Roughly seven in ten online shopping carts are abandoned — a figure the Baymard Institute has tracked across dozens of studies for years. That means the majority of the people who liked your product enough to add it to a cart still walked away. A well-built cart recovery flow is how you win a meaningful share of them back, and it's one of the highest-return automations in ecommerce. But most advice stops at "send an abandoned cart email," and a single templated reminder leaves most of that revenue on the table.
A flow that actually converts is a designed sequence, not a send: the right message doing the right job at the right moment, across the right channels — and it starts before the cart is ever abandoned. This guide walks through the whole thing step by step, and it's deliberately platform-agnostic. Whether you run Shopify, WooCommerce, or any email and SMS tool, the principles are the same; only the buttons differ.
First, know what you're actually recovering
"Abandonment" is really three different behaviours, and they call for different responses. Blurring them together is the first mistake most stores make.
Browse abandonment — someone viewed products but never added to cart. Lowest intent; recover with gentle "still looking?" nudges and recommendations, only if you have their consent.
Cart abandonment — they added items but never reached checkout. Clear intent; the classic recovery flow.
Checkout abandonment — they started checkout, entered details, then stopped. Highest intent and often the most recoverable, because something specific blocked them at the final step.
The higher the intent, the more urgent and the more valuable the recovery — and the more likely a specific, fixable obstacle is to blame. That's why the most profitable move often isn't a better email at all.
The best recovery is preventing abandonment
Before you build a single message, look at why people are leaving, because no email undoes a broken checkout. Baymard's research is consistent on the culprits: once you set aside the large share of shoppers who were simply browsing or not ready to buy, the top reasons are unexpected extra costs (shipping, taxes, fees) at checkout, being forced to create an account, a checkout that's too long or complicated, worries about payment security, too few payment options, and slow delivery estimates.
Notice that almost every one of these is something you control. Surface shipping costs early instead of springing them at the final step; offer guest checkout; trim the number of form fields; add the payment methods your customers actually use; and show trust signals where the card details go. Much of this is upstream of the cart entirely — it lives on the product page and in the checkout experience, and finding those leaks is exactly what a proper funnel audit is for.
The order of operations Fix why people abandon first. A recovery flow should win back the ones you couldn't prevent — not paper over a checkout that's quietly leaking sales.
The anatomy of a flow that converts
With prevention handled, the flow itself. The core idea: it's a short sequence in which each message has a distinct job, spaced over a day or two, ideally across both email and SMS. Because the whole thing runs on triggers and timing, it lives in your email marketing automation platform, whichever one you use. A single "you left something behind" email is fine; a sequence that reminds, then reassures, then — only if needed — incentivises, converts far more. The jobs, in order, are remind, reassure, and incentivise, and that order is the whole secret.
Step-by-step setup
Step 1 — Capture identity and consent early
You can't recover a cart from an anonymous visitor. The flow only works if you've captured an email or phone number and permission to use it, which is why cart recovery is legal in the first place: it runs on first-party data given with consent. Collect it as early as you reasonably can — a signup offer, an email field early in checkout — so a later abandonment still leaves you a way to reach out. This is the same owned-audience groundwork covered in building an email list from scratch.
Step 2 — Set the trigger and the exit conditions
The flow should fire when someone abandons and, just as importantly, stop the moment they no longer need it. Set clear exit conditions: remove anyone who completes the purchase, and suppress people who've abandoned very recently so you're not hitting the same inbox repeatedly. Getting the exits right is what keeps a helpful nudge from becoming harassment.
Step 3 — Message one: the friendly reminder (about 1–4 hours)
Send the first message within a few hours — long enough that you're not pouncing, soon enough that intent is still warm. Keep it simple and human: assume they were distracted, show exactly what's in their cart with a clear image and a one-tap link back, and resist any urge to discount. Most recovered revenue comes from this message doing nothing more than jogging a memory.
Step 4 — Message two: reassure and handle objections (about 24 hours)
If they haven't returned, the second message addresses why — the doubt that stopped them. This is where social proof (reviews and ratings), a clear returns policy, shipping and payment reassurance, and answers to common questions do their work. If there's genuine urgency — low stock, an ending offer — say so honestly here; manufactured countdowns erode trust. The job is to remove the friction, not to bribe past it.
Step 5 — Message three: the incentive, last and sparingly (about 48–72 hours)
Only now, if your margins allow, introduce an incentive — and prefer free shipping to a percentage off, since unexpected costs are a leading reason people abandon in the first place. Reserve richer offers for high-value carts where a single conversion justifies the giveaway. Many stores skip this message entirely for low-value carts, and they're right to.
Step 6 — Layer in SMS where you have consent
For a high-intent moment like an abandoned cart, a short text can outperform an email, landing where people actually look. Add it where you have explicit permission, and mind the rules — SMS consent is stricter than email (double opt-in in many regions, tight limits on how many messages and how soon). Used well, a single well-timed text slotted between emails lifts the whole flow.
Step 7 — Personalise dynamically
The flow should pull in the shopper's name, the exact items they left, and relevant recommendations automatically, and adapt to context — a back-in-stock note if an item sold out, different messaging for a first-timer versus a loyal customer. Modern tools make this straightforward, and the payoff in relevance is real, as we cover in how AI is changing email personalisation.
Step 8 — Protect deliverability and consent
None of this matters if the messages land in spam. Send only to people who opted in, suppress recent purchasers and anyone with an open support issue, and don't over-mail — engagement-based sending protects your sender reputation under the bulk-sender rules now enforced by the major inbox providers. A clean, permission-based list is what keeps a recovery flow deliverable over the long run.
| Message | Timing | Channel | Job | Incentive? |
|---|---|---|---|---|
| 1 | ~1–4 hours | Friendly reminder, show the cart | No | |
| 2 | ~24 hours | SMS or email | Reassure, handle objections, social proof | No |
| 3 | ~48–72 hours | Incentive (prefer free shipping), high-value carts | Sparingly |
The discount trap
The single most expensive mistake in cart recovery is leading with a discount. Put a code in message one and you teach your customers a lesson they'll remember: abandon the cart, wait, and a coupon appears. Over time you erode margin on shoppers who would happily have paid full price, and you manufacture abandonment on purpose. That's why the incentive comes last, sparingly, and ideally as free shipping or a one-off, new-customer-only offer rather than a standing discount. Remind first; reassure second; pay only when you must, and only where the maths works.
Measure what actually matters
Open rates flatter you; they don't pay the bills. Judge a recovery flow on recovery rate and revenue per recipient — and, above all, on incrementality. Here's the uncomfortable truth most dashboards hide: some people who abandon would have come back and bought anyway, with or without your emails. If you credit the flow with every one of those sales, you'll wildly overstate its value. The fix is a small holdout group that receives no messages; the gap between the treated group and the holdout is the revenue the flow genuinely added. That's the honest number, and it's the mindset we argue for in rethinking success metrics — measure the real lift, not the flattering one.
It also keeps you disciplined on incentives: if the discounted messages are mostly converting people who'd have bought at full price, they're costing you margin, not making it. Cart recovery is one channel in a wider store; the same proof-and-measurement rigour applies whether the sale closes by email, on a product page, or through newer routes like shoppable video and the social commerce channels driving sales in 2026.
The bottom line
A cart recovery flow that converts isn't a clever email — it's a system. Fix the reasons people abandon so the flow only has to catch the ones you couldn't prevent. Capture consent early, then run a short, permission-based sequence where each message has a job: remind, reassure, and only then incentivise. Add SMS where you're allowed, personalise the contents, protect your deliverability, and refuse to lead with a discount that trains customers to game you. Then measure the incremental revenue, not the vanity metrics, so you know what the flow is truly worth. Do that, and you'll turn a big chunk of that seven-in-ten into paying customers — profitably, and for the long run.
Turn abandoned carts into revenue — without training customers to wait for a discount.
We build and optimise cart recovery flows and the wider store experience that keep your margins intact.
Explore E-commerce Marketing →Frequently asked questions
How many emails should a cart recovery flow have?
Three is a sensible starting point, and a multi-message series reliably outperforms a single send. Give each message a distinct job: the first is a friendly reminder, the second reassures and handles objections, and the third — only if margins allow — introduces an incentive. Extend to four or five for high-value carts, but keep testing, and stop messaging anyone who has bought or clearly isn't returning.
When should the first abandoned cart email be sent?
Usually within a few hours — commonly one to four hours after abandonment. Sending instantly can feel intrusive and may catch someone briefly distracted, while waiting too long lets the intent go cold. Space the follow-ups over the next day or two (roughly 24 hours, then 48 to 72), so the sequence feels like helpful nudges rather than pestering.
Should cart recovery emails include a discount?
Not in the first message, and not every time. Leading with a discount trains customers to abandon deliberately and wait for the code, eroding margin on people who'd have paid full price. Save any incentive for later in the sequence, use it sparingly, and consider free shipping instead of a percentage off — since extra costs are a leading reason people abandon. Reserve richer incentives for high-value carts.
Can I legally send abandoned cart emails?
Yes, when you have consent. Cart recovery messages rely on first-party data — information a shopper gives you after opting in — which is compliant with privacy rules. You need a captured email or phone number and permission to use it, so collecting consent early (at signup or checkout) is what makes the flow possible. SMS has stricter requirements, including double opt-in in some regions.
How do you measure if a cart recovery flow is working?
Look past opens to recovery rate and revenue per recipient, and — most importantly — measure incrementality. Some people who abandon would have returned anyway, so raw recovered revenue overstates the flow's impact. A small holdout group that receives no messages shows how much revenue the flow actually added on top of what would have happened, which is the number that matters for the bottom line.
