In 2001, Amazon launched a feature that fundamentally changed ecommerce: "Frequently Bought Together." Powered by purchase history across millions of transactions, it showed shoppers that "customers who bought this drill also bought this drill bit set and this carrying case." The social proof of real purchase data made recommendations feel trustworthy rather than promotional.
Shopify stores don't have Amazon's transaction volume, but the concept still applies — and you can implement it effectively even with a modest order history.
Frequently Bought Together vs. general cross-sell recommendations
The distinction is in the data source:
- FBT (data-driven): Recommendations are based on actual co-purchase patterns. If 40% of people who bought Product A also bought Product B in the same order, that's a data signal. The recommendation is descriptive — "this is what real customers do."
- Cross-sell (curated): Recommendations are based on merchant judgment, product logic, or tag-based matching. The recommendation is prescriptive — "this is what we think you should also buy."
FBT recommendations tend to convert better when the data is strong (hundreds of co-purchase signals). Curated cross-sells tend to perform better for new products or slow-moving items that don't yet have co-purchase history.
Why Amazon's FBT works so well
Three psychological factors make FBT effective:
- Social proof: "Other customers bought this" is more persuasive than "we think you'd like this." Real behavior from real buyers is trusted more than merchant recommendations.
- Completeness: FBT often surfaces the items you "need" alongside the item you "want." Drill + drill bits. Camera + memory card. This completeness framing makes multi-item purchase feel pragmatic rather than impulsive.
- Efficiency: Adding multiple items in a single "add all" click removes the friction of hunting for accessories separately.
How to implement FBT on Shopify
You have three approaches, depending on your order volume:
For stores with 500+ orders/month:
Use an app that analyzes your actual order history for co-purchase patterns. These algorithms get better with more data. At 500+ orders/month, you'll have enough signal to generate meaningful FBT recommendations for your top 20–30 products within 2–3 months.
For stores with 50–500 orders/month:
Hybrid approach. Review your order history manually (Shopify makes this easy via the Orders export). Identify which products actually appear together in orders. Use those as your manual FBT assignments in a cross-sell app, with "FBT" framing in the widget copy.
For stores under 50 orders/month:
You don't have enough data for true FBT. Use curated cross-sell instead — it performs better at low data volumes because there's not enough purchase history to generate meaningful patterns. You can add the "FBT" label if the pairing is intuitive, but be honest: you're making an educated guess, not surfacing real purchase data.
The FBT widget anatomy
A well-designed FBT section typically includes:
- The main product (already selected, slightly grayed or checked)
- 1–3 recommended companion products with checkboxes
- Combined total price
- A single "Add all to cart" button
The checkbox format lets shoppers deselect items they don't want, which feels like control rather than a forced bundle. This flexibility usually converts better than a mandatory bundle at a slightly lower price.
FBT vs. "You may also like" vs. "Customers also viewed"
| Widget type | Data source | Best use case |
|---|---|---|
| Frequently Bought Together | Co-purchase history | Adding accessories/essentials |
| You may also like | Similarity / tags | Showing alternatives |
| Customers also viewed | Browse co-occurrence | Discovery / browsing |
Related reading
- Dropr vs Frequently Bought Together: Cart Drawer vs Amazon-Style Bundles
- The Shopify Cart Drawer Explained: What It Is, Which Themes Have It, and How to Add Cross-Sells
- Shopify Revenue Attribution Explained: What It Means for Cross-Sell Apps
- AI Product Recommendations for Shopify: What Actually Works in 2026
- Product Bundling on Shopify: The Honest Guide to What Works
FAQ
How many products should appear in a FBT section?
One to two companion products is the sweet spot. Three is the maximum. More than three creates decision fatigue and the "add all" price becomes intimidating. Amazon typically shows two, and their data supports that number.
Should I discount FBT bundles?
A small discount (5–10%) on the combined purchase can increase add rates. But it's not required — if the products genuinely belong together, shoppers will add them without a discount incentive. Test both and compare attributed revenue, not just click rate.
Does FBT work better than standard cross-sell?
When you have sufficient co-purchase data, yes — the social proof framing converts better. When you don't have the data, standard curated cross-sell with intentional product pairings performs similarly. The label matters less than the relevance of the recommendation.