Order bump vs upsell

An order bump is a checkbox on the checkout page. A one-click upsell is a full page that appears after the card has been charged. A downsell is what you show when the upsell is declined. Three slots in a sequence, three different psychological moments, and they are not interchangeable.

Most funnels we audit have one of the three, usually the wrong one for what they are selling. Here is what belongs in each slot.

The one-line difference

The pattern underneath: a bump removes a decision, an upsell asks for a bigger one, and a downsell reduces the size of a decision that has already been refused. Each slot has a different tolerance for friction, and that tolerance is what determines what you can put in it.

The order bump

A bump works when the answer is obvious before the customer has finished reading. That is the whole design constraint. If it needs a paragraph of persuasion, it is not a bump — it is an upsell in the wrong slot, and putting it there will lower your main conversion rate as well as failing on its own terms.

Practical rules that hold up across almost every offer:

• Price it well below the main item. A bump that is a large fraction of the order stops being an easy yes and starts being a second decision.
• It must complete the main purchase rather than compete with it. Extended warranty, priority scheduling, the consumable that the product needs, the rush option.
• Two or three lines of copy at most, with the price and the benefit both in the first line.
• One bump. Two checkboxes on a checkout is a menu, and menus create hesitation exactly where you least want it.
• Never pre-tick it. Charging for something the customer did not actively choose produces chargebacks and refunds that cost far more than the extra revenue.

In GoHighLevel the bump is configured on the order form step, not on the page. That trips people up when they duplicate a page and find the bump has not come with it — duplicate the step.

The one-click upsell

This is the page that appears after payment, offering something else on the card that was just charged, without asking for the details again. That mechanism is the whole reason it works: the friction of a second checkout is gone, and the customer is in the most agreeable state they will ever be in with you.

What belongs here is the logical next thing, not a random second product. Bought a single service? Offer the plan that covers the year. Bought the starter package? Offer the version with the part they will otherwise buy separately in three months. The test is whether a reasonable person, having just bought the first thing, would recognise the second as the obvious continuation.

Keep the page short. It is not a sales letter — the sale is already made and you are extending it. One headline, one paragraph on why now rather than later, the price, the accept button, and a plainly visible decline link.

Two constraints worth knowing before you design the sequence. One-click upsells need a payment provider that supports charging a stored card — Stripe and NMI do, PayPal does not. And the upsell step has to come after the order step in the same funnel; it cannot be a page on a website, which is one of the harder edges of the funnel versus website decision.

The downsell

Someone declined the upsell. A downsell offers less of the same thing at a lower price — three months instead of twelve, the core service without the extras, a payment plan instead of a single charge.

The rule that makes downsells work is that they must reduce the size of the commitment, not switch to a different product. Offering an unrelated item to someone who has just said no reads as pestering, and it costs you goodwill with a customer you already have.

And never punish the decline. The customer bought something. The exit from the downsell must be as easy and as unembarrassing as the exit from the upsell — one clear link that takes them to their confirmation.

Sequencing the whole thing

A complete paid funnel runs: sales page, order form with one bump, upsell, downsell if declined, confirmation. That is the maximum for most offers. Adding a second upsell after the first is where fatigue sets in and where refund rates start climbing, and the extra revenue rarely survives the support cost.

Where this sits in the bigger picture is covered in high converting funnel structure — this sequence is the third of the four shapes, and it belongs to offers the page itself can close.

Pricing the sequence

There is no universal ratio, and anyone quoting you one is quoting an average of businesses unlike yours. What there is, is a shape: the bump is a fraction of the main offer, the upsell is at least equal to it, the downsell is a portion of the upsell.

The number that decides whether the whole sequence was worth building is average order value — total revenue divided by number of orders, before and after. If you add a bump and average order value does not move, the bump is wrong, not underpriced. That is a measurement you take from your own orders over a defined window, and it is the only benchmark on this subject worth having. Setting your own baseline covers how to establish one properly.

Watch refund rate alongside it. A sequence that lifts average order value and doubles refunds has not made you money; it has made you work.

Where service businesses get this wrong

Most local service businesses do not sell online at all, so they conclude none of this applies. It does — it just moves. The bump becomes the option offered while booking: priority scheduling, the add-on treatment, the extra room. The upsell becomes the plan offered at completion, when the customer is happiest. The downsell becomes the smaller commitment offered when the plan is declined.

The mechanics are workflows rather than checkout pages, but the sequence is identical and the psychology is the same. Our offer and tripwire snapshots ship the on-page version, and every snapshot in the catalogue ships the post-completion version as part of its lifetime nurture layer — the car wash funnel snapshot is the clearest example, because a membership upsell at completion is the entire economics of that business.

The short version

Bump: obvious, cheap, one checkbox, never pre-ticked. Upsell: the logical next thing, full price, short page, easy decline. Downsell: less of the same, never something new. And what goes inside each of those offers so that it is worth buying at all is the offer stack.