More sales from the traffic you already have
Fifty studies over nineteen years, from Forrester in 2010 to Uptain in 2025, and an average of 70.22%. I put every documented cart abandonment rate in one sortable table so nobody ever has to cite a random blog again.
Cart abandonment is the most quoted and worst cited number in ecommerce. Half the articles on the internet say “roughly 70%” and link to each other in a circle. The actual primary source is Baymard Institute’s registry, which collects every published study it can verify, currently 50 of them spanning 2006 to 2025, and computes the average: 70.22%. This entry puts the entire registry in one sortable table, then adds what the abandonment is made of, because the composition is where the money is.
Click a header to sort. Every row is a separate published measurement, as catalogued by Baymard with its retrieval date.
| Rate ⇅ | Study ⇅ | Data year ⇅ |
|---|---|---|
| 71.72% | Uptain | 2025 |
| 79.53% | SaleCycle | 2023 |
| 68.70% | Fresh Relevance | 2022 |
| 79.30% | Kibo Commerce | 2021 |
| 59.22% | Fresh Relevance | 2021 |
| 81.08% | SaleCycle | 2021 |
| 57.60% | Fresh Relevance | 2020 |
| 84.27% | SaleCycle | 2020 |
| 77.13% | Barilliance | 2019 |
| 56.82% | Fresh Relevance | 2018 |
| 75.60% | SaleCycle | 2018 |
| 81.40% | AbandonAid | 2017 |
| 76.90% | SaleCycle | 2017 |
| 78.00% | Listrak | 2016 |
| 75.50% | Adobe | 2016 |
| 68.80% | Barilliance | 2016 |
| 74.52% | SaleCycle | 2016 |
| 71.39% | Barilliance | 2015 |
| 68.95% | IBM | 2015 |
| 75.00% | Listrak | 2015 |
| 75.60% | SaleCycle | 2015 |
| 68.38% | IBM | 2014 |
| 72.00% | Listrak | 2014 |
| 62.30% | Fireclick | 2014 |
| 69.20% | Vibetrace | 2013 |
| 74.00% | Barilliance | 2013 |
| 67.41% | IBM / Coremetrics | 2013 |
| 78.00% | AbandonAid | 2013 |
| 60.32% | Triggered Messaging | 2013 |
| 75.00% | Listrak | 2013 |
| 74.23% | SaleCycle | 2013 |
| 67.00% | Comscore | 2012 |
| 80.30% | Rejoiner | 2012 |
| 61.85% | IBM / Coremetrics | 2012 |
| 74.76% | Fireclick / DigitalRiver | 2012 |
| 76.00% | Listrak | 2012 |
| 72.31% | Fireclick / DigitalRiver | 2011 |
| 62.31% | IBM / Coremetrics | 2011 |
| 72.00% | SeeWhy | 2011 |
| 71.00% | SeeWhy | 2010 |
| 55.00% | Forrester Research | 2010 |
| 63.68% | IBM / Coremetrics | 2010 |
| 69.38% | Fireclick / DigitalRiver | 2010 |
| 62.14% | MarketLive | 2009 |
| 71.00% | Forrester Research | 2009 |
| 63.19% | IBM / Coremetrics | 2009 |
| 68.00% | SeeWhy | 2009 |
| 62.01% | IBM / Coremetrics | 2008 |
| 61.36% | IBM / Coremetrics | 2007 |
| 59.80% | MarketingSherpa | 2006 |
Source for the full registry: Baymard Institute, last updated September 22, 2025, read July 2026.
Sort by data year and a pattern appears that a single average cannot show. The early studies, 2006 through 2012, mostly live between the high 50s and low 70s. The recent studies, 2017 onward, mostly live in the mid 70s and above, with SaleCycle measuring 84.27% during the pandemic year and 79.53% in 2023.
The usual explanation, and I find it convincing, is the device shift. Mobile now carries roughly three quarters of ecommerce traffic in Dynamic Yield’s panel, and phones are where people browse on the couch, park items as bookmarks, and get interrupted. More window shopping per session means more abandoned carts, with nothing broken.
Vertical spread does the rest. In the same Dynamic Yield data through June 2026, Beauty and Personal Care carts were abandoned at 79.81% over the trailing year while Pet Care sat at 51.1%. If your benchmark did not come from your vertical, it is barely a benchmark.
The registry answers “how much”, but the useful question is “why”. Baymard’s survey of US shoppers found 43% abandon because they were just browsing and not ready to buy. That part is weather. Among the rest, the ranked reasons are: extra costs too high, 39%. Delivery too slow, 21%. Did not trust the site with card details, 19%. Forced account creation, 19%. Checkout too long or complicated, 18%. Unsatisfactory returns policy, 15%. Site errors, 15%. Could not see the total cost up front, 14%. Not enough payment methods, 10%. Card declined, 8%.
Two thirds of your abandonment is weather. The last third is a hole in the roof, and it has a published list of coordinates.
Baymard prices the fixable part precisely: solvable checkout usability issues are worth an average 35.26% conversion lift for large sites, or about 260 billion dollars in recoverable orders across the US and EU. I walked through how I hunt those exact leaks in my checkout audit sequence, and the distinction between a parked cart and an interrupted checkout, which matter differently, lives in abandoned checkout vs abandoned cart.
The table reproduces Baymard Institute’s public registry of documented cart abandonment statistics, read in July 2026, last updated by Baymard on September 22, 2025. Each row is a separate vendor study with the data year as catalogued; retrieval dates for each study are listed on the source page. The era means in the chart are simple unweighted averages I computed from the table rows falling in each period (19 studies for 2006 to 2012, 18 for 2013 to 2016, 13 for 2017 to 2025, counting by data year), and the three era means recombine to exactly Baymard’s 70.22% overall average, which is a decent arithmetic check. Baymard’s own average of 70.22% is likewise unweighted across all 50. None of these numbers are mine; the arithmetic on eras is the only thing I added, and you can recompute it from the table in a minute.
When you need one number for a slide: 70.22%, Baymard average, cite the registry. When you need a benchmark for your store: filter the table to studies from the last five years, then adjust your expectation up if you are mobile-heavy or in beauty, down if you sell pet food on subscription. And when someone quotes you “cart abandonment is 88%” from a vendor landing page, you now have fifty receipts saying otherwise.
The most defensible answer is 70.22%, which is Baymard Institute's calculated average across all 50 published studies it tracks, last updated September 2025. The most recent single study in the registry, Uptain 2025, measured 71.72%, almost exactly on the long-run average.
By itself, no. It is the documented industry norm, and a large share of it is structural: Baymard's survey found 43% of US shoppers abandon because they were just browsing, comparison shopping or saving items for later. Worry when your rate sits well above the mid 70s, or when it jumps without a traffic mix change, because then something specific is usually broken.
Sample and definition. Each vendor measures its own client base, and verticals differ wildly: in Dynamic Yield's panel over the twelve months to June 2026, Beauty and Personal Care carts were abandoned at 79.81% while Pet Care sat at 51.1%. Device mix matters too, since mobile abandons more than desktop, so a mobile-heavy panel reports higher numbers.
Baymard estimates that fixable checkout usability problems alone are worth a 35.26% average conversion lift for large ecommerce sites, which across the US and EU translates to about 260 billion dollars in recoverable orders. The just-browsing crowd is not recoverable by design changes, but the cost-surprise and forced-account crowds are.