Email Automation Flows: Welcome, Cart, Post-Purchase, Winback

Flows are where email stops being a newsletter and starts being a system. Campaigns can have a bad week. A broken welcome is a leak every hour.

This is the build order we use on email marketing retainers: the four (plus VIP) email automation flows that show up in almost every ecommerce account, and the mistakes that make them look busy while they print nothing.

If your Klaviyo still has default templates in these flows, start with a free email audit.

The only order that does not waste the month

People want a 12-flow diagram. They need four flows that actually fire.

Browse, back-in-stock, sunset, and site abandon are real. They are not week-one work. Lifecycle marketing 101 is the map. This article is the plumbing.

Welcome: stop sending a coupon and calling it a brand

A default Klaviyo welcome is a logo, a 10% code, and a grid of bestsellers. That teaches the list that the only reason to stay is a discount. Then every campaign has to beat 10%.

A welcome that works usually has four to six emails over two weeks, not twelve. Jobs:

1. Confirm the world they joined. What you sell, who it is for, what you will never do (daily spam, fake urgency).
2. One path to first purchase. A hero product or a quiz, not the entire catalog.
3. Proof. Reviews, UGC, a founder note if it sounds like a person.
4. A reason to stay without a code. Education, a drop calendar, a standard.

Filter: exclude people who already bought. They belong in post-purchase, not a “here is 15% off your first order” they cannot use.

{% if event.extra.line_items %}
  {# they already checked out — skip this welcome branch #}
{% endif %}

In Klaviyo you do this with a flow filter (Placed Order zero times since starting this flow), not a Liquid stunt. The point is the same: do not talk to buyers like they are still strangers.

Length: if you cannot explain the series on one whiteboard, it is too long. See common email mistakes.

Abandoned cart: speed, then reason, then (maybe) a code

Cart is not a morality play. People leave carts. Your job is to be the reminder that does not feel like a mugging.

Email 1 (1 hour): the cart. Image, price, one button. No essay.

Email 2 (24 hours): objection. Shipping, returns, size, “still in stock.” This is where copy earns its keep.

Email 3 (optional, 48–72 hours): a reason. Not automatically 10% off. If your margin and brand can survive a code, use it here, once. Luxury lists often should not.

SMS can take email 1 if they opted in — that is SMS vs email strategy. Do not fire both at +1 hour with the same line.

Exclude purchasers. Exclude people who emptied the cart. Exclude unengaged-from-email-for-months if you care about inbox placement.

Post-purchase: the second order is the product

First purchase is expensive. The flow after it is where LTV happens.

Day 0–1: receipt energy. What they bought, when it ships, how to reach a human. This is not a cross-sell slot.

Day 3–7: how to use it. Returns drop. Reviews go up later because they actually used the thing.

Day 10–21: the pair or the refill. This is merchandising, not “you might also like” scraped from a feed with no logic.

Day 21–45: replenishment if the product is consumable. 21 days is a starting point, not a religion — map to real usage. The beauty case study on this site is the repeat purchase version.

Ask for a review after they have had time. Asking at “order confirmed” is how you get one-star shipping complaints.

Winback: a reason, then silence

Winback is where brands panic. Three emails, three increasing discounts, then a fourth “we miss you.” The customer did not miss the coupon. They missed a reason to care.

Better shape:

1. What is new since they last bought (product, formula, fit, restock of their SKU).
2. A specific use they may have skipped.
3. One offer if the economics work — not a ladder.

Then stop. Unengaged people go to sunset and suppression. Mailing them “because they are on the list” is how you hurt the domain for the people who still buy.

VIP should not be in the same winback as a one-time $32 buyer. Segment first.

How to know a flow is working

A/B tests: one variable, enough volume. Subject lines in cart email 1 are fair. Rewriting the whole post-purchase based on 200 sends is superstition. See A/B testing.

Implementation notes that save you a rebuild

• One metric owner per flow. If five people can edit welcome, nobody owns it.
• QA as a customer. Place a test order. Abandon a cart. Use a real phone.
• Don’t stack discounts. Welcome 15% + cart 10% + campaign 20% is a margin bug.
• Design last. A true mobile template on a smart flow beats a pretty collage on a default. Mobile email design comes after the logic.

If you want these flows built in Klaviyo the way an operator would maintain them — not a 40-email diagram — talk to an expert.