Upselling is offering a customer a higher-priced or upgraded version of what they're already buying. A coffee shop offering a large instead of a medium is upselling. A SaaS prompt to switch from a $29 plan to the $99 tier is upselling. Done well, it raises what each buyer spends without you having to find more buyers.
For digital sellers, the three strategies that lift average order value (AOV) most are post-purchase one-click upsells, order bumps at checkout, and tiered pricing inside the checkout. Below, seven strategies, when to use each, and how to measure them.
What is upselling?
Upselling is a sales tactic where you offer the customer a higher-priced or upgraded version of the product they're already buying. The upgrade can be a larger size, a higher plan tier, an annual term instead of monthly, or a version with more access or support. The buyer has already decided to buy; the upsell changes what they buy, not whether they buy.
That last part is what separates upselling from most other revenue tactics. The offer lands when the buying decision has already been made, which is why it tends to convert far better than any offer made to cold traffic.
At the largest scale, this is how Amazon operates: McKinsey's retail research attributed 35% of what consumers purchase on Amazon to algorithm-driven product recommendations, much of it in the form of upsells and cross-sells. You don't need a recommendation engine to get the benefit, though. For digital sellers, one or two well-matched offers in the checkout flow do the same job on a small scale.
What is upselling vs cross-selling vs order bumps
These three tactics get mixed up constantly, though they occur at different points in the buying flow and serve different purposes.
Tactic | What it is | When to use | Expected AOV impact | Use this if… |
|---|---|---|---|---|
Upselling | A higher-priced or upgraded version of the product the buyer is already choosing | Pre-purchase, in-checkout, or post-purchase | Medium to high (varies by industry) | You have a clear "better" version of your core product, and the upgrade reason is obvious to the buyer |
Cross-selling | A complementary product offered alongside the main purchase | Pre-purchase or at checkout | Low to medium (varies by industry) | You sell several products that work together and you can name the pairing in one sentence |
Order bump | A pre-payment tickbox that adds a low-cost item before the buyer pays | In-checkout, before payment is confirmed | Low to medium per bump, compounding across volume | You have a small, high-margin add-on under roughly 25% of the cart total |
In short: upselling sells the same buyer a bigger version of what they're already choosing, whereas cross-selling sells them something complementary. Order bumps are a delivery mechanism (a tickbox at checkout) that can be used for either, though in practice they're most often used for low-cost cross-sells.
A buyer who chooses a $19 monthly subscription and sees an “annual plan, $190/yr (2 months free)” toggle is being upsold. The same buyer seeing a "tick to add the workbook for $9" box is taking an order bump.
Seeing "people who buy this also buy the templates pack" is being cross-sold, and all three can run in the same checkout flow.
7 upselling strategies that work
Each strategy below covers what it is, when to use it, a real example, and a rough AOV lift. If a strategy doesn't fit your product, skip it. An irrelevant offer at checkout hurts conversion faster than it lifts AOV.
1. Post-purchase one-click upsells
A one-click upsell is an offer shown immediately after the buyer has paid for the original product. The card is already charged, the payment method is stored, and accepting the upsell adds a second charge with a single click, with no new checkout and no re-entered card details.
When to use it: When you can finish this sentence in one go: "If you liked X, you'll want Y because…". Works well for digital products, courses, coaching, and subscriptions where Y is the obvious next thing. If you can't name Y in one sentence, skip the upsell.
Real example: A creator selling a $49 Notion templates pack offers a one-click upsell to a $79 full template library after payment. The buyer just paid for Notion templates; the only friction left is one click.
Expected AOV lift: Take-rates on well-matched post-purchase upsells generally land between 10% and 25%. The numbers move with industry, offer relevance, and price ratio. The strongest take-rates come when the upsell is roughly 50-100% of the original price and obviously related to it.
Checkout Page take: This is the single tactic that most consistently lifts AOV across our Stripe sellers. Stripe doesn't support post-purchase one-click upsells out of the box. Setup and mechanics live in our one-click upsells on Stripe guide.
2. Order bumps at checkout
An order bump is a small add-on offered with a tickbox during checkout, before the buyer pays. The buyer adds it (or doesn't) with one click and pays for everything in a single transaction. Order bumps work on pre-payment, which is where they differ from one-click upsells.
When to use it: When you have a small, high-margin add-on at roughly 10-25% of the cart total. Bumps are tick-the-box decisions, not considered purchases. A $9 add-on next to a $49 cart works. A $49 add-on next to the same cart usually doesn't.
Real example: A course seller offers a $19 workbook as a bump on a $97 course. About one in five buyers ticks the box, which adds roughly $4 per transaction (20% × $19) without changing the original price or the checkout flow.
Expected AOV lift: Take-rates on well-matched bumps tend to run 15-30%. Smaller than a post-purchase upsell in absolute terms, but the bump runs on every transaction, so it adds up. Bumps that feel forced or unrelated fall to single digits and start dragging down the main product's conversion.
Checkout Page take: Order bumps are standard on Checkout Page and can run alongside a post-purchase one-click upsell. Bump at checkout, upsell after payment; same buyer, two different commitment moments.
3. Tiered pricing displayed at checkout
Tiered pricing shows two or three versions of the same product side by side at the point of purchase, with the middle or higher tier styled as the recommended option. The buyer's default frame shifts from "should I buy this?" to "which one should I buy?".
When to use it: When you have a real feature ladder (basic, standard, pro) and three or four bullets can describe the gap between tiers. If the tiers only differ in price, the selector won't help.
Real example: ConvertKit's checkout shows three plans side-by-side, with Creator Pro pre-selected and monthly and annual prices both visible. The annual toggle puts the saving on the page, and Pro carries the "recommended" styling. It works because the differences between tiers (advanced reporting, deliverability, priority support) sit next to the price as a scannable list.
Expected AOV lift: Highlighting a middle or top tier moves buyers up the ladder compared to a single-product checkout. How much depends on how distinct the tiers feel side-by-side. Pull the number from your own pre/post data; industry averages will mislead you.
Checkout Page take: Tier selectors are a checkout setting, not a separate product. You configure them in the checkout builder.
4. Bundle upsells
A bundle upsell offers two or more related products together at a price below the sum of their separate prices. The bundle can appear pre-purchase (as the headline offer next to the individual product), at checkout (as an upgrade option), or post-purchase (as a one-click upsell from the single product to the full bundle).
When to use it: When you sell several products to the same buyer, that buyer would plausibly want more than one, and the bundle is meaningfully cheaper than buying the parts. Each component needs to hold up on its own. A "bundle" of one solid product and three filler extras reads as padding and costs you trust.
Real example: A photography educator selling a $99 Lightroom presets pack offers a $179 bundle: the presets, a 60-minute editing course, and a follow-up Q&A recording. Buyers can switch to the bundle in one click before paying, and the bundle saves about 25% versus buying the parts.
Roughly a third of buyers who'd otherwise have taken only the presets choose the bundle.
Expected AOV lift: A bundle priced at 1.5-2× the entry product, with each component named and described, tends to lift AOV by 20-40%. The range is wide because the quality of the underlying offers varies so much.
Checkout Page take: Bundles can run as a tier toggle in the checkout (basic product vs bundle) or as a one-click upsell after the basic product is paid for. The right shape depends on whether your buyers decide upfront or after committing.
5. Subscription and payment-plan upgrades
If you sell anything recurring, your upsell is built into the billing: monthly to annual, basic plan to pro, one-time payment to payment plan. The most common version is the monthly-to-annual prompt. A buyer on a $19/month plan is offered $190/year, which works out to two months free. The pricing math is the upsell.
When to use it: When you sell anything recurring and your buyers are on the monthly plan. The upgrade prompt fits at checkout (as a toggle), inside the customer portal (as a one-click switch), or as a one-click post-purchase upsell. Avoid it for new buyers who haven't used the monthly version yet.
Real example: ConvertKit's checkout has an annual/monthly toggle with the annual saving spelled out. A buyer who'd planned to pay monthly sees that the annual price equates to about 10 months, and many switch before completing the purchase.
Expected AOV lift: Moving even 15-20% of monthly subscribers to annual lifts blended ARPU noticeably, because annual buyers pay upfront and tend to churn less. See our guide on how to sell payment plans for the recurring-billing mechanics.
Checkout Page take: Upgrade prompts can run in the initial checkout (annual/monthly toggle) or as a one-click post-purchase upsell after the monthly version is paid for. The post-purchase version works because the buyer has just said yes once; the second yes is easier than it'll ever be.
6. Premium add-ons (faster delivery, concierge, priority access)
Premium add-ons are paid extras that upgrade the service around the product, not the product itself. Faster shipping, concierge onboarding, priority support, expedited access. They work because the buyer has already chosen the product and is now choosing how much attention they get with it.
When to use it: When your product has a service or experience layer worth upgrading, and a decent share of buyers would pay for the better version. Skip it when the "premium" tier feels like the regular product with a fresh label.
Real example: A course creator selling a $299 cohort-based course offers a $99 add-on at checkout: two 30-minute coaching calls with the instructor during the course. About 20% of buyers add it. The upsell isn't more content; it's more access.
Expected AOV lift: Add-ons priced at 25-40% of the core product that solve a real problem (more help, faster access) tend to convert at 15-25%.
Checkout Page take: Premium add-ons can run as order bumps, one-click upsells, or separate tiers. Anything under roughly 25% of the core price fits as an order bump. Above that, a one-click upsell or a tier toggle reads more naturally.
7. Thank-you page upgrade offers
The thank-you page is the screen the buyer lands on after paying, and it's the most overlooked spot in the checkout flow. A relevant offer here (a related course, a higher tier, an annual plan) can convert at decent rates, because the buyer has just said yes but isn't being pushed to say yes again right now.
When to use it: When you have a follow-on product or upgrade that's relevant but doesn't need to be decided inside the checkout. Good fit for offers that benefit from a beat of breathing room: an annual upgrade after a monthly subscription, an advanced course after a beginner one, a community membership after a one-off purchase.
Real example: A self-paced course platform shows a thank-you page with a $19/month paid community offer, with one-click sign-up using the stored payment method. The conversion rate runs about half of a one-click post-purchase upsell, but it carries no risk of disrupting the primary checkout.
Expected AOV lift: Lower than a one-click post-purchase upsell, because the buyer has visually "finished" the purchase before seeing the offer. Take-rates on well-matched offers tend to sit at 5-12%.
Checkout Page take: The thank-you page is configurable in the checkout builder. You can replace the default confirmation screen with an upsell page, redirect to a separate landing page, or keep the default and add an offer further down. The right pattern depends on how much attention the offer needs.
How to upsell in your checkout
Placement matters as much as the offer itself. Three rules cover most cases.
Match the placement to the size of the ask. Low-cost add-ons (under 25% of cart) belong in an order bump. Buyers will tick a box for a small extra but won't pause to evaluate one. Higher-cost upgrades (50-100% of the original purchase) belong in a post-purchase one-click upsell, because by then the buyer has already said yes once and the second yes is easier than it'll ever be.
Don't hide the original purchase behind the upsell. A buyer who can't see what they just bought leaves the funnel angry. Confirm the original purchase visibly, and offer the upsell alongside or after the confirmation, never instead of it. Our high-converting checkout guide goes deeper on placement and UI.
Make "no thanks" easy. Every upsell page needs a visible, plain-text decline link. Not buried, not styled as a barely-visible secondary option, not framed as a guilt-trip ("No, I don't want to grow my business"). A clear decline link lifts upsell conversion, not the other way around. Buyers who don't feel trapped say yes more often than buyers who do.
In Checkout Page, order bumps, one-click upsells, and the decline UI are checkout settings, not separate integrations. The shape of your checkout sets the shape of the upsell flow.
Upselling mistakes to avoid
Five mistakes come up over and over.
Upselling before the buyer has committed. An upsell shown before the original purchase is confirmed forces the buyer to make two decisions at once. Conversion on the primary product declines, and the upsell rarely closes the gap. Order bumps are the exception, because they're a tickbox, not a decision.
Offering something irrelevant. "People who buy X also buy Y" works in retail when the volume justifies the algorithm. For digital sellers with smaller catalogues, the upsell has to be obviously connected to what the buyer just paid for. An irrelevant upsell trains the buyer to ignore future ones, and signals that the seller isn't paying attention to what they're selling.
Over-pricing the upsell. A one-click upsell priced at 3-5× the original product converts terribly. The post-purchase window is high commitment but low deliberation. Buyers will say yes to a $79 add-on after a $49 purchase. They won't say yes to a $300 one. For one-click formats, keep the upsell at 50-100% of the original price.
Stacking too many upsells. Two upsells back-to-back compound friction faster than they compound revenue. Buyers who said yes to one don't necessarily want to say yes again, and buyers who said no are now being asked twice. One well-targeted upsell beats three loose ones almost every time.
No easy "no thanks" path. Hiding the decline link, styling it as a faint secondary option, or framing the decline as a confession ("No, I don't want to save money") all make the buyer feel trapped. Trapped buyers don't accept upsells. They refund. The clearer the decline path, the higher the take-rate runs.
How to measure upsell success
Four numbers tell you whether an upsell is working. Track them per offer.
Take-rate. The share of buyers shown the upsell who accept it. A well-matched one-click upsell runs 10-25%; an order bump runs 15-30%. Below 5% means the offer is wrong and the upsell is probably hurting more than helping.
AOV before vs after. Compare your average order value over a sample of transactions before the upsell, against the same metric after. The difference times transaction volume is the upsell's real revenue impact.
Refund rate on the upsell vs the core product. If the upsell refund rate is noticeably higher, you're probably converting buyers who didn't really want to say yes. Refunds on small-margin upsells can wipe out the AOV lift faster than the take-rate number suggests.
Downstream churn for subscription upgrades. For monthly-to-annual upgrades, the number that decides whether the upgrade was worth it is whether those buyers stay subscribed. An annual buyer who churns after one renewal is worth less than a monthly buyer who stays for fifteen months. Track 12-month retention on the upgrade cohort against the non-upgrade cohort.
The cost side matters too. On upsells priced under $20, Stripe processing fees can eat a meaningful share of the margin, especially on international cards. Run the math before launching an aggressive low-price upsell strategy.
From the Checkout Page team
We build checkouts on Stripe for digital sellers, course creators, and small subscription businesses. Of the seven strategies above, the post-purchase one-click upsell is the one we see lift AOV most consistently, when the offer is clearly relevant. The pattern that holds across customer accounts: take-rates of [X%] on well-matched upsells, near-zero (and refund-spiking) on irrelevant ones.
The most common mistake we see is the wrong moment, not the wrong offer. Sellers add a one-click upsell to a product where the buyer hasn't seen value from the original purchase yet, and the upsell underperforms not because the offer is wrong but because the timing is.
Not every cart needs an upsell
If your AOV is already close to the ceiling of what your audience will pay, the upsell will hurt more than help. If your product has no natural next step (a one-off purchase with no related products and no upgrade path), the upsell will feel forced. If your buyers had to think hard before clicking buy, a second offer right after reads as opportunistic, not generous.
The right shape for those checkouts is no upsell, a tight confirmation screen, and a careful follow-up email a few days later.
Upselling FAQ
What's the difference between upselling and cross-selling?
Upselling offers the same buyer a higher-priced or upgraded version of what they're already choosing. Cross-selling offers them a complementary product. A coffee shop upselling pushes you from a medium to a large; cross-selling pushes you from a coffee to a coffee and a pastry.
When should I add an upsell to my checkout?
When you have a clear "next step" product and can articulate in one sentence why it's the obvious next thing. If you can't articulate that, the upsell isn't ready, and it's better to skip it than to add a weak offer that erodes trust on every transaction.
How much can upselling lift my average order value?
It varies by industry, offer quality, and price ratio. Well-matched post-purchase one-click upsells generally see take-rates in the 10-25% range based on practitioner reports. Run the math on your own numbers before committing to a target; generic industry stats tend to overstate what a typical seller sees.
Does upselling hurt conversion?
It can, when done badly. An upsell shown before the buyer has committed, or one that's irrelevant to what they just bought, can drag down conversion on the primary product without compensating for the lift. The post-purchase one-click upsell is the safest format, because the original purchase is already confirmed by the time the upsell is shown.
What's a one-click upsell?
An offer shown immediately after the buyer has paid for the original product. The card is already charged, the payment method is stored, and accepting adds a second charge with a single click. Stripe doesn't support this natively; setup and mechanics live in our Stripe one-click upsell guide.
What's a good upsell take-rate?
For a post-purchase one-click upsell, 10-25% is a reasonable target. For order bumps, 15-30%. Take-rates below 5% on either format mean the offer needs work (usually relevance, occasionally price, occasionally copy). Treat any number below 5% as a signal to revisit the offer rather than to add more upsells.
