A cross-sell is an offer to purchase a different product that complements what a customer is already buying. While an upsell suggests a better or bigger version of the same product, a cross-sell suggests a related product from the same or adjacent category. Classic cross-sell examples in ecommerce include "Customers who bought this face serum also bought this SPF moisturizer," "Complete your routine with our toner," or "Add this travel bag to protect your supplements."
Cross-Sell Revenue = (Number of Cross-Sell Acceptances) × (Cross-Sell Product Price)
If 150 customers accept a ₹449 cross-sell product added to their cart from a cross-sell recommendation block:
150 × ₹449 = ₹67,350 in incremental revenue per period
Like upsells, this revenue is generated from an already-acquired customer — making it highly efficient on a CAC-adjusted basis.
Why Cross-Sell Matters for Ecommerce
Cross-selling increases AOV by adding incremental items to a transaction that the customer may not have sought out independently. For D2C brands with a multi-SKU catalog, cross-selling also helps expose customers to more of the product range — increasing the chance they find a second product they want to buy regularly, which has a compounding effect on customer lifetime value (CLV).
A skincare brand with an AOV of ₹600 for a single face wash can increase that to ₹1,000+ if a significant percentage of customers also add a serum or moisturizer to their order. The incremental margin on the second product is high because the shipping cost is already absorbed by the first product in the same order.
Cross-sells that are well-matched to the primary purchase also improve customer experience. A customer who buys a hair oil and is reminded to add a comb or shower cap has a more complete hair care routine — and a better outcome from the product — because of the cross-sell. Good cross-sells serve the customer's goal, not just the brand's revenue target.
Real-World Example
Bellavita cross-sells on their perfume product pages by showing a "Frequently Bought Together" section: the featured perfume paired with a matching body lotion and deodorant from the same scent range. The cross-sell makes sensory sense — customers who love a fragrance often want it reinforced by complementary products. The section is placed just above the "Add to Cart" button, making it visible before the primary purchase decision is complete. Their cart value for customers who accept at least one cross-sell is approximately 65% higher than for single-product purchases.
How to Improve / Optimize Cross-Sell
- Base cross-sells on real purchase data. "Frequently bought together" recommendations should reflect actual co-purchase patterns from your order history — not manual selections based on intuition. Data-driven cross-sells convert 2–3x better than random recommendations.
- Place cross-sells at the right moment. On the product page (below the CTA), in the cart drawer (before checkout), and on the order confirmation page (post-purchase) are the three primary cross-sell placements.
- Limit to 2–3 recommendations. Too many cross-sell suggestions creates choice paralysis. Two or three well-chosen complementary products outperform a generic "You might also like" grid of 12.
- Show the logic. "Complete your morning routine" or "Pairs well with" framing explains why the cross-sell is relevant — making it feel helpful rather than promotional.
- Bundle the cross-sell into a discounted kit. If the cross-sell naturally combines with the primary product, offer them as a bundle at a slight discount. This increases both acceptance rate and basket size.
Cross-Sell in A/B Testing
Cross-sell placement, recommendation logic, and framing are all testable. CustomFit.ai lets you experiment with whether data-driven "frequently bought together" recommendations outperform editorial "we recommend pairing this with" selections, and whether cross-sell placement in the cart drawer produces more incremental revenue than placement on the product page.
Run smarter A/B tests with CustomFit.ai — 14-day free trial, no credit card required.