Frequently Bought Together on Shopify

Frequently bought together (FBT) offers suggest products that make sense alongside the item a customer is considering. The useful question is simple: what would help with this purchase? Order history can inform the answer, but relevance, availability and the right variant still matter.

Related products versus useful companions

A bestseller list shows what is popular. A useful pairing explains what belongs with this particular purchase. A compatible accessory, useful consumable or complementary gift item can suit the original purchase. Sharing a category is a starting point, not proof of compatibility.

What purchase history can tell you

Products repeatedly appearing in the same orders can reveal pairings worth trying. Promotions, seasonality and small samples can also create misleading patterns. Two products bought together once are not enough to establish a reliable recommendation.

Recommendation systems may combine purchase history with product information and fallback rules. Their behaviour depends on the system and the data available. More time installed does not guarantee improvement if few eligible customers see offers or complete purchases.

Start sensibly when data is limited

Manual recommendations let you select suggested products yourself. Use them for a new product, limited order history or a deliberate selection. Choose a short list of available companions that fits the baskets where it appears. Check purpose, variant compatibility, price and margin. Review the actual suggestions customers receive, including new products and less common baskets. Our manual versus AI guide covers this decision.

Choose a placement customers encounter

Product-page offers can help customers discover a set. Cart offers reach people reviewing a basket. Neither placement guarantees better performance for every store. Check what shoppers actually see from collection quick-add, product-page Add to cart and the header cart, on mobile as well as desktop.

A cart offer should be easy to accept without losing required product options. Items needing personalisation may need a visit to the product page; see our personalised-product guide.

A discount is optional

Begin with a pairing worth buying at its stated price. A modest automatic discount can be a test if the margin allows, but discounting an irrelevant product will not make it useful. Specify eligible items and quantity, then confirm the final saving at checkout with any other active promotion.

Measure what happened

Track visible recommendation impressions, accepted adds and completed paid purchases. Compare products and placements using both counts and rates. One accepted suggestion from five impressions is encouraging, but too little evidence for a broad performance claim.

Keep free gifts separate from paid recommendations. Also distinguish attributed purchases from incremental revenue: a tracked item may be something the customer would have bought anyway.

Using Cart Uplift

Cart Uplift supports recommendations and discounted bundles in the shopping journey. Start with one clear offer and use available manual controls where you have a strong pairing in mind. Keep checking available variants and checkout behaviour as the catalogue changes.