Key Cart Abandonment Statistics at a Glance
| Metric | Value | Population / Scope | Data Period | Source |
|---|---|---|---|---|
| Overall cart abandonment rate | ~70–75% (aggregate) | Multi-study aggregate, global ecommerce | Ongoing | Baymard Institute |
| Mobile cart abandonment rate | >85% | Mobile web sessions, global | 2024 | Contentsquare / Barilliance |
| Desktop cart abandonment rate | ~65% | Desktop sessions, global | 2024 | Contentsquare |
| Travel industry abandonment rate | >90% | Online travel booking funnels | 2024 | Barilliance Industry Benchmarks |
| Recovery email open rate | ~40–45% | Automated cart emails, e-commerce senders | 2024 | Klaviyo / Omnisend |
| Unexpected costs as top exit reason | 48% of respondents | US online shoppers surveyed | 2024 | Baymard Institute Survey |
A published rate of 65% and a published rate of 85% can both be correct for the same period if one counts registered-account carts with a 24-hour window and the other counts all add-to-cart events in a session. Before comparing any two figures, confirm they use the same counting unit, the same observation window, and the same population. The section on definitions and the formula below explains these distinctions in detail.
What Is Cart Abandonment?
Cart abandonment is the condition where a shopper adds at least one item to an online shopping cart but does not complete a purchase within the measurement window the analyst defines. That last clause — "within the measurement window" — matters as much as everything before it. A purchase completed 48 hours after adding to cart might count as a conversion in one system and as an abandoned-cart recovery in another.
Several related terms appear in the same research, and mixing them produces rates that cannot be meaningfully compared.
A qualifying cart does not lead to a completed purchase within the defined observation window. Begins when the cart is created or when a qualifying add-to-cart event fires.
A shopper enters the checkout flow but does not complete the purchase. Only a subset of cart abandoners reach checkout, so checkout abandonment rates are lower than cart abandonment rates for the same store.
A shopper views product pages without adding anything to a cart. A distinct behavior that requires a separate definition and measurement approach.
A previously abandoned cart leads to a later purchase, attributed under a defined attribution rule. Recovery attribution does not by itself show that the recovery action caused the purchase.
Purchases divided by a specified population — sessions, users, or visits. Not the same as cart-to-purchase rate, even though both describe conversion. The denominator differs.
Total acquisition spend divided by customers acquired. A business-level metric distinct from any per-cart or per-session conversion figure.
[(Eligible carts − Carts converted to purchases) ÷ Eligible carts] × 100
Understanding this formula in context is easier if you are already comfortable with how rates and percentages work in general. The site's introduction to basic probability and the mean, median, and mode explainer both provide relevant grounding for reading benchmark data correctly.
Overall Cart Abandonment Rate Statistics
The most frequently cited overall figure comes from the Baymard Institute, which aggregates published studies on shopping cart abandonment from across the academic and commercial research literature. Their current aggregate sits around 70–75%. Because Baymard pools multiple studies with different methodologies, this is more accurately described as a weighted average of reported rates than as a single population-level measurement.
Different platforms and research firms that publish their own platform-telemetry data (rather than survey aggregates) tend to report rates at the higher end of the range. Contentsquare's 2024 Digital Experience Benchmark report, which draws from observed session data across thousands of ecommerce sites, reported an overall rate closer to 73%, with meaningful variation by vertical and device.
| Finding | Rate | Measurement Basis | Data Period | Source |
|---|---|---|---|---|
| Baymard aggregate (multi-study) | ~70–75% | Aggregate of published studies; mixed methods | Ongoing, updated 2024 | Baymard Institute |
| Contentsquare observed session data | ~73% | Platform telemetry; sessions with cart events | 2024 | Contentsquare Digital Experience Benchmark |
| Barilliance tracked ecommerce sites | ~77–78% | Tracked sites; add-to-cart sessions without purchase | 2023–2024 | Barilliance Cart Abandonment Statistics |
| Dynamic Yield benchmarks | ~72–74% | Platform data across global ecommerce clients | 2024 | Dynamic Yield Personalization Benchmarks |
| Practical working range | 70–80% | Across methods; most stores fall in this window | 2023–2024 | Synthesized from above sources |
Source: Baymard Institute, "Cart Abandonment Rate Statistics," 2024 update; Contentsquare Digital Experience Benchmark Report, 2024; Barilliance "Shopping Cart Abandonment Statistics" report, 2024; Dynamic Yield Ecommerce Personalization Benchmarks, 2024.
Baymard's own research notes that roughly 58% of US online shoppers have abandoned a cart because they were "just browsing" or not yet ready to buy. This matters when interpreting the total value of products sitting in abandoned carts: not all of it represents recoverable revenue. Treating the full cart value as foregone income overstates the actual opportunity.
Cart Abandonment Rate by Device
Device-level data shows a consistent pattern across all major benchmark datasets: mobile shoppers abandon at higher rates than desktop shoppers. This gap has persisted for several years despite improvements in mobile checkout design, suggesting it is partly structural — reflecting differences in purchase intent and context rather than just screen-size friction.
Mobile
Highest abandonment rate across all device categories. Mobile users browse more and purchase less per session than desktop users. Cross-device purchases (browsing on mobile, buying on desktop) are not always attributed to the mobile session.
Desktop
Consistently lower abandonment rate. Desktop sessions tend to have higher purchase intent and fewer friction points in the checkout flow. Most payment credential autofill features work more reliably on desktop browsers.
Tablet
Falls between mobile and desktop in most datasets. Tablet traffic has declined as a share of total ecommerce visits over recent years, making tablet-specific benchmarks less common in current reports.
| Device | Abandonment Rate | Denominator | Data Period | Source |
|---|---|---|---|---|
| Mobile (web) | 85–90% | Sessions with cart event; mobile web only | 2024 | Barilliance; Contentsquare |
| Desktop | ~65% | Sessions with cart event; desktop browsers | 2024 | Contentsquare |
| Tablet | ~75–78% | Sessions with cart event; tablet devices | 2024 | Barilliance |
Source: Barilliance "Shopping Cart Abandonment Statistics by Device," 2024; Contentsquare Digital Experience Benchmark, 2024. Mobile web figures exclude native app sessions unless otherwise stated.
A shopper who browses and adds items on mobile but purchases on desktop will appear as an abandoned mobile session and a new desktop conversion. The mobile session gets no credit for the eventual sale. This means device-level abandonment rates overstate mobile underperformance when cross-device behavior is common in the audience. Logged-in account data and cross-device identity tools reduce this problem but do not eliminate it.
Cart Abandonment Rate by Industry
Industry differences in abandonment rates reflect differences in purchase complexity, research intensity, and how shoppers use carts in each category. A travel booking involves significant planning, timeline constraints, and comparison across many options — conditions that structurally push abandonment higher. A fashion purchase involves less friction and lower commitment, which pulls the rate somewhat lower.
| Industry | Reported Rate | Notes |
|---|---|---|
| Travel and hospitality | >90% | Booking funnels; multi-session comparison behavior common |
| Financial services | ~83–85% | Insurance quotes, loan applications; high research phase |
| Automotive (parts and accessories) | ~80% | Compatibility checks and fitment research extend decision |
| Electronics | ~77–80% | Price-sensitive; comparison shopping delays purchase |
| Sporting goods | ~73–76% | Size and specification considerations add friction |
| Home and garden | ~74% | Delivery logistics and room-planning delays common |
| Fashion and apparel | ~70–74% | Closer to overall average; high return rates complicate net conversion |
| Beauty and personal care | ~69–72% | Repeat purchase category; loyalty program effects |
| All industries (Barilliance median) | ~77% | Median across all tracked verticals; 2024 data |
Source: Barilliance "Cart Abandonment Rate by Industry," 2024 report. Figures are reported medians for each category using Barilliance's platform methodology. Industry labels follow Barilliance's own classifications. Travel and financial services figures reflect booking/application funnel abandonment, which uses the same formula as cart abandonment but applies to a different conversion event.
If one benchmark source reports fashion at 70% and a separate source reports electronics at 77%, those figures cannot be placed in the same ranking unless both used the same measurement methodology applied to comparable populations. Industry comparisons are only reliable within a single consistent dataset from the same period.
Reported Reasons for Cart Abandonment
Baymard Institute's 2024 checkout usability survey asked US online shoppers why they had abandoned a checkout in the past quarter. The survey excluded respondents who said they were "just browsing" from the primary analysis of avoidable abandonment. The percentages below come from respondents who had a genuine purchase intent but did not complete it. Respondents could select multiple reasons, so these figures do not sum to 100%.
Source: Baymard Institute "Why Do Users Abandon the Checkout?" survey, 2024. US respondents (n = 4,384) who had abandoned a checkout in the past quarter, excluding "just browsing" respondents. Multi-select question; percentages do not sum to 100%.
Self-reported reasons for abandonment are useful for generating hypotheses about checkout friction. They are not experimental evidence of causal effect sizes. A store that eliminates mandatory account creation will not automatically recover 24% of abandoned carts — the actual impact depends on what share of its abandoning visitors cited that reason and whether the fix introduces other friction. Treat these figures as diagnostics that point toward where to look, not as predictions of recovery magnitude.
Abandoned Cart Recovery Email Statistics
Recovery email benchmarks describe the performance of automated messages sent to shoppers who left items in a cart. These figures come from email service provider platform data, which tracks delivered messages and attributed outcomes using that platform's own attribution rules. Two important constraints apply before reading any of these numbers.
First, not every abandoned cart is eligible for a recovery email. The shopper must be identifiable — meaning they are either logged in or have previously provided an email address. Platform data suggests that roughly 10–25% of abandoning visitors are identifiable, though this varies considerably by store and traffic mix.
Second, attributed orders are not the same as orders caused by the email. A shopper who was already planning to return and complete the purchase will appear as a "recovered" cart in the attribution data even if they would have converted without any message.
| Metric | Value | Denominator | Source |
|---|---|---|---|
| Open rate | ~40–45% | Delivered messages | Klaviyo, Omnisend 2024 |
| Click-through rate | ~8–10% | Delivered messages | Klaviyo, Omnisend 2024 |
| Click-to-open rate | ~20–22% | Measured opens | Klaviyo 2024 |
| Placed-order rate (per recipient) | 5–8% | Delivered messages | Klaviyo, Omnisend 2024 |
| Revenue per recipient | $5–$12 | Delivered messages; varies by AOV | Klaviyo 2024 |
| Attribution window (typical) | 5–7 days | After email delivery; platform-defined | Klaviyo / Omnisend default settings |
Source: Klaviyo Benchmark Report, 2024; Omnisend Email Marketing Statistics, 2024. All figures are platform medians across ecommerce senders using automated cart recovery flows. Open rate tracking on iOS devices has been affected by Apple Mail Privacy Protection since late 2021; these open rates may be inflated for senders with significant iPhone audiences.
Multi-Email Sequence Performance
Omnisend's 2024 automation benchmark data found that three-email recovery sequences outperform single-email flows on attributed revenue per eligible cart by a meaningful margin. Their data showed sequences generating roughly 69% more attributed orders than a single email for comparable populations. That figure describes attributed outcomes under their platform's attribution model — it does not control for whether multi-email recipients were more engaged to begin with.
| Message in Sequence | Typical Send Delay | Average Open Rate | Placed-Order Rate |
|---|---|---|---|
| First message | Within 1 hour | ~45% | ~5–6% |
| Second message | ~24 hours later | ~40% | ~3–4% |
| Third message | ~72 hours later | ~35% | ~2–3% |
Source: Omnisend Email Marketing Statistics Report, 2024. Figures represent median performance across stores using three-message automated cart recovery sequences. Rates decline across the sequence partly because purchasers who respond to earlier messages exit the flow before receiving later ones.
Recovery emails attribute purchases that occur within a window after delivery. Some fraction of those purchasers would have returned and bought without any message. The only way to measure how many additional purchases the email actually caused is to run a holdout group — a randomly selected set of eligible abandoners who receive no recovery email. Without a holdout, attributed orders overstate the causal impact of the campaign.
Worked Measurement Examples
The following examples are hypothetical and constructed to illustrate how the formulas work. These numbers do not represent any real store's performance and should not be cited as benchmarks.
A store observes a cohort of 1,000 eligible carts over a defined purchase window
| Outcome | Number |
|---|---|
| Eligible carts | 1,000 |
| Carts reaching checkout | 400 |
| Carts producing a completed purchase | 250 |
| Carts without a completed purchase | 750 |
| Checkout starts without a completed purchase | 150 |
Calculations:
Cart abandonment rate = 750 ÷ 1,000 × 100 = 75%
Checkout abandonment rate = 150 ÷ 400 × 100 = 37.5%
Cart-to-purchase completion rate = 250 ÷ 1,000 × 100 = 25%
The cart abandonment rate (75%) counts all 750 carts that did not convert. The checkout abandonment rate (37.5%) counts only the 150 carts that entered checkout but did not complete — using those 400 checkout starters as the denominator, not all 1,000 carts. Both rates are valid; they describe different stages of the same funnel. Reporting the checkout abandonment rate in place of the cart abandonment rate would understate the overall dropout significantly.
After the initial window, a recovery email is sent to identifiable abandoners
Of the 750 abandoned carts above, 200 are contactable (the shopper provided an email address). One email is delivered to each of those 200 abandoners. Within the attribution window, 14 purchases are attributed to the email.
Placed-order rate per delivered email = 14 ÷ 200 × 100 = 7%
Those attributed purchases as share of all abandoned carts = 14 ÷ 750 × 100 = 1.9%
Note that the 7% figure (per delivered email) and the 1.9% figure (per abandoned cart) describe the same 14 purchases. The difference is the denominator. Neither figure indicates how many of those 14 purchases were caused by the email rather than occurring independently.
Statistical Interpretation Notes
Cart abandonment statistics are reported in ways that produce interpretive errors when readers apply them without checking the underlying definitions. The most common problems are listed here with brief explanations. For a broader treatment of how to read statistical data in a business context, see the site's business statistics guide and the correlation vs. causation explainer.
- Percentage points vs. relative change. A drop from 75% to 70% is a 5 percentage-point decrease, or a 6.7% relative reduction. These are not the same thing. Reporting a "5% improvement" when the absolute change is 5 percentage points doubles the apparent effect size. For a deeper explanation, see the relative frequency guide.
- Mean vs. median store-level rates. A handful of stores with very low abandonment (well-optimized, loyal customer bases) can pull the mean downward. The median is more representative for most stores. Benchmark reports that give only a mean without noting distribution width conceal how much rates vary. See the site's mean, median, and mode guide.
- Device-mix confounding. A store that gets 80% of its traffic from mobile will have a higher overall abandonment rate than a store with 50% mobile traffic, even if both stores perform identically on each device. Changes in traffic source composition can shift the aggregate abandonment rate without any change in actual checkout performance.
- Attribution vs. causation. Orders attributed to a recovery email within a 7-day window are not necessarily orders caused by the email. Without a holdout group, the incremental value of the campaign cannot be measured. This mirrors a broader principle in data analysis covered in the correlation vs. causation page.
- Incomplete observation windows. If the analysis window closes before all potential conversions have had time to occur, the measured abandonment rate will be higher than the true rate for that cohort. Long-cycle categories — high-value electronics, furniture — often see purchases 5–10 days after initial cart creation.
- Survey vs. observed behavior. Reported reasons for abandonment come from surveys. Observed behavioral data (session recordings, funnel analytics) tells you where shoppers drop off but not why. Both types of evidence answer different questions and should not be substituted for each other.
For confidence interval calculations relevant to testing checkout changes, the site's confidence interval reference and A/B testing statistics guide provide the relevant methodology.
How to Apply These Benchmarks to Your Store
An industry-level benchmark is useful only as a directional reference point. The relevant comparison for any store is its own historical baseline for the same metric, defined the same way. The sequence below gives a practical starting point.
Define your cart and checkout events precisely
Specify which event marks a cart as eligible (add-to-cart, session-persistent cart creation, or account-linked cart). Set a consistent observation window — typically 24 to 72 hours — and stick to it across all reporting periods.
Choose a consistent counting unit
Carts, sessions, and users are not interchangeable denominators. A shopper who creates three carts across three sessions counts as one user, three sessions, and three carts. Pick the unit that best reflects your business question and use it consistently.
Calculate your baseline before comparing to benchmarks
Record your current rate using the definitions above before looking at external benchmarks. Checking the benchmark first introduces confirmation bias — the tendency to find data that supports what you already believe about your store's performance.
Segment by device and traffic source
Your aggregate abandonment rate mixes audiences with very different purchase intent. Mobile-organic, desktop-paid, and email-referred sessions typically behave differently. Segmentation reveals where friction is highest and which audience segments are underperforming against their own baseline.
Select benchmarks with comparable definitions
Match your benchmark source to your measurement method. If your rate measures carts with at least one item added and a 48-hour window, look for benchmark sources that describe the same setup. Benchmarks built from checkout-initiation data are not directly comparable to cart-creation data.
Test changes and measure completed purchases
The abandonment rate is a diagnostic, not a target. Changes to checkout design, pricing display, or shipping presentation should be measured against completed purchase volume and revenue, not just against the abandonment rate. A lower abandonment rate that produces the same or fewer completed purchases has not improved the business outcome. For test design, see the site's statistics for A/B testing guide.
Data Sources and Methodology
All statistics on this page carry explicit source attribution, the population covered, and the data period. The table below identifies the primary research organizations and notes relevant limitations for each source used.
| Organization | Coverage on This Page | Methodology | Key Limitation |
|---|---|---|---|
| Baymard Institute | Overall aggregate rate; survey reasons for abandonment | Multi-study aggregate + proprietary usability survey; US-focused | Aggregate pools studies with different methods; survey reasons are self-reported |
| Contentsquare | Overall rate; device-level rates | Platform telemetry across thousands of global ecommerce sites | Contentsquare client base may not represent all ecommerce segments |
| Barilliance | Overall rate; device rates; industry breakdown | Tracked ecommerce sites; session-level event data | Barilliance-enabled sites; potential self-selection in client mix |
| Klaviyo | Recovery email open rate, CTR, placed-order rate | Platform telemetry; automated cart flow emails across Klaviyo senders | Apple MPP inflates open rates; Klaviyo senders skew toward mid-to-large ecommerce |
| Omnisend | Recovery email sequence performance; timing data | Platform telemetry; automated flows across Omnisend senders | Attribution window set by platform default; incremental recovery unmeasured |
| Dynamic Yield | Overall cart abandonment rate cross-reference | Platform data across global enterprise ecommerce clients | Enterprise-skewed client base; not representative of all store sizes |
Cart abandonment benchmarks shift as platform tracking capabilities, mobile commerce behavior, and checkout technology evolve. The figures on this page reflect research available as of October 2026, with data periods noted for each finding. Older research is included only where a more recent equivalent is unavailable and the finding remains directionally relevant. This content is for research and planning purposes only and does not constitute commercial, legal, or professional advice. Statistics described as estimates are subject to change as primary sources update their methodologies.