bluebarry.
ReviewsPricingResources
Sign inBook demoBook a 15 min demo
PersonalizationConversion

Frequently Bought Together vs. Frequently Viewed Together

Last updated: September 14, 2026
JR
Jelmer Reitsma

Co-founder of bluebarry

Table of contents
Complements vs Substitutes: The Core Behavioral SplitStrategic Placement: Product Page vs Cart DrawerBasket Observations and Algorithmic Pair ChoicesCold-Start Handling and Thin Data Fallback TiersCart State Hygiene and Technical CompatibilityTesting Algorithmic Models and Merchant ControlsFrequently asked questions

Share this article:

Frequently bought together uses purchase patterns to suggest combinations. Frequently viewed together uses browsing patterns to suggest products worth comparing. Neither signal tells the whole story: products bought together may not be compatible, and products viewed together may serve different needs. Choose the strategy that fits the page and check the actual pairs.

Many merchants deploy these two recommendation models interchangeably or rely on a single automated widget across every page template. This setup misinterprets shopper intent. Presenting alternative substitutes inside a slide-out cart triggers hesitation and increases checkout abandonment, while displaying minor accessories to a shopper who is still evaluating core specifications creates distraction. Selecting the right model between co-purchase and co-view logic lets you align your product catalog with the exact stage of the customer decision journey.

Complements vs Substitutes: The Core Behavioral Split

In retail economics, products relate to each other either as complements or substitutes. A complement adds functional or aesthetic value to a primary item, creating an incremental sale. A substitute satisfies the same primary consumer need, representing an alternative choice rather than an additional purchase. Your recommendation engine must respect this split to generate relevant suggestions.

Frequently bought together models analyze transaction logs to detect items that frequently appear inside the same completed orders. When implemented effectively, this cross-selling approach lifts average order value without requiring shoppers to search through product categories. Understanding cross-selling and upselling strategies helps merchants distinguish between presenting a higher-tier replacement and bundling an essential accessory alongside the base item.

Frequently viewed together algorithms analyze collective session telemetry rather than order receipts. When multiple visitors repeatedly examine three distinct outdoor jackets during the same session, the system registers them as substitutes. These shoppers are comparing technical fabrics, fit profiles, or price points. Surfacing competing options during this phase helps the shopper complete their evaluation and pick a winner.

Strategic Placement: Product Page vs Cart Drawer

A shopper moves through distinct psychological phases as they navigate your store. During initial consideration on the product detail page, they seek assurance that they are choosing the right product. Once they click the add-to-cart button, their mindset switches to fulfillment and purchase finalization. Recommendation widgets must adapt to this transition.

Product Detail Page: Balancing Comparison and Add-Ons

The upper and middle sections of a product detail page benefit from co-view recommendations. If a visitor lands on a product that lacks their preferred color variant or falls outside their budget, alternative options keep them on site. Lower down the page, after the shopper has inspected specifications and imagery, a frequently bought together bundle module captures intent by offering complementary items before they proceed to checkout.

Slide-Out Cart and Checkout: Exclusive Cross-Sell Focus

The cart drawer must focus entirely on co-purchase logic. Surfacing alternative products here invites doubt, prompting the customer to reconsider their initial choice and delay checkout. Merchants running native Shopify product recommendations often discover that uncurated feeds blend related items with substitutes. High-converting cart drawers restrict suggestions to low-friction accessories, warranties, and consumable add-ons.

Basket Observations and Algorithmic Pair Choices

In this illustrative worked example, we examine how raw behavioral data from a specialty home espresso store translates into distinct recommendation outputs. The algorithm must differentiate between high-frequency joint purchases and exploratory browsing comparisons to select the appropriate widget.

Relying solely on co-occurrence numbers can lead to misleading pairings. If twenty shoppers examine two different espresso grinders before purchasing one, the raw correlation is high, yet bundling both into a single recommendation produces an absurd offer. The table below illustrates how behavioral inputs dictate merchandising actions.

Observed Behavioral SignalPrimary ProductCandidate Paired ProductRecommendation TypeMerchandising Decision
140 completed orders contain both items togetherEspresso MachinePrecision 58mm TamperFrequently Bought TogetherDisplay as one-click add-on bundle on product page and cart
85 shopper sessions view both items within 30 minutesCompact Espresso MachineDual-Boiler Espresso MachineFrequently Viewed TogetherDisplay in product page comparison carousel; exclude from cart
52 completed orders contain both items togetherBurr GrinderOrganic Grinder Cleaning TabletsFrequently Bought TogetherDisplay in slide-out cart drawer as low-consideration upsell
74 shopper sessions view both items; 0 joint ordersManual Hand GrinderElectric Burr GrinderFrequently Viewed TogetherDisplay as alternative option on product detail page only

Cold-Start Handling and Thin Data Fallback Tiers

Transactional recommendation engines struggle when catalogs contain new arrivals, niche items, or seasonal products with limited order history. When co-purchase data is sparse or absent, naive systems display empty modules or irrelevant storewide bestsellers that break customer trust.

A structured fallback hierarchy prevents these dead ends. When insufficient order records exist for a specific SKU, the engine defaults to category-level co-purchase patterns, followed by brand-affinity accessories, before relying on curated merchant rules. This sequence ensures that a newly launched trail running shoe displays compatible running socks and hydration flasks rather than an unrelated kitchen blender.

bluebarry connects product data, search behavior, and recommendation rules across the store. When shopper interaction data is limited, the platform utilizes shared catalog attributes and verified category relationships to suggest logical accessories, maintaining merchandising coherence across every collection.

Cart State Hygiene and Technical Compatibility

Displaying an accessory that the shopper has already placed in their cart wastes valuable screen space and signals poor site intelligence. Recommendation engines must monitor cart state in real time and suppress active items from frequently bought together widgets. When a customer adds a phone case to their basket, the widget should automatically recalculate and display a screen protector or charging cable instead.

Physical and technical compatibility rules are equally essential. In an illustrative electronics store, pairing an audio interface with a microphone requires checking connector standards and mounting hardware. Suggesting an incompatible mount creates friction, generates negative customer reviews, and increases support tickets.

Effective merchandising combines algorithmic pattern detection with explicit boundary rules. Merchants must retain control to enforce variant compatibility, suppress clearance items from cross-sell slots, and ensure that only in-stock SKUs with active inventory are presented to prospective buyers.

Testing Algorithmic Models and Merchant Controls

No single recommendation model delivers superior conversion across every retail vertical. Apparel retailers often see strong engagement from visual co-view carousels that aid style exploration, whereas replacement parts distributors generate higher revenue through precise co-purchase bundles.

Continuous split testing validates which model performs best in specific page locations. By measuring recommendation lift through controlled A/B experiments, merchants can isolate average order value improvements, click-through rates, and total revenue per visitor across competing widget configurations.

bluebarry enables merchants to test automated co-purchase and co-view suggestions against custom merchandising rules. With AI-assisted recommendations operating under clear merchant guardrails, you can optimize product pairings while preserving full authority over brand presentation and margin targets.

Frequently asked questions

Yes, provided they occupy different sections of the page. Place frequently viewed together carousels directly beneath primary product specifications to help shoppers compare alternatives. Position frequently bought together bundles lower on the page or adjacent to the main purchase button to capture buyers who are ready to complete their setup.

Most collaborative filtering algorithms require at least 15 to 30 joint transactions between two specific SKUs to establish statistically reliable recommendations. For products below this volume, configure attribute-based fallback rules to ensure that presented add-ons remain relevant.

Discounts are not mandatory to drive bundle conversion. When the complementary accessory provides clear functional utility, such as specialized cables for electronics or protective cases for eyewear, convenience alone motivates customers to add the bundle to their cart.

Boost Your Store Order Value with bluebarry

Integrate intelligent co-purchase and co-view recommendations that respect inventory availability, customer intent, and your specific merchandising rules.

Book a personal demo

Put this into practice

  • Product recommendations
  • Personalized product bundles

Keep reading

  • Shopify Product Recommendations: Related, Complementary, and Personalized
  • 9 product recommendation placements: examples for your store
  • Cross-Sell vs. Upsell: Which Offer Fits the Purchase?
  • Do Product Recommendations Increase Sales? How to Measure Incremental Lift

Further reading

  • Shopify Theme Product Recommendations Documentation

See bluebarry in your store

Bring your catalog and your next growth idea. We will show you how to connect them.

Book a demo
Table of contents
Complements vs Substitutes: The Core Behavioral SplitStrategic Placement: Product Page vs Cart DrawerBasket Observations and Algorithmic Pair ChoicesCold-Start Handling and Thin Data Fallback TiersCart State Hygiene and Technical CompatibilityTesting Algorithmic Models and Merchant ControlsFrequently asked questions

See bluebarry in your store

Bring your catalog and your next growth idea. We will show you how to connect them.

Book a demo
bluebarry.

Helping beauty, health and outdoor brands reach their dream AOV

[email protected]+31 6 57 16 10 87De Ried 10, 9285KK Buitenpost (The Netherlands)
Martijn DoumaStan van RooyAnco PostmaJelmer Reitsma

The guys that are increasing your AOV.

Platform
Quiz funnelsLanding pagesProduct QuizRecommendationsSearchIntegrationsPricing
Resources
CasesHelp centerFAQCompare
Company
Contact usPartners & affiliateReviewsRequest demo
Book a 15 min demo
© 2026 bluebarry. All rights reserved.
Privacy PolicyCookie PolicyTerms & Conditions
English/Nederlands/Deutsch
bluebarry