Cart abandonment and checkout abandonment are two different numbers
Cart abandonment is the share of shoppers who put something in a basket and never paid. Checkout abandonment is the share who started the checkout and never paid. Both use paid orders as the numerator, so the only thing separating them is the denominator - and that difference is large enough that quoting one figure while meaning the other will misdirect a whole quarter of work.
The reason is structural. Adding to a cart is a low-commitment act: on most stores it is how people compare prices, check shipping, or save an item for later. Starting a checkout is a high-commitment act, because the shopper has agreed to hand over an address and a card. So the population inside checkout is far more intent-heavy than the population inside the cart, and its failure rate is correspondingly lower. If your cart abandonment is 75% and your checkout abandonment is 55%, those two numbers are consistent, not contradictory.
Which one you should manage depends on where the loss is. A high cart rate with a low checkout rate means the problem sits before checkout - shipping costs revealed late, price comparison, an unclear cart page. A cart rate and a checkout rate that are close together means most people who add an item do reach checkout, and the loss is concentrated in the payment flow itself: forced account creation, a slow form, a declined card, an unexpected tax line. This calculator prints both, plus the cart-to-checkout drop-off that sits between them, so you can tell those two stories apart.
Baymard Institute's aggregation of published studies puts the average documented online shopping cart abandonment rate close to 70%, which is a useful sanity check on your own tracking: a figure far below that usually means your cart event is only firing for high-intent shoppers, and a figure near 100% usually means your order event is not firing at all.
The formula, and why the recovery figure is an assumption rather than a result
Each rate is one minus a survival ratio. Divide the orders by the stage count, subtract from one, multiply by 100. Written out: cart abandonment is (1 − o ÷ c) × 100 and checkout abandonment is (1 − o ÷ k) × 100. The cart-to-checkout drop-off in the middle is (1 − k ÷ c) × 100.
The three do not add up, and that trips people up. They are conditional rates on different bases: in the worked example below, 45.00% of carts never reach checkout, and 54.55% of the ones that do reach it still fail to pay. Combine them by multiplying the two survival ratios: 0.5500 × 0.454545 = 0.2500, so 25.00% of carts convert and the cart abandonment rate is 75.00% - exactly the figure the direct division gives. Multiply survival ratios; never add drop-offs. Adding 45.00 and 54.55 gives 99.55, a number with no meaning at all.
The money side has two layers, and only the first is arithmetic. The value sitting in abandoned carts is (c − o) × V. That is a real, checkable figure. The recoverable revenue multiplies it by an assumed recovery rate r, and that rate is an input you supply, not something this calculator can know. Any tool that hands you a recovery number without asking you for the rate has silently chosen one for you.
Treat the recovery rate the way you would treat a discount rate in a valuation: state it, defend it, and test the answer against alternatives. If you already run a win-back flow, do not assume - divide the flow's attributed revenue by the value that was in the carts it targeted, and you have your own r. Then multiply by gross margin, because a recovered order carries its cost of goods with it. On a 45% margin, $73,800 of recovered revenue is $33,210 of gross profit, and the campaign cost has to come out of the second figure.
One measurement caution built into the arithmetic: abandonment counts sessions or carts, while a single shopper can abandon three times in a week and buy on the fourth visit. A session-scoped abandonment rate therefore overstates how many people walked away. Keep the scope consistent between numerator and denominator and read the rate as a property of visits, not of customers.
Worked example: 12,000 carts, 6,600 checkouts, 3,000 orders
A store logs 12,000 carts created in a month, 6,600 checkouts started, and 3,000 paid orders. Average order value is $82, gross margin is 45%, and the win-back flow historically converts 10% of the carts it emails.
- Cart abandonment. 3,000 ÷ 12,000 = 0.25 survive, so 1 − 0.25 = 0.75 = 75.00%.
- Checkout abandonment. 3,000 ÷ 6,600 = 0.454545 survive, so 1 − 0.454545 = 54.55%.
- Cart-to-checkout drop-off. 6,600 ÷ 12,000 = 0.55 survive, so 45.00% never reach checkout.
- Check that they reconcile. 0.55 × 0.454545 = 0.25, and 1 − 0.25 = 0.75. The two stage rates multiply back to the headline figure.
- Abandoned carts. 12,000 − 3,000 = 9,000.
- Value in those carts. 9,000 × $82 = $738,000. This is the figure vendors like to quote; it is not money you were ever going to collect.
- Recoverable at 10%. $738,000 × 0.10 = $73,800 of revenue.
- Gross profit on the recovery. $73,800 × 0.45 = $33,210. Subtract the flow's cost from this, not from the $73,800.
Now find where the bigger prize is. Cutting the cart-to-checkout drop-off from 45% to 40% would send 12,000 × 0.60 = 7,200 shoppers into checkout instead of 6,600. At the same 45.45% checkout survival that is 7,200 × 0.454545 = 3,273 orders, or 273 more - worth 273 × $82 = $22,386 of revenue a month. Cutting checkout abandonment from 54.55% to 50% instead would give 6,600 × 0.50 = 3,300 orders, or 300 more, worth $24,600. The two projects are comparable in size, so the tie-breaker is which is cheaper to build.
How to read your two rates together
Read the pair, never the headline alone. The gap between cart abandonment and checkout abandonment tells you which half of the journey is losing the money, and the gap is the actionable part.
When the cart rate is much higher than the checkout rate, the loss is concentrated before checkout. The usual culprits are cost surprises revealed at the cart step, shipping thresholds that are unclear until the last moment, and the fact that many shoppers use the cart as a shortlist. Fixes are informational: show delivered cost earlier, show stock and delivery date on the product page, make the saved-cart function explicit so comparison shopping does not register as failure.
When the two rates are close, almost everyone who adds an item reaches checkout and the loss is inside the payment flow. That is a technical problem with technical fixes: guest checkout, address autocomplete, fewer required fields, retryable card declines, a wallet option, and a checkout that works on a mid-range phone on a slow connection.
Segment before you act. Split both rates by device and by whether the shopper is a returning customer. Mobile checkout typically carries a higher abandonment rate than desktop on the same store because the form burden is heavier on a small screen, and a blended rate hides which platform is carrying the loss. Compare each segment against its own trailing figure rather than against an external benchmark, and attach a margin of error before you call a movement real - the method is in the conversion rate calculator.
Finally, be honest about what a win-back flow does. Some of the orders it is credited with would have arrived anyway, because a shopper who saved a cart on Tuesday often returns on Friday. The only way to separate incremental revenue from re-attributed revenue is to hold out a control group and read the difference with the A/B test significance calculator, sizing the holdout first with the A/B test sample size calculator.
Recoverable revenue per 1,000 abandoned carts
| Average order value | 2% recovered | 5% recovered | 10% recovered | 15% recovered |
|---|---|---|---|---|
| $40 | $800 | $2,000 | $4,000 | $6,000 |
| $60 | $1,200 | $3,000 | $6,000 | $9,000 |
| $80 | $1,600 | $4,000 | $8,000 | $12,000 |
| $100 | $2,000 | $5,000 | $10,000 | $15,000 |
| $150 | $3,000 | $7,500 | $15,000 | $22,500 |
These are revenue figures. Multiply by your gross margin to get the profit a flow has to cover its own cost from - at a 45% margin the $8,000 cell is $3,600 of gross profit.
Where abandonment numbers go wrong
- Quoting the cart rate as though it were the checkout rate. They differ by tens of points on most stores, and the fixes they point to are completely different.
- Adding the stage drop-offs. 45% plus 54.55% is not 75%. Multiply survival ratios: 0.55 × 0.454545 = 0.25.
- Treating the value in abandoned carts as lost revenue. It is the size of the pool, not a forecast. Only the recovery-rate line turns it into money, and that rate is your assumption.
- Taking credit for orders that would have happened anyway. Without a holdout group, a win-back flow's attributed revenue includes shoppers who were already coming back.
- Counting sessions and calling them people. One shopper abandoning three times before buying registers as three abandonments and one order.
- Letting express checkout skip the cart event. A wallet button that jumps straight to payment produces orders with no cart, which pushes the measured cart abandonment rate down for no real reason.
- Comparing your rate to an aggregate benchmark. Free-shipping thresholds, price point, and how easily your cart doubles as a wishlist move the rate more than checkout quality does.
- Forgetting margin. A recovery campaign is funded out of gross profit, so a 45% margin turns $73,800 of recovered revenue into $33,210 to work with.
Abandonment is not the same as lost demand
A cart that never converts is not automatically a customer you lost. Some of those shoppers bought from you on a later visit and are counted as an order in a different session; some were never going to buy at any price; some were bots. The measurable, honest version of the metric is a statement about visits: three out of four carts created this month did not end in an order in the same visit. Any stronger claim needs a customer-level analysis, which is what the customer lifetime value calculator and the customer retention rate calculator are for.
What to measure next
Abandonment tells you the size of a leak at one point in the journey. Three neighbouring measures tell you whether fixing it is the best use of your time.
Order value. Recovered revenue scales linearly with average order value, so a 10% lift in order value is worth exactly as much to your recovery figure as a 10% lift in recovery rate. Free-shipping thresholds and bundles move both the cart rate and the order value at once, in opposite directions sometimes; measure both with the average order value calculator.
Acquisition cost. Every abandoned cart was paid for once already through media spend. Reducing abandonment lowers your effective cost per order without buying a single extra click, which is the same mechanism the cost per acquisition calculator and the ROAS calculator measure from the spend side.
Whole-funnel structure. Cart and checkout are the last two stages of a longer chain that starts with a session and a product view. Modelling the full sequence in the sales funnel drop-off calculator shows whether your worst stage is really checkout or whether it is upstream, and the multiplicative arithmetic there is the same as the reconciliation step in the worked example above.
One last practical note on payment: card declines and processing costs both sit inside the checkout stage and both show up as abandonment or as margin loss rather than as a separate line. If your checkout abandonment is unusually high on one payment method, price the fee side with the payment processing fee calculator before you decide the flow is at fault.