping periods. The important distinction is that advertising comes after the brand understands the customer behavior it wants to amplify.
Rather than using advertising dollars to force an arbitrary combination onto shoppers, brands can put additional visibility behind a purchasing relationship customers have already demonstrated. That generally creates a stronger foundation for efficient growth.
Turn Bundle Testing Into an Ongoing Process
Customer purchasing behavior changes. New products launch. Seasonality affects demand. Advertising introduces customers to different parts of the catalog. Promotions change buying patterns. Competitors enter the market. For that reason, market basket analysis shouldn’t be a one-time exercise. Brands should continually review their data to identify:
- New product affinities
- Changes in existing product relationships
- Seasonal bundle opportunities
- High-value basket combinations
- Underperforming bundles
- Opportunities for new virtual bundle tests
- Successful tests that may justify permanent multipacks
The strongest Amazon bundling strategy is therefore not a fixed collection of product combinations. It is an ongoing testing program informed by evolving customer behavior.
Build Bundles Around Customer Behavior, Not Assumptions
Amazon already contains an enormous amount of information about how customers interact with your catalog. Market basket analysis turns part of that information into an actionable merchandising strategy. Instead of asking which products should go together, start by examining which products customers already choose together. Then test.
Use virtual bundles to validate demand. Experiment with discount levels. Measure the impact on basket size and average order value. Compare performance across seasonal events. Identify which combinations deserve greater investment and which don’t.
As the Palais des Thés case study demonstrates, this process can translate purchasing insights into measurable growth. By using market basket data to guide initial bundle recommendations and then expanding successful tests, WBX helped the brand achieve 40% sales growth, a 12% lift in AOV, and a 3.5% lift in basket size year over year.
For Amazon brands, the opportunity is clear: your customers may already be telling you what your next bundle should be. You just need the data to see it.
Find Your Next Amazon Bundle Opportunity With WBX
WBX helps brands turn Amazon customer behavior into actionable growth strategies. Through WBX Glass, market basket analysis, virtual bundle testing, advertising insights, and ongoing marketplace optimization, brands can identify product combinations with real customer demand and determine how to scale them profitably. Rather than guessing which products belong together, use the purchasing behavior already happening across your Amazon channel to guide the strategy.
Contact WBX to learn how market basket data can help uncover new opportunities to increase basket size, improve Amazon AOV, and drive incremental sales.