How it worksPricingLog InSign Up Free

What is upselling? Difference from Cross-sell

Definition

An upsell encourages a customer to choose a higher-value version, configuration, tier, or upgrade of the product they are already considering.

Upsell

Upsell

The defining idea is an upgrade path. If the customer is looking at a 128GB phone and chooses the 256GB version, that is an upsell. If the customer keeps the 128GB phone and adds a case, that is a cross-sell.

Not every more expensive recommendation is an upsell. The higher-priced option should still solve substantially the same underlying job while offering additional value that is relevant to the customer.

Upsell changes the primary choice

A simple test is: > If the recommendation succeeds, does the customer replace or upgrade the current choice rather than merely add another complementary item? Examples:

  • 128GB phone → 256GB phone: upsell
  • standard mattress → premium mattress: upsell
  • basic software-controlled appliance → model with additional capacity/features: upsell
  • phone → case: cross-sell
  • coffee machine → filters: cross-sell

Adobe Commerce's current product-relationship documentation similarly describes up-sells as products that may be higher quality, more popular, or offer better margin, while cross-sells are commonly surfaced as additional purchases. That implementation language is useful, but the durable distinction is upgrade versus complement.

The upgrade must explain its incremental value

A weak upsell says: "Spend $80 more." A strong upsell says why that $80 matters:

  • double the storage
  • longer warranty
  • larger capacity
  • stronger material
  • faster performance
  • features the customer has signaled they need

Price difference is only the cost of the upgrade. The merchandising job is to make the value difference legible.

Figure 1

A relevant upsell pairs a price step with a benefit step

Standard$180
  • Manual milk frothing
Plus$240, +$60
  • Automatic milk system
  • Buys convenience
Pro$340, +$100 more
  • Automatic milk + dual boiler
  • Buys throughput and control
Different relationship, not this ladder
Adding a case to a phone: cross-sell, not an upgrade
The best tier is customer-dependent. Promoting the most expensive option because it produces the most revenue is a poor upsell when the shopper has shown no need for the extra features.
Each tier up should explain what the extra money buys. "Spend $60 more" is not a reason; "automatic milk system" is.

Placement should support comparison

Upsells are often strongest while the customer is still comparing the primary product. Useful surfaces include:

  • variant or configuration selector
  • product comparison table
  • product detail page
  • plan/tier chooser for products with service components
  • cart, when the upgrade can be changed without restarting the buying process

Late-stage upsells can be disruptive if they force the customer to reopen decisions they had already completed. For example, replacing a carefully configured laptop from the cart with a different model can invalidate accessories, delivery estimates, or customization choices.

Upgrade paths can be variant-level

An upsell does not require a separate product page. Examples include:

  • larger size or capacity variant
  • premium material
  • extended configuration
  • higher quantity pack

But the merchant should not label every variant jump as an upsell. A different color at the same price is simply another choice. A larger apparel size is generally not a commercial upgrade merely because it is a different variant. The value must increase in a meaningful way.

Margin and revenue are not the same result

Suppose the Standard machine sells for $180 with $60 gross profit, while Plus sells for $240 with $92 gross profit. Moving 20 customers from Standard to Plus adds:

  • $60 revenue per upgraded order
  • $32 gross profit per upgraded order
  • $640 gross profit across 20 upgrades

That looks attractive, but measure the whole funnel. If an aggressive comparison causes 15 other shoppers to abandon because the store makes the Standard option look intentionally crippled, the net result can be worse. Upselling should improve incremental contribution without damaging base conversion.

Personalization can improve the upgrade path

Different customers value different improvements. A photographer may care about storage and battery life. A parent may care about durability. A business buyer may care about warranty or serviceability. Behavioral signals can help choose which upgrade benefit to emphasize, but a signal should not override the actual product relationship. The safest sequence is:

same customer job → meaningful higher-value option → relevant benefit → understandable price step.

Measure upgrades, not only AOV (average order value)

Useful metrics include:

  • upsell impression rate
  • upgrade selection rate
  • incremental revenue per eligible shopper
  • incremental gross profit per eligible shopper
  • base-product conversion rate
  • cart abandonment after upgrade prompt
  • cancellation and return rate by upgraded option

AOV can rise because only the highest-spending customers remain. That does not prove the upsell improved the business. Where possible, compare eligible shoppers with a credible control or experiment so the merchant can estimate incremental upgrade behavior.

Upsell versus cross-sell in one line

Upsell improves or expands the primary choice; cross-sell adds a complementary choice. Both are merchandising relationships. Keeping that distinction clear makes recommendation logic, placement, measurement, and customer expectations easier to reason about.

Worked example

Suppose a shopper is choosing among 3 coffee machines:

  • If the shopper is viewing Standard, promoting Plus can be a relevant upsell when convenience matters
  • 1. What extra outcome does each tier create?
  • 2. Which customer signals suggest that outcome matters?
  • 3. Is the price step understandable?
  • 4. Does the upgrade preserve the shopper's original job?

Promoting Pro only because it produces the most revenue can be a poor upsell if the shopper has shown no need for the features and the $160 jump creates decision anxiety. A useful ladder asks:

What passes and what does not

  • Tier design can make upgrades easier to understand, but manipulative patterns can backfire
  • Examples of poor upselling include:
  • hiding essential functionality until the shopper reaches the higher tier
  • making the base option artificially unusable
  • showing a premium option whose benefits have no connection to the shopper's task
  • switching the customer to a higher-priced configuration without clear confirmation
  • using fake scarcity to pressure the upgrade
  • An upgrade should be a choice with a clear tradeoff, not a trap

Comparison

A worked option ladder

OptionOptionPriceMain difference
StandardStandard$180manual milk frothing
PlusPlus$240automatic milk system
ProPro$340automatic milk + dual boiler
Try OnVoard free

Every app on every plan. Connect your store and switch on the flows in an evening.

Sign Up Free
Share
Commonly confused with

Sources

AdobeProduct relationshipsexperienceleague.adobe.com/en/docs/commerce-admin/marketing/promotions/product-relationships/product-relationships
OnVoardProduct Recommenderonvoard.com/product-recommender
ShopifyProduct recommendations with Shopify Search & Discoveryhelp.shopify.com/en/manual/online-store/storefront-search/search-and-discovery-recommendations