Guide · Lifecycle
The email and SMS flowsthat actually carry revenue
Half of all second orders land within thirty days of the first. Three quarters land within ninety. Most brands spend that entire window silent, then buy the same customer again from Meta. This is what to send inside it, message by message, and which channel each one belongs on.
01 · Why timing beats copy
The second orderhas a schedule
Roughly one in four to one in five customers ever places a second order. Published twelve-month repeat rates across DTC run 25 to 30 percent, and a February 2026 analysis of 156,110 customers put the aggregate closer to 19 percent. Those are the odds.
What makes them actionable is that the ones who do come back are not spread evenly across the year.
When second orders actually happen
The median time to a second purchase clusters between roughly fifteen and thirty-five days. So the decisive period is short, specific and entirely predictable, and it happens while the customer still remembers ordering.
Most brands go quiet on day two and start advertising again on day sixty.
There is one more finding that changes what you send inside the window. About 77 percent of second orders are a reorder of the same product rather than a cross-sell. The instinct is to introduce the range. The data says ask whether they want another one.
That is the whole case for building flows rather than scheduling campaigns. A campaign goes out when someone remembers to send it. A flow fires on day fourteen because day fourteen is when this particular customer is deciding, and it does that for every customer independently, forever.
The revenue split follows. Automated flows are about 5 percent of the messages a typical ecommerce brand sends and produce roughly 41 percent of email revenue, at close to eighteen times the revenue per recipient of a campaign. On SMS the pattern is sharper still: flows are 7.6 percent of sends and 45 percent of revenue. The economics of that sit on the email page. What follows here is the build.
02 · The reference
Every flow,message by message
Pick a flow. Each message shows when it fires, which channel it belongs on and what job it does. Enter two numbers and it sizes the flow against its published conversion band, where one exists.
Welcome series
Conversion bands are published targets for a well-built flow rather than averages across all brands, and where no figure is published consistently the tool says so instead of inventing one. The sequences themselves are how we build them. Nobody publishes message-level timings, so treat those as a starting architecture to test against rather than as a benchmark. The revenue-share weights match the ledger on the email page so the two do not contradict each other.
03 · The cross-channel rule
When a messageearns an SMS
Email costs nothing to send. SMS costs a carrier fee and, more importantly, costs subscribers. So every text has to clear a bar that an email does not, and four questions decide it.
Does three minutes beat three hours?
SMS is a speed instrument. If the outcome changes because the message arrived immediately rather than at the next inbox check, it earns a text. Cart recovery, back in stock and a genuine deadline pass. A brand story does not.
Is the intent already high?
Someone who reached checkout has told you something. Someone who looked at a product page has told you very little. That is why browse abandonment stays on email in our builds and cart abandonment does not.
Is it marketing or is it the order?
Shipping and delivery notifications ride on the consent given at checkout. Add a promotional line to one and it becomes a marketing message needing the higher written consent standard. Keep them separate. The consent rules are here, and this is not legal advice.
Can it fit in one line?
If the message needs explaining, it is an email. The text should be one sentence and a link, because anything longer is being read on a lock screen by someone who did not ask for a newsletter.
Applied across the six flows, this produces a programme that is mostly email with a handful of texts in the places where speed genuinely changes the outcome. Opt-out rates stay in the 0 to 1.5 percent band that marks a well-run programme, and the list keeps growing faster than it burns.
04 · Sequencing
Build themin this order
Not by revenue potential, by time to first dollar. Every flow below runs on traffic and customers you have already paid for, so the only question is which one starts paying soonest.
- Cart and checkout abandonment. Highest intent, highest recovery rate, and it is running on people who were seconds from buying. A single reminder recovers 2 to 3 percent. A multi-step sequence across email and SMS recovers 8 to 12 percent, and the difference is the number of steps rather than the copy in any one of them.
- Welcome. Catches every new subscriber the signup form produces, at an 8 to 12 percent conversion target. Load the strongest offer into the first email, because it carries 60 to 90 percent of the whole series' revenue.
- Post-purchase. The highest-attention message you will ever send, usually wasted on a receipt. It is also where the post-purchase survey belongs, which is the cheapest fix available for the attribution problem.
- Winback and replenishment. Fires at the category's natural reorder point, inside the thirty-day window where half of second orders happen. Ask for a reorder of the same product before offering something new.
- Browse abandonment. Lower conversion rate, far more traffic. Frequently skipped because the percentage looks unimpressive next to cart, and frequently worth more in absolute revenue.
- Sunset and re-engagement. Last because it looks like deleting revenue. It is buying deliverability, and without it the other five gradually stop reaching the inbox.
First three live inside two to three weeks. All six plus segmentation takes six to eight weeks. If authentication is not in place, that goes before any of it, because there is no point building flows that land in spam.
Common questions
Straight answers
How many emails should a cart flow have?
More than one, and that is the single highest-leverage sentence on this page. A single reminder recovers 2 to 3 percent of carts. A multi-step sequence combining email and SMS recovers 8 to 12 percent.
Three to five touches over 72 hours is the working shape. The recovery rate is a function of the number of steps far more than the quality of the writing in any one of them, which is why brands agonising over subject lines on a one-email flow are optimising the wrong variable.
Should we discount in the cart flow?
Not in the first message. The first touch should assume the cart was abandoned for a practical reason, because most of them are, and a discount trains people to abandon deliberately.
If an incentive appears at all it belongs late in the sequence and only where the margin carries it. Work out what an order actually contributes before deciding what you can afford to give away.
What goes in a post-purchase flow that is not a receipt?
Four things: how to get value from the thing they bought, a post-purchase survey asking how they found you, a review request once they have had it long enough to have an opinion, and a reorder primer timed to the category's natural cycle.
The survey is the underrated one. It is the only attribution source no ad platform controls, it costs nothing, and it consistently tells a different story from the dashboards.
Do flows stop working after a while?
The structure does not, but the content ages. A welcome flow written two years ago offering a discontinued product is not a working flow, it is a liability that fires every day.
Quarterly review is the right cadence. Check that offers are live, products are in stock, the timing still matches your delivery times, and the copy has not been overtaken by a repositioning nobody told the email platform about.
Should the same person get an email and an SMS from the same flow?
Yes, when the messages do different jobs and are spaced properly. What does not work is sending the same content twice on two channels within an hour, which reads as a system malfunction rather than a follow-up.
The pattern that works is email for the explanation and SMS for the nudge, separated by hours rather than minutes, with the text suppressed if the email already converted.
What share of revenue should flows be producing?
Mature programmes run 50 to 60 percent of email revenue from flows and the rest from campaigns. If campaigns are producing more than 70 percent of your email revenue, the flow architecture is the gap rather than the send calendar.
The ledger on the email page prices that gap against your own revenue, flow by flow.
Methodology
Sources
Benchmarks are published. Sequences are ours. The distinction matters and we would rather draw it than blur it.
- Second-order timing. A February 2026 analysis of 156,110 DTC customers, which found half of second orders landing within 30 days, three quarters within 90, a median time to second purchase clustering between 15 and 35 days, and 77 percent of second orders being reorders of the same product.
- Repeat purchase rates. Twelve-month DTC benchmarks of 25 to 30 percent from industry aggregators alongside the 18.8 percent aggregate from the same 156,110-customer study. The measurement window explains the gap.
- Flow against campaign performance. Klaviyo's 2026 ecommerce benchmark report across more than 183,000 brands: flows at 5.3 percent of email sends producing roughly 41 percent of email revenue, at close to 18 times the revenue per recipient, and SMS flows at 7.6 percent of sends producing 45.2 percent of SMS revenue.
- Welcome flow benchmarks. Target conversion of 8 to 12 percent and open rates of 40 to 60 percent for Klaviyo-powered ecommerce brands, with the first email carrying 60 to 90 percent of the series' revenue.
- Cart recovery rates. Top performers recovering 8 to 12 percent of carts through multi-step flows combining email and SMS, against 2 to 3 percent for a single-email flow, with average revenue per abandonment email around $5.81.
- Flow share of email revenue. Mature programmes running 50 to 60 percent of email revenue from automated flows, with campaigns above 70 percent treated as the marker of an underdeveloped flow architecture.
- SMS opt-out and consent. Opt-out benchmarks below 3.5 percent per send with well-run programmes at 0 to 1.5 percent, and the prior express written consent standard for marketing messages, reinstated by the FCC in August 2025.
Message-level timings are not published by anyone and are presented here as our build, tested and adjusted per account. Where a flow has no published conversion band the tool above says so rather than filling the gap with a plausible-looking number.
What to do with this
Start with the cart.
Finish inside a month.
Three flows live in two to three weeks covers most of the available revenue. Everything on this page runs on customers you have already bought, which makes it the cheapest thing on your roadmap.