Cart abandonment is the most-discussed and least-understood number in ecommerce. The headline figure has barely moved in fifteen years, sitting somewhere between 67% and 73% globally, and it feeds an endless supply of "fix your funnel!" advice that mostly hides the real problem.
The real problem has two layers. There is a structural layer of carts that were never going to convert — the visitor was researching, comparing prices, or window-shopping. And there is a recoverable layer, where someone genuinely wanted to buy and got blocked by something fixable. This guide separates the two and recommends recovery only for the second.
What is cart abandonment?
Cart abandonment is the percentage of visitors who add at least one item to their cart but do not complete checkout. The standard formula:
cart abandonment rate = 1 − (completed orders / cart-reachers)
Two notes worth knowing. First, "cart-reachers" excludes people who never added an item — they are not abandoning, they are just browsing. Second, the denominator is per-session, not per-visitor: one customer who comes back three times and buys on the fourth counts as one completed order and four cart-reaches.
Cart abandonment is not the same as checkout abandonment, the narrower share of visitors who started checkout but did not finish. They overlap but answer different questions:
| Metric | Numerator | Denominator | 2026 average |
|---|---|---|---|
| Cart abandonment rate | Cart-reachers minus completers | Cart-reachers | 70.19% |
| Checkout abandonment rate | Checkout-starters minus completers | Checkout-starters | 28-35% |
| Browse abandonment rate | PDP visitors minus cart-adders | PDP visitors | 88-93% |
Most "cart abandonment" articles blur the three together, but the fixes differ. Browse abandonment is a content and product-page problem. Cart abandonment is a friction problem. Checkout abandonment is friction plus trust.
What do the 2026 numbers say?
The headline 70.19% hides wide variation. Five slices worth knowing.
Baymard Institute, 2026
Baymard 2026
By device. Mobile abandons more than 20 points higher than desktop. As mobile traffic grows (now 73% of all ecommerce sessions per Shopify analytics 2026), the blended rate climbs mechanically.
By industry. Apparel sits around 67%, beauty around 71%, electronics around 73%, home and furniture around 79%, jewellery and luxury around 82%. Higher order values correlate with higher abandonment, because expensive purchases get compared more.
By traffic source. Direct and email abandon least (60-65% on desktop) because the visitor arrived with intent. Paid social abandons most (85-90%) because that traffic is interruption-driven.
By time of day. Late evening (8-11 PM local) abandons most; lunchtime (12-1 PM) completes most. It tracks with attention: midday buyers are usually finishing earlier research, while late-night browsers are still discovering.
By cart age. Carts under five minutes old convert at 35%; carts older than an hour convert at 8%. The window for any recovery action is short.
Why do customers abandon carts?
The 2026 Baymard reasons-for-abandonment study ranks the top causes. The top five account for roughly 78% of stated abandonment:
- Extra costs surprised at checkout (shipping, taxes, fees) — 47%
- Account creation required — 24%
- Long or complicated checkout process — 22%
- Could not see total order cost upfront — 18%
- Did not trust the site with credit card info — 17%
(Respondents could pick more than one reason, so the percentages exceed 100%.) Content was rephrased for compliance with licensing restrictions.
Two things stand out. Four of the top five are checkout-page problems, not cart-page problems — the abandonment happens downstream of the friction. And the number one cause, surprise costs at checkout, has topped this list every year since Baymard began the study in 2014. It is the most consistently failed piece of ecommerce UX there is.
How much of it can you actually recover?
The structural floor — visitors who were never going to buy — sits around 50-55% even on the best stores. Roughly 30% of cart-adders are explicit comparison-shoppers, 15-20% are window-shoppers who will not buy this session, and 5-10% are bots or test traffic.
Your recoverable share is the gap between your current rate and that floor. At an $85 average order value and 50,000 monthly visitors converting at 2.5%, recovering 11 of those 20 points lifts revenue from $106,250 to $152,000 — a $45,750 a month improvement.
The cart abandonment calculator runs this on your own funnel. Use it to set a realistic target before you pick a recovery strategy.
Which recovery channels work in 2026?
Ranked by the lift they produce today. The lower ones still have a place, but they are where most teams overspend.
In-session help (the highest-return channel)
The onsite conversion agent — covered in the field guide — steps in before the visitor leaves. It watches for signs of abandonment intent, like the cursor heading for the exit, the tab losing focus, or a long idle on the cart page, and responds with something useful in the moment: a coupon that applies in one tap, a comparison, a quick answer. Cost per recovered order is the lowest of any channel, because it runs on the same page the shopper is already viewing.
Cart abandonment email
The 2018-2020 standard: a three-step sequence at 1 hour, 24 hours, and 72 hours. Recovery rates were 8-12% back then; the 2026 median is 3-7%. Still useful, just smaller. The decline has a few causes:
- Email open rates have fallen broadly — now 24-28% across ecommerce, down from 32-38% in 2019.
- iOS Mail Privacy Protection (2021 onward) made open signals unreliable for triggering, so sequences fire less precisely.
- Shoppers who would once have re-engaged by email now often re-engage in the session if an in-session layer is present.
Keep email. It is just no longer the leading edge.
SMS
TCPA in the US requires explicit prior written consent for SMS marketing; GDPR and ePrivacy in the EU require similar opt-in proof; CASL in Canada is stricter still. Getting this wrong is expensive. Where you have the consent, SMS works best as a single message 4-6 hours after abandonment. Multi-message SMS sequences drive unsubscribe-and-block rates that damage your sender reputation over time.
Retargeting ads
The most-cited channel and the lowest-return. Cost per recovered order on retargeting has climbed about 28% since 2022 as auctions have crowded. It still contributes for most stores, but at margins 5-10x worse than in-session help. One 2026 pattern that does work: retarget people who saw the cart but not the product page with brand content rather than a discount.
Exit-intent popups
A fading category. Banner blindness on exit-intent now runs around 92% on desktop and 96% on mobile (Nielsen Norman Group, 2026). The honest recommendation is to skip exit-intent popups in 2026 unless you have nothing else running.
Why is mobile cart abandonment so high?
Mobile abandons more than 20 points higher than desktop. Three changes do most of the work, in order:
- Add Apple Pay and Google Pay above the email field. Wallet-pay completes 1.7x more often than a typed card number on mobile.
- Use device-native form types. Setting
inputmode="numeric",autocomplete="cc-number", andautocomplete="postal-code"is the highest-leverage two-line change in mobile checkout. - Move shipping cost into the cart, not the checkout. Surprise costs at checkout are the top reason for abandonment; remove the surprise and you remove the leak.
How do you run a recovery program?
Not all at once. Six weeks is about right for a clean rollout you can actually measure against a holdout.
Week 1. Add in-session help. Trigger it when the cart is not empty and the visitor shows exit intent. Start with one move: a coupon that applies in a single tap.
Week 2. Add a three-step email sequence (1 hour, 24 hours, 72 hours), personalised to the cart contents.
Week 3. Pull the holdout-versus-treatment data and work out the lift per channel.
Week 4. Add SMS for the opted-in subset. One message at four hours.
Week 5. Audit checkout against the Baymard top five — surprise costs, forced account creation, form length, total cost upfront, trust signals — and fix the worst two.
Week 6. Re-pull the data.
By the end of week six the rhythm is set, and the program runs with a monthly holdout reset and a quarterly review.
What does the calculator tell you?
The cart abandonment calculator models the recovered revenue from any reduction on your own funnel. Three outputs are worth tracking: monthly recovered revenue, payback period on the tool itself, and the lift on completed orders (the cleaner number for comparing across stores).
For the example store above, recovered revenue lands near $48,000 a month. Whether the payback is fast comes down to how that figure compares with your plan; see /pricing and model it against your own numbers rather than anyone's averages.
What are the compliance rules?
Recovery has a compliance footprint that grows with the channel.
In-session help. Lowest footprint. It reads behaviour signals in the moment and does not store personal data until the visitor engages, which keeps it inside legitimate-interest scope under GDPR and ePrivacy; the 2026 ICO guidance is explicit on this. No cookie-consent banner is required for the help itself.
Email recovery. Needs opt-in proof. The cart-creation step is sometimes treated as opt-in if the visitor entered an email there; confirm with your DPO. Stable across the EU, US, and UK with proper consent.
SMS recovery. Highest footprint. TCPA in the US requires explicit prior written consent, and the EU and UK require similar. The opt-in at cart-creation must specifically mention SMS marketing; "by entering your phone you agree to our terms" is not enough and has driven class actions in 2024-2026.
Retargeting. The Meta Pixel and Google Ads tag both require consent in the EU and UK under ePrivacy and the 2026 EU AI Act transparency provisions. Most stores manage this with Cookiebot, OneTrust, or Klaro. Get the consent layer right before you turn retargeting on.
Further reading
- GuideHow onsite conversion agents workThe in-session layer behind the highest-return recovery channel.
- GuideThe proactive-engagement referenceThe behaviour cues behind in-session recovery.
- GuideThe 2026 CRO playbookWhere cart recovery fits in the wider funnel.
- ToolCart abandonment calculatorModel recovered revenue on your own funnel.
- CompareOnsite conversion agent vs chatbotProactive in-session help versus the corner bubble.
Frequently asked questions
Cart abandonment is the share of visitors who add to cart but do not complete checkout. Checkout abandonment is the narrower share who reach the checkout page but do not finish. The 70.19% figure is cart abandonment; checkout abandonment alone is 28-35%.
Last updated May 31, 2026.