Cart Abandonment Statistics: What the Benchmarks Really Tell You (And What They Dont)

The number everyone quotes and nobody checks

It's in the deck. Slide four, under a heading about conversion. "70% of shoppers abandon their carts." Someone nods. Someone else writes it down. The meeting moves on to the budget.

Ask the room where that number came from and you'll get a shrug. Ask what it was measured on - which stores, which categories, which years - and the shrug gets uncomfortable. It's a statistic everyone has absorbed and nobody has read.

That matters, because the number is being used to make decisions. It's used to justify a checkout rebuild. It's used to set a target. It's used to explain a bad quarter. And it's being applied to your store, which is not the average of fifty other stores.

Here's what this page does. It gives you the numbers that are actually worth knowing, clearly attributed. It shows you how to tell a statistic you can trust from one you can't. And it makes the case that the only benchmark that should ever set your targets is the one sitting in your own analytics.

By the end, you'll be the person in the room who knows where the number came from.

The one statistic worth knowing: 70.22%

If you're going to carry one published figure in your head, carry this one.

Baymard Institute reports an average documented online shopping cart abandonment rate of 70.22%, based on 50 studies.

That's the anchor. It's the most cited number in ecommerce, and it's cited correctly more often than most - which is not the same as being used correctly.

First, what it actually is. Baymard didn't run one study and publish the result. They collected 50 separate studies of cart abandonment, each with its own sample, its own methodology, and its own definition of what counts as an abandoned cart. Then they averaged them. That's a meta-analysis: a study of studies. It's a legitimate and useful way to find the center of a noisy field.

Now the part that gets skipped. An average across 50 studies is a sanity check, not a target.

A sanity check tells you whether you're in normal territory. If your rate is near 70%, nothing is structurally broken. If your rate is 92%, something is. That's genuinely useful information, and it's the correct use of the figure.

A target is something else entirely. A target implies that 70.22% is a destination your store should be moving toward, and that's not what the number means. It's the midpoint of a very wide distribution. The studies inside that average span every category, every price point, every device mix, and every level of checkout quality. Some of those stores were excellent. Some were broken. Averaging them produces a number that describes the field, not your store.

So use it to orient yourself. Don't use it to set a goal.

The documented reasons for abandonment

The other Baymard dataset worth knowing is the list of documented reasons shoppers abandon. These are self-reported reasons from real shoppers, and they're the closest thing the industry has to a map of the problem.

All figures: Baymard Institute.

Read that table and notice what's missing from the top. Price isn't the leading reason. The two biggest entries are timing (they weren't ready) and surprise (the total changed at the last step). Those are different problems with different fixes, and neither one is solved by a discount code.

Notice something else, too. The shares don't add up to 100%, and they shouldn't. Shoppers were asked to select reasons, not to rank a single cause. Most abandonments have more than one reason behind them. A shopper who hit a slow checkout on a phone, then saw a shipping charge she didn't expect, then decided she'd think about it - that's three rows in this table and one abandoned cart.

That's the honest reading. The table tells you which reasons are common. It does not tell you which reason is yours.

Why a single average hides everything

Here's the limitation that makes this whole category of statistics tricky, and it's worth being blunt about.

A single average hides everything that matters.

Abandonment behavior varies enormously by:

• Device. Mobile and desktop behave differently, and the gap is usually about checkout friction rather than shopper intent.
• Product category. Something bought on impulse and something bought after three weeks of research do not abandon for the same reasons.
• Price point. A $8 consumable and a $4,000 piece of furniture have structurally different natural rates. The $8 buyer decides in seconds. The $4,000 buyer is comparing, measuring, and waiting for a second opinion. Same metric, completely different behavior.
• Traffic source. A shopper who arrived from an email she opted into is not the same as a shopper who arrived from a cold paid ad.
• Checkout step. Where the drop happens tells you more than how big it is.

So when someone offers you a "cart abandonment benchmark by industry," ask what's inside it. Which stores? What price points? What device mix? What year? A benchmark that averages a $12 candle brand with a $6,000 mattress brand is describing neither of them.

We're not going to give you segment-level benchmark figures here, because reliable ones don't exist in public form. If a source hands you a precise number for "apparel, mobile, 50–100 carts," ask which stores contributed to it and how they were sampled. In most cases, nobody can tell you.

The honest position is this: segment-level benchmarks should come from your own history. Your mobile rate last quarter is a real benchmark for your mobile rate this quarter. Your paid-traffic rate is a real benchmark for your paid-traffic rate. Those numbers have a known sample, a known definition, and a known context. Published segment figures have none of that.

How to read a statistic critically

This is the skill that separates a useful number from a decorative one. Five checks, in order.

Check the sample size

A statistic built on 200 shoppers is a hint. A statistic built on 50 studies is a pattern. Neither is a law, but they carry very different weight. If a source doesn't state its sample, treat the number as an anecdote until proven otherwise.

Check the date

Ecommerce behavior shifts. Checkout technology changes. A statistic from 2014 is describing a world with different payment options, different device mix, and different delivery expectations. Baymard's meta-analysis spans studies from multiple years, which is part of why it's presented as a range of documented rates rather than a single truth. Always ask when the data was collected, not when the article was written.

Check whether it's a meta-analysis or a single study

A meta-analysis averages many studies and smooths out individual quirks. A single study reflects one company's customers, one category, one moment. Both can be useful. They are not interchangeable, and a single study presented as an industry benchmark is one of the most common errors in this space.

Check whether it measures carts or sessions

This one is subtle and it changes the number dramatically. An abandonment rate calculated on created carts measures shoppers who added something and started a purchase. A rate calculated on sessions divides abandoned carts by everyone who ever visited - including blog readers and people who bounced in four seconds. The second number looks much better and means much less. If a statistic doesn't tell you which denominator it used, you can't compare it to yours.

Check who published it, and why

This is the check nobody runs. Ask who paid for the research and what they sell.

A checkout software vendor publishing abandonment statistics has an interest in abandonment looking like a checkout problem. A recovery tool vendor has an interest in it looking like a recovery problem. A payments company has an interest in it looking like a payments problem. None of that makes the data wrong - but it does shape which questions get asked and which findings get published.

Baymard's figures are widely trusted because the research is independent and the methodology is published. That's the standard to hold other sources to. When a statistic arrives without a methodology, without a sample, and from a company that sells the fix, treat it as marketing until you can verify it.

Run these five checks on any number before it goes in a deck. Most won't survive all five. The ones that do are worth keeping.

The statistics that actually matter for your store

Here's the argument this whole page has been building toward.

Published benchmarks are context. Your own numbers are the decision.

Four proprietary numbers matter more than every statistic on this page combined:

1. Your own cart abandonment rate. Calculated consistently, on created carts, month over month. This is your baseline, and it's the only thing you can meaningfully beat.
2. Your own step-level drop-off. Which step in your checkout loses the most shoppers. This tells you where to work. A published average can never tell you this, because it's about your checkout.
3. Your own segment splits. Your mobile rate versus desktop. Your paid traffic versus email. Your high-value carts versus low. These gaps are specific to your store and they point directly at your problems.
4. Your own holdout-measured lift. What actually changed when you did something. Not what a case study claims happened at another company - what happened in your store, measured against a control group that received nothing.

Notice the pattern. Every one of these has a known sample, a known definition, and a known context. Every one of them is about your customers. And every one of them is more actionable than a number averaged across fifty other businesses.

The published statistics are useful for orientation and for framing. They tell you whether you're in normal territory and they give you a shared vocabulary. But the moment you start setting targets from them, you're optimizing toward someone else's average.

What to do with the numbers once you have them

Once you have your own rate, your own step-level drop-off, and your own segment splits, you have a diagnosis. Two places to go next:

• If you know where the leak is but not why - start with the causes. Our guide to how to reduce cart abandonment walks through every documented reason and the fixes in priority order.
• If you need to calculate or interpret your own rate first - that's covered in cart abandonment rate: what yours actually means, including the formula and the denominator mistake that flatters your number.

Statistics tell you what's normal. Your data tells you what's wrong. Get the second one, then act on it.

How we measure at ConvertFlux

One note on measurement, because it's the reason we care about this topic.

When we run recovery campaigns, we don't report attributed revenue. Attributed revenue counts every order that followed a message, including the orders that would have happened anyway. It's a flattering number and a misleading one.

Instead, we hold back a small random slice of shoppers who receive nothing, and we compare. The difference between the two groups is the incremental lift - the part that actually came from the campaign. Then we subtract the cost and look at incremental profit.

It's a slower way to report. It's also the only way to know whether the work is doing anything. If you want to see how that's set up, it's documented at our analytics page.

FAQ

What is the average cart abandonment rate?

Baymard Institute's meta-analysis of 50 studies puts the average documented online shopping cart abandonment rate at 70.22%. It's the most reliable published figure available, and it's best used as a sanity check rather than a target.

What is a meta-analysis, and why does it matter here?

A meta-analysis combines the results of many separate studies into one average. Baymard's figure averages 50 studies, each with its own sample and methodology. That makes it a robust description of the field - but it also means it describes the midpoint of a very wide range, not any single store.

What are the most common reasons for cart abandonment?

Baymard's documented reasons: 42% were just browsing or not ready to buy, 40% found extra costs too high, 20% cited slow delivery, 19% didn't trust the site with card details, 18% faced forced account creation, 17% found checkout too long or complicated, 17% hit site errors or crashes, and 13% were dissatisfied with the returns policy. Shoppers could select multiple reasons, so the shares don't total 100%.

Are there reliable cart abandonment benchmarks by industry?

Not in any form we'd stake a decision on. Public industry benchmarks rarely disclose their sample, their price points, or their methodology. Segment-level benchmarks are far more useful when they come from your own history - your mobile rate, your paid-traffic rate, your high-value cart rate - because those have a known sample and a known definition.

Why is 70.22% not a good target?

Because it's an average across 50 studies spanning every category and price point. A store selling $8 consumables and a store selling $4,000 furniture have structurally different natural rates. Use 70.22% to confirm you're in normal territory, then set targets against your own baseline.

How do I know if a cart abandonment statistic is trustworthy?

Run five checks: does it state its sample size, does it state when the data was collected, is it a meta-analysis or a single study, does it measure carts or sessions, and who published it and what do they sell? A statistic that fails most of these is marketing, not research.

What's the difference between measuring carts and sessions?

A rate based on created carts measures shoppers who added something and started a purchase. A rate based on sessions divides abandoned carts by all visitors, including people who were never buying. The session-based number looks better and means less. Always check which denominator a statistic used before comparing it to yours.

Which statistics should I actually track?

Four: your own abandonment rate, your own step-level drop-off, your own segment splits, and your own holdout-measured lift. Those have known samples, known definitions, and direct relevance to your store. Published benchmarks are context; your numbers are the decision.

The short version

Baymard's 70.22% across 50 studies is the one published figure worth carrying - as a sanity check, not a target. The documented reasons tell you what's common, not what's wrong with your store. Read any statistic with five checks: sample size, date, methodology, denominator, and who published it. Then go get your own four numbers, because those are the only ones that can tell you what to fix.

External sources

• Baymard Institute - Cart Abandonment Rate Statistics: https://baymard.com/lists/cart-abandonment-rate - supports the 70.22% figure and the meta-analysis explanation
• Baymard Institute - Reasons for Cart Abandonment: https://baymard.com/lists/cart-abandonment-rate/reasons - supports the eight-row documented reasons table
• Baymard Institute - Checkout Usability Research: https://baymard.com/research/checkout-usability - supports the critical-reading section on methodology