In short
Four carry most of the automated revenue for a typical store: a welcome sequence for new subscribers, a checkout abandonment sequence, a post-purchase sequence, and a win-back for customers who have gone quiet. Build those before anything else and build them properly rather than building ten thin ones. After that, the flows worth adding depend on your catalogue: back in stock and price drop for scarce or discounted ranges, replenishment for consumables, and list health flows that most stores never get to.
Build these four first, and build them well
The welcome sequence works because it reaches people at their most interested. Somebody who just subscribed is paying more attention than they ever will again, and a sequence that introduces the brand, explains why the products are made the way they are, and shows the range does more than a single discount code.
The checkout abandonment sequence recovers orders that were nearly complete. It has the highest revenue per recipient of anything you will build, because the audience had already chosen a product and started paying.
The post-purchase sequence turns a buyer into a repeat buyer. It reduces returns by helping people use what they bought, collects reviews, and sets up the second order. Stores under-invest here more than anywhere else.
The win-back reaches customers who bought once and stopped. It is cheap to run, it uses data you already have, and it targets the people most likely to buy again of anyone on your list. A single well-timed message to lapsed customers usually outperforms a month of campaigns.
The next tier, in the order most stores should add them
Browse abandonment comes next for stores with enough traffic from identified visitors. It reaches somebody who looked at a product and left without adding it, which is a weaker signal than a cart and still a real one. Keep it to one message and make it feel helpful rather than watchful.
Back in stock follows for any store that sells out. The people who asked to be told are the warmest audience you have, and the flow needs almost no persuasion, only speed.
Replenishment matters for consumables, and its value depends entirely on getting the interval right. Base it on how long the product actually lasts at normal use rather than on a convenient number of weeks.
A second-order flow deserves a place too. The gap between the first and second purchase is where most customers are lost, and a sequence aimed specifically at that transition, with a reason to return rather than a discount, changes the shape of the whole customer base.
Some flows only make sense for certain catalogues
Price drop notifications suit stores with a genuine sale cycle and a catalogue people watch. They suit premium brands poorly, because they train customers to wait.
Size and fit follow-ups suit apparel and footwear, where a message before delivery explaining how the item runs, or after delivery explaining the exchange process, reduces returns and the cost that comes with them.
Subscription flows are their own category: upcoming charge notices, skip and swap prompts, and a save sequence before a cancellation completes. These affect retention more than any campaign and they are frequently left at the app defaults.
Care and usage flows suit anything with a maintenance cycle: leather, equipment, plants, machinery. They are useful, rarely commercial in tone, and they build the sort of relationship that produces the next purchase without asking for it.
List health flows are the ones nobody builds
A re-engagement sequence for subscribers who have not opened anything in months protects the deliverability of everything else you send. Sending repeatedly to people who never engage is one of the surest ways to end up filtered.
Pair it with a sunset rule. If somebody does not respond to the re-engagement attempt, stop sending campaigns to them. It feels like shrinking the list, and it improves the results of every message afterwards.
A preference flow gives people a way to hear from you less often rather than not at all, which converts some unsubscribes into a quieter relationship. It also gives you data on what people actually want to receive.
And a data hygiene flow, cleaning bounces and invalid addresses, belongs in the same group. None of these produce revenue directly, and all of them determine whether the four flows in the first section reach anyone at all.
Build order for a store starting from nothing
- 1
- Welcome sequence, three to five messages
- 2
- Checkout abandonment, three messages
- 3
- Post-purchase, anchored to delivery
- 4
- Win-back for lapsed customers
- 5
- Browse abandonment and back in stock
- 6
- Replenishment or second-order sequence
- 7
- Re-engagement and sunset rules
An illustrative order for a store with limited time. Doing the first four properly returns more than doing all seven thinly, which is the usual failure mode.
Related questions
How many messages should a welcome sequence have?
Three to five, spread over a week or so. One message cannot introduce a brand, explain what makes the products different and show the range without becoming a wall of text. More than five and the later messages arrive after the initial interest has faded.
Do flows replace regular campaigns?
No. Flows reach people at the right moment in their own timeline, and campaigns reach the list at moments that matter to your business, such as a launch or a season. Most stores earn a large share of automated revenue from flows and still need campaigns for anything time-bound.
Which flow should a store with limited time build first?
Checkout abandonment, because the audience is closest to buying and the revenue per recipient is highest. Then the welcome sequence, because it feeds everything else by turning new subscribers into first-time customers rather than leaving them to decide on their own.