Skip to content
All articles
Cross-SellingUpsellingAverage Order ValueEmail MarketingE-CommerceReyo

Cross-Selling in E-Commerce: A Practical Guide

Cross-selling in e-commerce: which mechanics actually lift the basket, how to build recommendations that connect, and what German online shops really send.

By Reyo Editorial · · 11 min read

EnglishDeutsch
AI-generated 3D illustration: a blue shopping bag holding a sneaker, next to it a care spray and laces as the matching add-on, with a newsletter envelope behind

Cross-selling is the sale of the item that belongs with the first one. Not the pricier version, but the thing that goes with what is already in the basket: the case for the device, the care product for the shoe, the lamp for the table. It is the least glamorous revenue source in e-commerce and the only one that costs no media spend. Someone already at checkout does not have to be won again.

Yet in most shops it lives on the product detail page and stops there. In the newsletter, the surface you own, far less happens than you would expect.

The examples and market figures here come from newsletters sent by German online shops, captured by Reyo in the four weeks from 14 September to 11 October 2026. We do not name the shops: this is about the patterns, not the senders. And the limit of our data is clear: we see what was sent, not how it performed.

Cross-Selling, Upselling, Bundling: three levers, three goals

The three terms get used interchangeably, and the confusion produces emails that achieve nothing. They differ in what they change.

Lever What you offer What changes When it works
Cross-selling the complementary item items per order when the complement obviously belongs
Upselling the higher tier price per item before the decision, not after
Bundling the finished package both, at one price when the package beats the sum of parts

The practical rule follows: upselling belongs before the purchase, cross-selling after it. Push the pricier version after someone has ordered and you create regret. Show the matching accessory and you create a second order. Bundling does both at once, which is why it works best in email: a package needs no explanation, it needs a price.

Why email is the cheapest surface for it

On the product page your recommendation competes with the buy button. In the inbox it competes with nothing, and delivery costs no price per click. Three things make the email the better place for it:

  • You set the occasion. The shop waits for someone to arrive. The email arrives when the season, the consumption cycle or the gap since the last order makes it sensible.
  • You have room for the reason. Two sentences on why two products belong together will not fit into a shop widget. They fit into an email.
  • You can send after the purchase, when readiness for the complement is highest.

What German online shops actually send

Now the reality check against the inbox. Roughly one in three captured emails contained a product recommendation block at all. Almost a third of shops sent none in four weeks, and only a small minority used them in nearly every send. Where a block was present the median was five products: not a single complement, but a selection.

The most revealing finding sits in the combination with discounts. Emails without a price reduction carried a recommendation block considerably more often than promotional emails, a good four in ten against a good quarter. The email that wants to sell shows fewer products than the one that only informs. The discount crowds out the recommendation.

By assortment the spread is sharper. Fashion, shoes and accessories lead by a distance, with two thirds of those emails carrying recommendations. Electronics sits at the bottom at roughly one email in ten. That is remarkable, because electronics has the most obvious accessory business. Case, cable, memory card, warranty: nowhere is the complement more natural, and nowhere is it shown less often.

The mechanics the market uses

Thresholds by item count

  • A retailer with fashion and home ranges gave 15 percent from three items up. The voucher was embedded in the email, a toggle had to be activated in the basket, and no minimum order value applied.
  • A fashion retailer ran three for two across the whole shop for two days, cheapest item free, explicitly including reduced stock.
  • Another offered two for one over three days, limited to children's clothing.

How to read it: This is cross-selling in its purest form, because it does not cut the price, it makes the item count the condition. It leaves the choice open: whoever assembles three pieces decides which three. The waiver of a minimum order value in the first example matters. Without a value floor the mechanic also reaches the small basket, and that is where the third item is the biggest relative gain.

Thresholds by basket value

  • A home furnishing retailer tiered an instant discount by basket value: 20 euros from 100, 50 euros from 250, 100 euros from 500. The exclusion list was long, naming reduced stock, individual brands and whole product groups.
  • A health products retailer put three tiers into one email: 15 percent on everything, 25 percent from 125 euros, and from 170 euros 25 percent plus a gift.
  • A drinks shop included a glassware set from 50 euros basket value, for one day.

How to read it: The value threshold is the most common way to lift a basket and the riskiest. It rewards not the matching second item but any second item. Someone 50 euros short of the next tier adds whatever is closest to hand, not what fits best. So it lifts order value without improving relevance. Used properly it is only the frame, and the recommendation block answers what fills the gap. As orientation: the median stated minimum order value was just under 70 euros, the median free shipping threshold 50 euros. Derive your own from your median basket, not from the industry.

Gifts and sets

  • A baby products retailer added a bag worth close to 20 euros from three units up, with a second tier for a larger quantity.
  • A DIY chain included a battery starter set with the purchase of a tool, usable only within the same battery platform.
  • A supplements retailer gave 10 percent on four pre-assembled packages, another sold a set of three 10 percent below three single bottles.

How to read it: The gift is the lever with the best margin, because the perceived value of an add-on sits above its cost. The DIY example is the smartest in the field: the bundled battery set only works inside one system, which makes every further tool in that system more likely. That is cross-selling that continues by itself. The set of three shows the more honest route: quantity against price, understandable in one line.

The recommendation block

  • A home furnishing retailer assembled 24 articles in one email with no discount, from the wardrobe through bed frame, rug and floor lamp to the candle.
  • A bike retailer sent lights, mudguards and accessories for the autumn commute, also without a reduction.
  • A garden and forestry equipment retailer introduced chain oils and fuel and linked the matching accessories alongside.
  • A designer fashion retailer listed 16 brands, with no prices and no promotion.

How to read it: The first three are cross-selling, the last is not. The furniture retailer sells a room, not 24 articles. The bike retailer sells a safe commute in the dark. The equipment retailer sells the operation of a machine his customers already own. The fashion retailer sells a list. All four look identical in an analysis, because all four carry a block. The difference is the axis: room, occasion and consumption connect products, a brand list places them side by side.

That is the most commonly wasted opportunity in the German market. The block appears in every third email, but only about one in twenty names any relationship between the products, whether accessory, complement or combination. The block is there, the reason is missing.

The condition decides, not the percentage

One figure says more than any discount distribution: value thresholds appeared in roughly one email in ten and at almost a third of all shops, while thresholds by item count showed up in a handful across the whole period. And the most classic cross-selling mechanic of all, a discount on the second item, occurred in two emails in four weeks. One was a contact lens retailer giving 50 percent on the second pack, running to the end of January.

That is the gap. The German market steers the basket through euro thresholds and almost never through item count, even though item count is what cross-selling is about.

How to build it into your flows

Cross-selling rarely works as a campaign, because it presupposes a purchase. As an automated flow it almost always does. Four are enough to start:

  1. Post-purchase. Three to seven days after delivery, the complement to the item bought, with no discount. Best effort-to-return ratio of the four.
  2. Consumption. For consumables, a reminder after the typical usage period, with the repeat purchase and the matching accessory beside it.
  3. Basket below the threshold. When someone abandons a basket below your shipping threshold, do not show a discount, show three items that close the gap.
  4. Occasion and season. A change of season is the best reason for an accessory email that gives nothing away. The bike example above is exactly that.

Selecting the products needs no elaborate logic. The category of the last purchase is enough, and it sits in every shop system. How to build such flows is covered in our guide to marketing automation for e-commerce, and the steering behind them in the guide to CRM marketing for e-commerce.

Common mistakes

  • The recommendation is a variant. Three more jumpers next to a jumper is not a complement, it defers the decision. Cross-selling needs the other product group.
  • Everything comes with a discount. The complement needs a reason, not a reduction. Discount every recommendation and you teach customers to wait for the next promotion.
  • Too many items. Five per block is the market median, three is usually better. A selection of 24 works only with a visible axis, such as a room.
  • Only the value threshold. It lifts order value, not fit. Without a suggestion for what fills the gap, the nearest item goes in.
  • The timing comes from the send calendar. The post-purchase email belongs to the purchase date, not the editorial plan.

How Reyo helps

Your shop data tells you what your customers bought. It does not tell you which mechanics your competitors use to lift their baskets. Reyo supplies that half:

  • Competitor monitoring shows the conditions behind other shops' promotions: minimum order values, item thresholds, gifts and bundles, including the exclusions in the small print.
  • The campaign planner places your flows next to the market calendar, so your accessory email does not land in the week five competitors give 30 percent off everything.
  • Market predictions show from previous years when your assortment comes under promotional pressure, and therefore the weeks when complements carry further than percentages.

The dates for the current quarter are collected in the Q4 guide 2026. How deep the market discounts is covered in our guide to discount strategy for e-commerce.

Conclusion

Cross-selling is not a question of recommendation technology, it is one of reasoning. The market shows it clearly: blocks are widespread, the connection between products almost never stated, and baskets steered through euro thresholds instead of item counts.

Show the complement along an axis, whether room, occasion or consumption, and switch the condition from euros to units, and you are in a field German retail has barely occupied. Four automated flows and the category of the last purchase are enough to begin. That is less work than the next discount campaign, and it lasts longer.

Want to see which mechanics your competitors use to lift their baskets? Book a demo.


This article was published on 15 October 2026.

Reading about the market is good. Seeing your competitors live is better.

30 minutes, your competitors set up before the call, one recommendation to take away.