Here’s a conversation we have more often than we’d like. A store owner wants help with ads. We look under the hood and find no welcome series, no checkout recovery, nothing after the receipt. Which means every dollar of ad spend is renting a visitor once, with no system for keeping them.
Ads pour traffic in. Flows stop it pouring straight back out. Build the bucket before you pay for the water.
The good news is that “the bucket” is five automations, they run themselves once built, and none of them requires you to become an email person. Here they are, in the order we’d build them, with the email counts that current benchmarks actually support. Not the counts that sound impressive. There’s a difference, and it surprised us too.
1. Abandoned checkout (build this first)
Someone put items in their cart, started paying you, and stopped. These are the warmest leads that exist anywhere in your business, and most stores send them nothing.
What good looks like: 2 to 3 emails. Not 7. Klaviyo’s own research across thousands of stores found 2 to 3 messages is the sweet spot, and more actually performs worse. First email within a few hours, a plain reminder showing the exact items left behind. If you use a discount at all, hold it until the second or third email. Offering it immediately trains your entire customer base to abandon carts on purpose, which is a customer training program nobody meant to run.
Why first: fastest payback of the five. You are recovering sales that were seconds from happening.
2. Welcome series
What new subscribers get after they hand over their email. If your answer is “the discount code and then silence,” you’re introducing yourself with a coupon and then ghosting, which is a strange way to start a relationship you asked for.
What good looks like: 4 to 7 emails over one to two weeks. Deliver the promised incentive instantly, then use the rest of the series to do the actual introducing: your story, your best products, real customer proof, and answers to whatever stops people buying. Split it so existing customers don’t get the “nice to meet you” treatment.
Why second: highest engagement you will ever get. People read these because they just asked to.
3. Post-purchase
The sale is not the end. It’s the audition for order two, and order two is where DTC economics stop being painful.
What good looks like: 3 to 5 emails over the weeks after delivery. Thank them like you mean it, help them actually use the thing, ask for a review once they’ve had time to form an opinion, and point to the natural next purchase. Split first-time buyers from repeat customers, because “welcome to the family” reads differently on order six.
4. Browse abandonment
The gentle one. They looked at a product and left, without carting anything. Intent is real but light, so the flow should be too.
What good looks like: 1 to 2 emails, maximum. Show the actual product they viewed, not a generic “still thinking it over?” And make the flow exit cleanly the moment they add to cart, so your emails don’t chase people who already moved on to the next step. Two is the ceiling here, not the floor. This is the flow most likely to tip from helpful into “the website is following me.”
5. Win-back
Customers who bought, then went quiet. They already trusted you once, which makes them cheaper to recover than strangers are to acquire.
What good looks like: 2 to 3 emails, and Klaviyo’s guidance caps it at three. The trick is timing, not volume: trigger it based on your actual buying cycle. A shampoo customer is “lapsed” after two months. A furniture customer is just living with their couch. End the sequence with a graceful exit toward your sunset process rather than emailing the unresponsive forever, which helps nobody and hurts your deliverability.
The order matters more than the speed
Notice the pattern in all five: the modern answer is almost never “more emails.” It’s the right two or three, aimed properly, with an exit. The era of winning by volume is over, partly because inboxes got smarter and partly because everyone tried volume and readers noticed.
If you want to know exactly where your setup stands, we built a free Flow Grader that scores all five flows plus the foundations underneath them, and hands you a fix-first list in payback order. Takes two minutes, no email required to see your results, and it will be honest with you, because we don’t know how to make tools that aren’t.
And if the grade stings, that’s fine. Every store starts at F. The whole point of flows is that you only have to fix it once.