E-Commerce Bundles, Upsells and Cross-Selling Strategies

By Grocito17 October 2026

E-commerce bundles, upsells and cross-selling all aim to help a shopper find more of what they genuinely need in one order. An upsell suggests a better or larger version of the item being viewed, a cross-sell suggests a complementary item, and a bundle packages several items together. They work when the offer is relevant, honest and easy to decline, and when your stock and margins can support it.

Key takeaways

Relevance beats volume. A few complementary suggestions will do more good than a crowded row of random popular items.
Place offers where the shopper is still deciding, and keep the checkout itself calm and uncluttered.
Know your margins and stock levels before you promote anything, or a successful offer can quietly lose you money or oversell an item.
Never use pre-ticked add-ons, vague labels or hidden costs. Dark patterns win a sale and lose a customer.
Measure attach rate, average order value and the effect on conversion together, and change one thing at a time.

What is the difference between a bundle, an upsell and a cross-sell?

These terms get mixed up, so agree on them inside your team first.

Upsell: a higher-value alternative to the item in view, such as a larger pack size or a model with more features.
Cross-sell: a different product that goes with the one being bought, such as a case for a phone or a sauce for a pasta.
Add-on: a small extra attached to a specific item, often a service, such as gift wrapping, an extended warranty or installation.
Bundle: two or more products sold together as one offer, usually for a single price or a modest saving.
Frequently bought together: a suggestion block that shows items other shoppers often purchase with the current one, either chosen by hand or generated by the platform.

Each one answers a different shopper question. An upsell says "is there a better option?", a cross-sell says "what else will I need?", and a bundle says "can someone just sort this out for me?"

Where should you place offers on an online store?

Placement matters as much as the offer itself. A good offer in the wrong place feels like an interruption.

Product page. The best place for cross-sells and "frequently bought together" blocks, because the shopper is still forming the basket. Upsells work well here too, as a clear comparison between versions.
Cart. Suitable for small, low-effort additions that clearly complete what is already in the cart. Keep it to one or two suggestions.
Checkout. Handle with great care. The shopper has committed, and anything that slows the payment step risks the whole order. A single optional add-on that is obvious and easy to ignore is the most you should consider.
Post-purchase page. A confirmation or "thank you" page can offer an item to add to the order that has just been placed, provided your fulfilment process can still amend it.
Email and messages. Follow-up messages with consent can suggest refills, accessories or the next logical purchase. Our guide to email marketing for small businesses covers the basics of doing this respectfully.

Our article on conversion optimisation from product page to checkout explains why each step of the journey should have one clear job, which is a helpful test for any offer you plan to add.

How do you choose offers that are actually relevant?

Start from the customer's task, not from what you want to clear out of the warehouse.

Think in complements

Ask what a person buying this item will need to use it, protect it, refill it or enjoy it fully. Those are your cross-sell candidates. A random bestseller next to an unrelated product is noise.

Be margin aware

Rank candidate offers by how they affect your gross margin after discounts, shipping and handling, not just by how likely they are to be accepted. An offer that is accepted often but costs you money on every order is a hidden loss.

Respect stock

Promoting an item that is nearly out of stock, or one that ships from a different location and delays the whole order, creates disappointment. Connect your offers to live inventory where you can.

Keep the catalogue tidy

Good offers depend on clean product data: correct categories, consistent attributes and clear relationships between items. See our guide to product catalogue management for online stores for how to structure that data so suggestions stay accurate.

How should you design a bundle?

Bundles are the most flexible tool, and the easiest to get wrong.

Fixed or build-your-own?

Fixed bundles are curated sets such as a starter kit or a gift set. They are simple to explain, photograph and stock.
Build-your-own bundles let the shopper choose items from a defined group. They feel personal and suit consumables, but need clearer rules, such as how many items to choose and which items qualify.

Pricing logic

Describe the reason for the bundle price in words a shopper understands: a small saving for buying together, a free extra item, or simply the convenience of one ready-made set. Make sure the bundle price is never confusing against the price of the parts, and keep it fair enough that shoppers do not feel tricked when they compare.

Inventory complications

Bundles quietly create stock puzzles. If one component sells out separately, the bundle may become unavailable. Decide whether the bundle is a real product with its own stock count or a virtual set that draws from each component. Decide how returns will work if only part of a bundle comes back. Settle these before launch, not after the first awkward order.

How do you avoid annoying shoppers and dark patterns?

The aim is to be useful, not to trap people.

Honest labels: say what the add-on is, what it costs and whether it is optional.
No pre-ticked boxes: an add-on must be chosen by the shopper, not slipped in by default.
Easy to decline: the route to continue without the offer should be obvious, not hidden in small grey text.
No fake urgency: never invent countdowns or scarcity.
No surprise costs: the total must be visible before payment.

Beyond ethics, there is a business reason. Customers who feel manipulated return items, dispute charges and leave poor reviews. Consumer protection rules on this topic vary by country and change over time, so ask a qualified adviser if you are unsure about a specific practice.

Manual or algorithmic recommendations?

In plain words, manual recommendations are chosen by your team: "this charger goes with this phone." Algorithmic recommendations are generated from behaviour, such as what other shoppers viewed or bought together.

Manual suits small catalogues, new stores and products where fit really matters. It is accurate but takes effort to maintain.
Algorithmic suits larger catalogues and changing ranges. It scales, but needs enough data and occasional human checks so that odd pairings do not appear.
A hybrid works well for most stores: let the system suggest, then add merchandising rules, such as never recommending out-of-stock items, excluding low-margin items from promotions, or always showing a specific accessory for a key product.

How do you measure whether offers work?

Use a small set of measures and read them together, without chasing any one in isolation.

Attach rate: how often an offered item is added when it is shown.
Average order value: whether baskets are getting larger over time.
Conversion rate: whether the extra offers are quietly reducing the share of visitors who complete a purchase.
Returns and complaints: whether offered items come back more often than average.
Margin contribution: whether the extra revenue is actually profitable.

A higher average order value that comes with falling conversion may mean you are adding friction. Our guide to e-commerce analytics KPIs can help you set up the reporting.

How should you test offers methodically?

Change one thing at a time: the position, the item, the label or the incentive. Run the comparison long enough to cover normal weekly patterns, and decide in advance what result would make you keep the change. If your traffic is small, prefer simple before-and-after notes and customer feedback over elaborate experiments.

What mistakes should you avoid?

Recommending irrelevant items because they are bestsellers.
Launching bundles without sorting out stock, returns and reporting.

A worked example

Imagine a small online store that sells coffee equipment. Shoppers viewing a manual grinder see a cross-sell for a matching brush and a storage canister, chosen by hand. The cart offers one optional extra, a descaling kit, with a plain label and no pre-ticked box. The team also launches a fixed starter bundle containing a brewer, filters and a scoop, tracked as one product with its own stock count. After a month they find the bundle sells well, but the canister suggestion rarely gets picked, so they replace it with a second type of filter. They check returns and margin on every offer before expanding.

FAQ

What is the difference between upselling and cross-selling?

Upselling encourages a shopper to choose a better or larger version of the item they are looking at. Cross-selling suggests a different product that goes with it. Both should be relevant to what the person is trying to do.

Where should upsell offers appear in an online store?

Upsells work best on the product page, where shoppers are still comparing options, and sometimes in the cart as a small, clearly labelled suggestion. Keep the checkout itself as clean as possible so that the payment step is not slowed down.

Are product bundles good for profit?

They can be, but only when the bundle price still leaves a healthy margin after discounts, handling and shipping, and when your stock can support the combination. Check returns and fulfilment effort as well as revenue before deciding that a bundle is working.

How many product recommendations should I show?

Fewer, better suggestions generally work better than a crowded row of items. Start with a small number of clearly relevant products in each place and adjust based on how shoppers actually respond.

Next steps

Pick one product page and write down the three things a buyer will most likely need alongside it. Offer one of them, test it, and check the effect on conversion and margin before adding more. If you would like help planning merchandising rules, recommendation blocks or a storefront that supports them, take a look at our e-commerce storefront or contact us to talk it through.