How to reduce cart abandonment, step by step

Roughly seven of every ten online carts never turn into an order. Baymard Institute's running average across 50 studies puts the documented cart abandonment rate at 70.22%, and that number has stayed stubbornly high through years of checkout redesigns.

Some of that is normal. People add things to a cart to save them, compare prices, or see what shipping costs, and plenty of them were never going to buy that day. The rest is revenue you can get back, and most of it traces to surprise costs and clunky checkouts, plus a quieter problem that checkout fixes can't touch: the shopper had a question and nobody answered it.

This guide works for ecommerce teams of any size, whether one person runs the whole store or a dedicated team owns checkout and support. The first half explains what normal looks like and why shoppers leave. The second half is six steps, and where the advice changes with team size or traffic, the step says so.

Cart abandonment vs. checkout abandonment

The two terms get used interchangeably, and that's how teams end up fixing the wrong thing.

Cart abandonment is when someone adds a product to their cart and leaves without buying. Checkout abandonment is narrower: the shopper started checkout, maybe typed in an email or an address, and left before paying. Every checkout abandoner is also a cart abandoner, but a lot of cart abandoners never reach checkout at all.

You can calculate both from the same funnel:

• Cart abandonment rate = (carts created − orders completed) ÷ carts created

• Checkout abandonment rate = (checkouts started − orders completed) ÷ checkouts started

Shopify's analytics and the GA4 ecommerce funnel both report add to cart, checkout started, and purchase as separate steps, so you can pull both numbers without new tracking. If cart abandonment is high and checkout abandonment is low, the problem sits earlier, in price, shipping, or doubt about the product. If checkout abandonment is high, look at the form, the payment options, and anything that breaks on mobile.

What's a normal cart abandonment rate?

Baymard's 70.22% is an average of 50 published studies, and it's the figure most people quote. Dynamic Yield, which measures live sessions across the retailers on its platform, reports a higher global cart abandonment rate of 77.54% over the trailing 12 months. The gap comes from method, since one averages studies and the other counts sessions from a specific set of stores. Either way, a rate somewhere between 70% and 80% is typical, and landing in that range doesn't mean your store is broken.

Cart abandonment rate by industry

Category matters more than cross-industry averages suggest. In Dynamic Yield's data, beauty and personal care sits at the top at 79.81%, and pet care is the lowest at 51.1%. One plausible reason is that pet supplies lean toward repeat purchases of things the shopper already knows, while beauty shoppers browse and compare shades. Compare yourself to your own category, and to your own numbers from last quarter, before you worry about a benchmark built from other people's stores.

Cart abandonment rate by device

Mobile runs higher. Dynamic Yield puts mobile abandonment at 79.92% against 69.19% on desktop, with tablets in between at 72.31%. Forms are harder on a small screen, and plenty of people browse on their phone and buy later on a laptop, which shows up as an abandoned mobile cart even when the sale happens.

Why shoppers abandon carts

Baymard regularly surveys US online shoppers about why they left. In its latest cart abandonment research, 42% said they'd abandoned a cart because they were just browsing or weren't ready to buy. Set those people aside, and here's why everyone else left during checkout:

• Extra costs were too high (shipping, tax, fees): 40%

• Delivery was too slow: 20%

• They didn't trust the site with their card details: 19%

• The site wanted them to create an account: 18%

• Checkout was too long or complicated: 17%

• The site had errors or crashed: 17%

• The returns policy wasn't good enough: 13%

• They couldn't see or calculate the total cost up front: 12%

The percentages add up to well over 100 because shoppers could give more than one reason.

Unexpected costs

Cost is the top reason by a wide margin, and the surprise often hurts more than the amount. Twelve percent left because they couldn't see the total up front, which is a separate complaint from the price being too high. A $9 shipping charge that appears for the first time on the payment step feels like a bait and switch, even when it's a fair price.

Checkout friction

Forced accounts, long forms, site errors, and declined cards together cover a big share of the list. These are the problems most checkout advice is written for, and they're worth fixing. They also have a ceiling, because once your checkout is short and works on a phone, polishing it further moves the number less.

Unanswered questions

Read the list again as a set of questions. "Delivery was too slow" can mean the shopper couldn't tell when the order would arrive. "Returns policy wasn't good enough" can mean they couldn't find it, or didn't understand it. Plenty of abandonment starts with something like: will this fit, will it get here before Saturday, can I send it back if the color's off, does it work with the one I already own? When the answer isn't on the page and there's nobody to ask, leaving is the easiest thing to do.

Shoppers who are just browsing

The 42% who were just browsing aren't a failure to fix. Some of them will come back on their own. Hitting every one of them with a discount teaches your customers to add to cart and wait for a coupon, and it eats margin on sales you'd have made anyway, so save incentives for carts where something specific went wrong.

Step 1: Ask the shoppers who left

Benchmarks tell you what's typical. Your own shoppers tell you what's happening on your store. A one-question exit survey on the cart page ("What stopped you from checking out today?") with four or five answer choices and an "other" box takes minutes to set up with most survey apps. Read your chat and email transcripts too. Search for "when will" and "what size," plus anything about returns, then count what comes up most. The answer tells you which of the next steps to do first.

Step 2: Show shipping and total cost before the cart

Cost is the top reason in Baymard's list, so it's the first fix to make. The goal is to move the number earlier so nobody sees it for the first time at checkout.

• Put your free shipping threshold on product pages and in the cart, along with how close the shopper is ("You're $12 away from free shipping").

• Show an estimated shipping cost and delivery date on the product page. Many Shopify themes and shipping apps can display this.

• Link the returns policy from the product page and the cart in one plain sentence, like "Free returns within 30 days."

• If tax gets added on top of the listed price, say so before checkout.

If your margins allow it, try building shipping into the product price and compare it against charging shipping separately. Some categories convert better with free shipping and a slightly higher price and some don't, so run it as a test on your own store.

Step 3: Fix the checkout basics

Most platforms handle these basics in settings, so they're usually quick wins.

• Guest checkout. Don't make an account the price of buying, since 18% of abandoners left over exactly that. Offer to create one after the order, when the email and address are already filled in.

• Express wallets. Shop Pay, Apple Pay, Google Pay, and PayPal let a returning shopper skip the form, which matters most on mobile.

• Buy now, pay later. Options like Klarna, Afterpay, or Shop Pay Installments can help on higher-priced carts. Show them on the product page too, because "4 payments of $37" changes the math before the shopper reaches checkout.

Then trim the form. Cut anything you don't need to ship the order, turn on address autofill, and skip the phone number unless your carrier needs it. Near the pay button, show the cards you accept and your payment processor's security badge, and make it easy to reach you if something goes wrong.

Finally, check out on your own phone, on cellular data. Site errors caused 17% of abandonments in Baymard's survey, and a checkout that only breaks on one browser can go unnoticed for months.

Step 4: Answer the question before the shopper leaves

Checkout fixes help the shopper who has decided to buy and hits a snag. The shopper who is still deciding, and has a question, usually leaves quietly, and the cart shows up as abandoned with no reason attached.

The usual ways of answering have limits. An FAQ page helps if people find it. Live chat helps when someone's free to pick it up. Email gets answered tomorrow, and by then the shopper has bought from someone else.

An AI shopping assistant covers that gap. It sits on the product page and the cart and answers questions about sizing, delivery dates, and return windows from your own catalog and policies at any hour. When a different size or product is the better choice, it says so, and it hands off to a person with the full conversation when it needs to. If you're evaluating one, Gladly's shopping assistant pulls the latest product information from your site, and your team trains it in plain language, so keeping it current doesn't take a developer.

Two setups do the most for abandonment:

• Proactive outreach on the cart page. Trigger a message once a shopper has lingered, so they don't have to go looking for help. Gladly's Proactive Chat can trigger on the page URL and time on page, and you can prioritize visitors by cart value. Its best practices guide suggests prompting cart visitors who've spent more than two or three minutes without checking out.

• Answers on the product page. A lot of cart doubt starts on the product page. A prompt like "Questions about sizing?" on a product with a known fit issue catches the hesitation before the item goes in the cart, the same idea behind website personalization that responds to what the shopper is looking at.

Here's what that looks like in practice. Smith Optics reached a 67% resolution rate with Gladly's shopping assistant. At KÜHL, AI took over repetitive questions, which gave the team time for fit and product conversations, and revenue per call went up 120%.

Conversations are one lever, and they won't fix a $15 shipping surprise. They work on the part of abandonment that comes from doubt, which is the part checkout changes don't reach. Further out, agentic checkout lets that same conversation complete the purchase, so the shopper never fills in a form at all.

If you run a small team

The assistant is often your first real coverage outside business hours. Start with the ten pre-purchase questions that came up most in step 1, turn on one proactive campaign on the cart page, and send anything the assistant can't answer to whoever checks the inbox first.

If you run a larger support team

You probably staff chat already, so the gains come at peak times and from freeing agents for conversations that lead to a sale. Put proactive campaigns on your highest-intent pages, use the estimated wait time trigger so outreach only fires when someone can respond quickly, and route handoffs to the people best placed to close the sale.

Step 5: Recover the carts that still leave

Some carts will still get abandoned, and an automated recovery flow is the cheapest way to win part of them back. Most email platforms, from Klaviyo and Shopify Email to Braze, include an abandoned cart template to start from.

Set expectations first. In Klaviyo's analysis of more than 143,000 abandoned cart flows, the average flow turned 3.33% of recipients into orders, and the top 10% of flows reached 7.69% (2023 data). Those are small numbers per send, and they add up because the flow runs on every abandoned cart without anyone touching it.

A few cart abandonment email basics:

• Send the first email within a few hours. Klaviyo suggests two to four hours after the cart is abandoned, while the product is still on the shopper's mind.

• Plan on a short sequence of two or three emails. Use the later ones to answer the likely objection, whether that's a shipping cost or a worry about sizing.

• Save any discount for the last email in the sequence.

• Add one SMS message for shoppers who opted in to texts.

• Give people a way to reply. A reply-to address that reaches a person, or a link to chat, turns the email into a way to ask the question that stopped them.

Retargeting ads can reach shoppers who never gave you an email address, but you pay for every impression, so they usually come after email and SMS. For more on the email side, see our guide to abandoned cart emails.

Step 6: Measure what you actually recovered

Cart abandonment rate is a diagnostic. Revenue is what you're trying to move, so translate each fix into orders and dollars.

Here's an illustrative example. Say your store gets 50,000 sessions a month, converts at 2%, and has an $80 average order value. That's 1,000 orders a month, and at a 70% abandonment rate those orders come from about 3,333 carts.

• Cut abandonment by one point, to 69%, and the same carts produce about 1,033 orders. That's 33 more orders, or roughly $2,667 a month and $32,000 a year.

• Cut it by two points, to 68%, and you get about 1,067 orders. That's 67 more a month, roughly $5,333 a month and $64,000 a year, and your conversion rate moves from 2.00% to 2.13%.

Plug in your own sessions, conversion rate, and order value and the math works the same way. A few habits keep the measurement honest:

• Track cart abandonment and checkout abandonment separately. Showing shipping earlier should move the first, and a shorter form should move the second.

• Count chat as part of the funnel. Gladly's chat widget can send chat events to Google Analytics, including conversations started and proactive campaigns triggered, so you can segment purchase behavior by chat activity in GA.

If your traffic is lower

A few thousand sessions a month usually isn't enough to A/B test a checkout change. Make one change at a time, compare the same weeks year over year so a holiday spike doesn't get the credit, and give each change at least a full month before you judge it.

If your traffic is higher

Run real tests. Split traffic, let each test reach the sample size your testing tool calls for, and keep a small holdout group out of your recovery emails so you can see what the flow adds beyond the shoppers who would have come back anyway.

For testing method and traffic math, our ecommerce CRO framework goes deeper, and ecommerce conversion rate benchmarks will help you set a baseline. Once more shoppers are finishing checkout, raising average order value is the next lever to pull.

Where to start

To get started this week, run step 1, then pick the one fix that matches what your shoppers told you and measure it before you move to the next.