What a downsell is

A downsell is an offer of a more affordable version of a product to a shopper who has refused the
original one or stalled over the price. The logic is simple: a smaller deal beats a deal that never
happens.

The technique completes a trio of instruments for working with order composition. The first two —
upsell and cross-sell — are widely known and appear in
every e-commerce handbook. The downsell is discussed much less often, because it intuitively looks
like turning money away. In practice it works on a different part of the funnel: not on people who
are already buying, but on people who are about to leave.

Upsell Cross-sell Downsell
What is offered A more expensive version of the same product A complementary product from an adjacent category A cheaper alternative for the same purpose
The goal Higher average order value A wider basket Saving the deal
The moment While the purchase decision is still open After the main product is chosen After a refusal signal or a price barrier
The main metric AOV Attach rate, AOV Conversion without losing gross profit
The main risk Scaring the shopper off with price Distracting from the main order Margin cannibalisation

The crucial difference is the moment. Upsell and cross-sell sit inside a positive scenario, a
downsell inside a negative one. Showing a lower-priced offer before any refusal signal has appeared
is a direct route to lost margin.

Scenarios in e-commerce

The item is out of stock. The safest scenario: the alternative does not compete with the main
product, because the main product simply is not there. Here the match runs on purpose and
specification rather than price as such — similar items filtered to
available to order.

The price is above expectations. The shopper never signals too expensive directly, but the
behavioural signs are readable: a long look at the product page without an add to cart, a return to
the listing and the application of a price filter, switching between several items at the bottom of
the category’s price range. Those signals accumulate in the
affinity profile and make it possible to show an alternative before
the shopper leaves the site.

An abandoned cart holding an expensive item. The classic point of return. An important
constraint: the personalization platform does not send emails or push messages itself — it supplies
a block that is embedded into the retailer’s own campaign or shown on the site at the next visit.

A subscription or renewal cancellation. Instead of a full cancellation — an offer of a cheaper
plan, a smaller volume or a longer delivery interval. Here the downsell saves not a single deal but
the whole future revenue stream from that customer, so it pays back best of all.

The risk of margin cannibalisation

This is the one genuinely dangerous side of the technique. Stated plainly: some of the shoppers who
see the lower-priced offer would have bought the expensive product anyway. For that group, the
business voluntarily lowers its revenue and gross profit.

Incremental effect = GP(test) - GP(control), per visitor,
where GP is gross profit

Cannibalisation is present if:
conversion(test) > conversion(control), but the incremental effect <= 0

The trap is that conversion under a downsell rises almost always. Looking at it alone, any mechanic
appears successful. That is why the decision is made on gross profit per visitor rather than on
conversion or order count.

⚠

The downsell is the only one of the three techniques that can improve every familiar funnel metric
while making the financial result worse. No version of it should be rolled out to 100% of traffic on
the strength of a conversion gain.

Hence the mandatory launch requirement: a holdout group — a share of traffic
that never sees the lower-priced offer at all. The comparison runs between that group and the test
group, not before and after.

How to launch it

  1. Define the refusal trigger. A downsell without a trigger is simply showing a cheap product to
    everyone. The minimum set: an out-of-stock item, an exit attempt from the cart, a subscription
    cancellation.
  2. Set the discount corridor. A working guideline is an alternative 20–40% cheaper than the
    original. A wider gap reads as a substitution: the shopper sees a different class of product and
    leaves.
  3. Preserve the purpose. The alternative has to solve the same problem. A $900 laptop is
    replaced by a $600 laptop, not by a tablet.
  4. Do not devalue the main product. Wording along the lines of this option is cheaper and just
    as good damages the whole upper half of the assortment. The right frame is what the shopper gets
    for less money, with no comparative judgement.
  5. Run an A/B test with a holdout group and decide on gross profit per
    visitor.
  6. Look at segments separately. A downsell often produces a positive incremental effect on new
    shoppers and a negative one on regulars who were buying expensive items anyway. In that case the
    mechanic is restricted to a segment rather than switched off entirely.

Common mistakes

  • Showing the lower-priced offer to everyone. Without a refusal trigger the technique turns into
    a discount for people who did not need one.
  • Judging it on conversion. Conversion will almost certainly rise; the thing to watch is gross
    profit.
  • Rolling it out without a holdout group. A before-and-after comparison does not separate the
    effect of the mechanic from seasonality and changes in traffic.
  • Offering an alternative from another category. A substitution by price without preserving
    purpose reads as an irrelevant block.
  • Placing the downsell next to the main product on the page. There it competes with the sale
    rather than rescuing it.