The first thing to understand about cart abandonment is that a large share of it isn't abandonment at all. The cart has become a general-purpose tool: a shortlist, a price-check, a wishlist, a reminder to come back on payday. Treating every unfinished cart as a lost sale means chasing people who were never buying today, and — worse — it hides the ones who were.
So before you install a recovery email, separate the two populations. They need opposite things.
Browsing carts and intent carts
A browsing cart is a bookmark. The person added an item to hold it, compare it, or think about it. They never entered checkout. They may return in a week. Recovery emails work reasonably well on this group, because the email is a reminder, and a reminder is what they wanted.
An intent cart belongs to someone who tried to buy and stopped. They entered checkout. They typed something. Then a specific thing happened, and they left.
Recovery emails are close to worthless for this second group. You're emailing someone who already decided to buy and then hit a wall you haven't removed — so the email invites them to hit it again. Every intent cart is a bug report.
The distinction is measurable: did they reach checkout? That single line splits your abandonment into "market to them later" and "fix this now."
What stops the intent carts
Almost always, a number they didn't expect.
Shipping revealed at the cart. The buyer priced the item on the product page. The cart quotes a different total. The gap isn't usually large — it's that it was withheld. A cost revealed late reads as concealment, and it converts a purchase decision into a trust decision, which you were not prepared to win.
Taxes, handling, and fees stacking on. The same mechanism, applied repeatedly, each one confirming the buyer's suspicion that the real price is a moving target.
A threshold that appears too late. "Add $12 more for free shipping" is a nudge to someone still shopping and an insult to someone who has finished. Placed after the decision, it re-opens the decision.
No delivery date. For anything time-bound — a gift, a part, a replacement — "when does it arrive" is a purchase-blocking question. A cart that can't answer it sends buyers to a store that can.
A forced account. Being asked to invent a password to hand over money is a demand for a relationship the buyer didn't ask for.
Notice the pattern: none of these are persuasion failures. They're information failures, and they're all failures of timing rather than content. The information was going to be revealed. It was revealed too late.
Prevention beats recovery, and it isn't close
Recovery reaches a fraction of abandoners, after a delay, in a channel they may not open. Prevention reaches all of them, at the moment of the decision.
The prevention list is short and mostly about moving information earlier:
Show shipping cost on the product page — or a clear, unconditional "free shipping over $X" banner sitewide, so the number is never new.
Surface taxes and handling as early as you can estimate them, rather than letting them accumulate at the total.
Give a delivery estimate early, and be specific enough to be useful.
State the free-shipping threshold while they're still adding items, when it's a nudge rather than a correction.
Offer guest checkout, and invite the account afterward, when there's an order to track.
Then check the whole sequence yourself, on a phone, as a new customer, and note the first moment a number changes. That moment is where you're losing intent carts.
Where recovery genuinely earns its place
For browsing carts, a well-made reminder is useful and welcome. Keep it honest: a discount in the first email teaches your regulars to abandon carts to receive one. Lead with the reminder, and hold any incentive for later, if at all.
And segment. Sending "you left something behind!" to someone who abandoned at your payment step because their card was declined is a slightly cruel non-sequitur.
The measurement that changes what you fix
Two stores with the same abandonment rate can have completely different problems: one is full of browsing carts and converting fine; the other is losing buyers at the shipping line. The rate cannot tell them apart. The split — cart abandons versus checkout abandons — can, and it's the single most valuable number here.
Beyond that split lies the field-level question: inside checkout, which specific input stopped them. That isn't visible from the outside, and it isn't visible in a funnel. It's visible only by watching real buyers, which is what the tracking snippet does. Defrixa's free scan will read your cart and checkout structurally first — costs revealed late, forced accounts, missing reassurance — and name the biggest one.
Common questions
No, and chasing the number down can be actively harmful. Some of it is normal shopping behaviour. The number worth watching is how many people enter checkout and don't finish.
Soon enough to catch the intent, late enough not to be creepy — within a few hours for the first, then spaced. But if you're recovering intent carts, fix the wall before you build the ladder.
Usually not. You'll train repeat customers to abandon carts deliberately, and you'll discount sales you would have made anyway.
Sometimes, as a last resort. But an exit-intent popup that offers free shipping is an admission that your shipping cost was the problem — which you could have solved earlier, for everyone, for free.