Ecommerce & CRO
Cross-selling
Cross-selling is recommending complementary products alongside the one a shopper is buying — the case with the phone, the belt with the trousers. Where upselling trades up on the same item, cross-selling adds a second item. Both raise AOV, and both work best when the recommendation is genuinely relevant rather than merely popular.
Why it matters
Cross-selling works when it removes a future problem for the shopper. Someone buying a camera will need a memory card, and finding out after the parcel arrives is a worse experience than being offered one at the right moment. Done well, it is service that happens to raise AOV.
Done badly it is noise, and shoppers learn to ignore the block entirely within a few visits. The difference is relevance, and relevance is a data problem rather than a design one — which is why so many recommendation widgets look fine and earn nothing.
There is a second benefit beyond the immediate order. A shopper who buys the complementary item is more likely to use the product properly and more likely to return, which shows up later in repeat rate rather than in this week's numbers.
How it works on Shopify
There are three sources of recommendations, and they perform very differently.
Shopify's product recommendations API generates suggestions from order history and product relationships automatically, which is a reasonable default and requires no maintenance.
Manual curation through a metafield reference list is the most accurate: a merchandiser picks the genuinely complementary products for each item. It scales poorly past a few hundred products but wins on relevance where it is applied.
App-driven recommendations sit between the two, trading a subscription and some page weight for behavioural targeting.
Placement matters as much as source. Product page cross-sells work when the connection is obvious; cart cross-sells work for small, low-consideration additions; post-purchase works for anything that does not require thought. Collection pages are generally the wrong place — the shopper has not chosen anything to complement yet.
Bundles are the strong form: rather than suggesting the accessory, sell the set at a price that makes buying separately look careless.
Common mistakes
- Recommending popularity. Store bestsellers shown on every product page are not cross-sells; they are a bestseller list, and shoppers read them as one.
- Suggesting substitutes. Showing four similar shirts next to the shirt in the cart invites reconsideration rather than addition.
- Ignoring what is already in the cart. Recommending the item the shopper added two minutes ago is the fastest way to look automated.
- Too many options. Three relevant suggestions convert better than twelve.
- No stock check. Recommending an out-of-stock item wastes the slot and irritates.
- Set once, never reviewed. Catalogues change. Manual relationships decay unless someone owns them.
When you need help
The signal is a recommendation block with a click-through rate close to zero. That is a relevance problem, and the fix is usually a shift from automated to curated relationships for the products that matter most — the top hundred by revenue, not the whole catalogue.
The other case is bundling that needs real implementation: bundle pricing through Shopify Functions, inventory that has to decrement across components, and bundles that must not break discount rules. That is engineering, and getting it wrong shows up as either oversells or unprofitable orders.
Related terms
- UpsellingUpselling is prompting a shopper to buy a higher-value version of what they already intend to buy — a larger size, a premium tier, a longer warranty. It raises AOV without new traffic. Placement matters more than the offer: post-purchase upsells convert well precisely because they cannot jeopardise the original order.
- AOVAOV, or average order value, is the average amount a customer spends per order — total revenue divided by number of orders over the same period. It is one of the three levers on ecommerce revenue, alongside traffic and conversion rate. Raising AOV through bundling, upsells, or free-shipping thresholds is usually cheaper than buying more traffic.
- MerchandisingMerchandising, in ecommerce, is deciding what products a shopper sees and in what order — collection sort rules, search relevance, badging, and curated edits. It is the digital equivalent of a store layout. Good merchandising moves inventory that would otherwise sit, and it is one of the highest-leverage jobs a marketing team can do without a developer.