How to build a Klaviyo abandoned cart flow that actually recovers revenue

Abandoned cart vs abandoned checkout: get this right first

In Klaviyo these are two different triggers, and most brands wire up the weaker one.

Started Checkout fires when a shopper reaches the checkout page and Klaviyo has captured their email. Added to Cart fires earlier, when they add an item but may never have shown real intent. Started Checkout is the higher-intent event and the one that recovers the most revenue per send, so it should be your primary trigger.

If your store sells on Shopify, the Started Checkout event flows in automatically once Klaviyo is connected. Cart-level data is patchier and depends on your theme and any onsite tracking, which is one reason a clean Klaviyo Shopify integration matters before you trust the numbers.

Trigger and timing: when each email should send

The flow trigger is the Started Checkout metric. From there, timing is everything.

• Email 1: 1 hour after Started Checkout. Long enough that genuine buyers have often finished, short enough to catch the rest while intent is warm.
• Email 2: 24 hours later. The forgetters and the on-the-fence.
• Email 3: roughly 48 to 72 hours after that. The last call, and the only place a discount belongs.

Add a flow filter so anyone who places an order exits immediately, and a conditional split that skips the rest of the sequence once a purchase happens. Without these, you email people who have already bought, which is the fastest way to earn unsubscribes and spam complaints.

Use smart sending to avoid colliding with campaigns, but do not let it suppress a high-intent cart email by accident. Check the smart sending window against your campaign cadence.

The 3-email sequence that converts

Each email has one job. Do not make all three the same email with a different subject line.

Email 1: the reminder

No discount. Show the exact product with a clear image, a single button back to the checkout, and one line of reassurance. Most of your recovered revenue comes from people who simply got distracted, and they do not need a code to come back.

Email 2: handle the objection

This is the email most brands skip, and it is where the money hides. Answer the real reason people stall: shipping cost and speed, returns, sizing, or trust. Reviews and a short FAQ block do more here than any image.

Email 3: the nudge

Now you can add urgency or a small incentive, because the people still here are the ones who needed a reason. A modest code or a genuine low-stock line works. Do not lead with the biggest discount you can afford, or you train customers to abandon on purpose.

If you are deciding what else to automate after this, the order of priority is set out in our guide to ecommerce email automation.

Common mistakes that quietly cost you

• Discounting in email one. You hand money to buyers who would have converted anyway, and you teach the rest to wait.
• One email instead of three. A single send catches only the people already coming back. Emails two and three do the actual selling.
• No purchase or exit logic. Emailing people who already bought is a deliverability problem, not just an awkward one.
• Ignoring deliverability. A recovery email that lands in the Promotions tab or spam recovers nothing. Sender reputation and authentication come first, which is why we treat email deliverability as part of the build, not an afterthought.
• Set and forget. Cart flows decay. Review the copy, offer and timing every quarter.

Benchmarks and when to A/B test

Treat published benchmarks as a sanity check, not a target. For DTC abandoned cart flows you are usually looking at:

• Open rate: 40 to 55 percent (high, because intent is high).
• Click rate: 5 to 10 percent.
• Recovery rate: 5 to 11 percent of abandoned checkouts, depending on price point and traffic quality.

Lower price points and impulse buys recover faster. Considered, high-ticket purchases recover slower but are worth more per save.

Once the flow is live and has volume, A/B test one variable at a time: the email-one subject line first, then the email-two objection angle, then whether email three needs an incentive at all. Give each test enough conversions to mean something before you call it. The how is covered in our Klaviyo A/B testing guide.

None of this is rocket science on its own. Doing all of it well, then keeping it tuned every quarter while you actually run the business, is the part founders underestimate. It is exactly the work we take off their plate, and where most of the recovered revenue comes from.