- One text per winback flow. I only use one-step flows for SMS, so the text has to be pointed.
- The text is a reminder, not the whole winback. Texts are reminders, emails are branding and awareness.
- Don't stop the winback early because you added a text. In my email data from 4 brands, winback email #3 earns about 2x email #1.
- My habit is to include an offer in any SMS automation after the first text. That's my view, not a rule, and you can test it.
- Only text people with SMS marketing consent, exit anyone who orders again, and leave quiet hours on.
Does winback SMS actually work?
It can, if you give it the right job.
SMS is a more acute form of marketing than email. It's timely and urgent and carries very little brand awareness. A lapsed customer doesn't need a brand story by text. They already know you. They need a nudge at the right moment.
That's why I think about the two channels like this: texts are reminders, emails are branding and awareness. In a winback, the emails carry the product benefits, the use cases and the reasons to come back. The text reminds people that those emails are sitting in their inbox, or that it's time to reorder.
The SMS platforms back the idea. Attentive's guide to cutting opt-outs recommends a winback journey for shoppers who haven't bought in a while and says these journeys are "a great way to drive sales and prevent subscriber opt-outs." Klaviyo's SMS best practices page notes that "SMS tends to perform better when sent before email."
What a winback text can't do is replace the flow. More on that below.
How many texts should a winback flow send?
One per person.
I only use one-step flows for SMS. In email I'm happy to run long flows, because emails are easy to ignore. A text isn't. Every extra text in a winback is another chance for someone who already drifted away to reply STOP.
Some brands go further. Postscript's post on SMS automations shows a two-message winback and suggests testing a two-message version where the second text is canceled if the first one converts. That's a fair setup to test. My starting point is still one text, placed where a reminder matters most.
It also helps to count winback texts toward your overall frequency. Klaviyo's best practice is 1 to 2 texts a week, and its suggested monthly mix counts flow messages in that total. If you're already near that, a winback text has to earn its spot. More in how often to send SMS.
When should the winback text go out?
After the winback has started on email, and at the step where a reminder makes the most sense.
First, the flow itself. Klaviyo's guide to building a winback flow says the first delay should be "a little longer than your business's average buying cycle." Postscript says SMS winbacks are "commonly" triggered 30, 60 or 90 days after the last purchase. Your buying cycle matters more than any one of those numbers.
For replenishable products, I email customers with no order in 60 days to bring them into an educational cycle: how to use it, why it works, what they're missing. That's a good place for a text. Let the education email go out, then send one text that tells them it's there, or that it's time to restock.
Where I'd put the text in a longer email winback:
- Not first. Let email open the conversation. The text lands better once there's something to point back to.
- Near the offer or the deadline. A reminder is most useful when there's a reason to act now.
- Never next to an email that says the same thing. If the text repeats the email an hour later, it's wasted attention.
For the full email timing, see how long should a winback flow be.
Why you shouldn't cut the winback short after the text
It's tempting to think the text does the heavy lifting, so the flow can end after it. The numbers I have say the opposite, and they're email numbers, so read them that way.
Across 4 brands, winback email #3 earned about 2x what winback email #1 did: about $0.23 per person vs $0.11. The later emails in a winback did more of the work, not less.
This is email data, not SMS data. I don't have a winback text result to share. But the point carries over. If adding a text tempts you to trim the sequence to one email and one text, you cut the step that earned the most. Keep the emails. Add the text on top, on its own path for SMS subscribers.
Should a winback text include a discount?
My habit: in any SMS automation, anything after the first text carries an offer. That's my view from running these, not a rule or a requirement.
How that plays out in a winback:
- If the text is the first touch of the offer, you can send it without one. A reminder and a link back to what they bought is enough for people who just forgot.
- If you add a second text, that one gets the offer. By then, a plain reminder has already had its shot.
- If your text sits next to the winback offer email, the text points to that offer.
Platform guidance leans toward an incentive too. Klaviyo's winback guide recommends including "an incentive or discount" and putting "a time cap on the incentive." The winbacks in Postscript's post use offers like 10% and 15% off, and Postscript suggests testing a discount against no discount. That test is worth running on your own list before you give margin away to everyone.
If you want the bigger picture on offers, I cover it in my email discount strategy post.
When does a winback text annoy people?
Most of the annoyance comes from five mistakes.
1. Texting people without SMS marketing consent. Klaviyo's guide to adding SMS to a flow says "email consent does not cover SMS." Its compliance page adds that consent for transactional texts, like order updates, doesn't subscribe someone to marketing. A past customer who got shipping updates by text hasn't agreed to a winback text. The CTIA's messaging principles say an opt-in "should apply only to the campaign(s) and specific Message Sender for which it was intended or obtained."
2. Texting people who already came back. Klaviyo's winback setup uses a flow filter of "Placed Order zero times since starting this flow." Nobody should get "we miss you" the day after they reordered.
3. Texting people who are long gone. Postscript's help center describes a separate list-cleaning winback for subscribers who haven't ordered in 365 days, have never clicked, and subscribed more than 180 days ago. That message isn't a sales pitch. It gives them a chance to opt out or say what they want. Don't mix those people into your normal winback.
4. Sending at bad hours. Klaviyo's quiet hours page says to avoid texting before 8 a.m. or after 9 p.m. in the recipient's local time. Its compliance page notes that some states, like Florida, limit sending to 8 a.m. to 8 p.m. In a flow, a text that hits quiet hours waits in a queue until it's allowed to send, and the person doesn't move on to later steps until it does.
5. Leaving out who you are. Klaviyo says to mention your brand in every text, for example with a "[Brand]:" prefix, and to include an opt-out keyword like STOP for numbers that can take replies. A lapsed customer may not recognize your number anymore.
This is general information, not legal advice. If you're not sure what applies to your list, ask whoever handles compliance for you.
How do you set up winback SMS in Klaviyo?
- Start with the winback email flow. Klaviyo's guide triggers it on Placed Order with the filter "Placed Order zero times since starting this flow."
- Pick one step for the text. Usually near the offer or after the education email.
- Split on SMS consent at that step. SMS subscribers get the text, everyone else gets the email. Then both paths rejoin. I walk through the split in email and SMS flow strategy.
- Leave quiet hours on. Klaviyo shows the setting on each text message in a flow.
- Turn Smart Sending off on the text. I keep Smart Sending off on automations so people don't miss critical reminders. Klaviyo checks quiet hours first, then Smart Sending, so a delayed text can still get skipped if Smart Sending is on.
- Check the cost. Every text is billed per segment, per person. A winback text to a big list of lapsed customers adds up. The breakdown is in Klaviyo SMS credits.
Here's the shape of a pointed winback text. It's a made-up example, not a client message:
[Brand]: It's been a while! Your favorite is still here. Restock in one tap: [link]. Reply STOP to opt out.
Short, says who it's from, one link, one reason to tap. If it's the second text in a test, that's where the offer goes.
FAQ
Does a winback flow need SMS?
No. Email does most of the work in a winback. A text is a reminder you add for customers who gave SMS consent, at the step where a nudge matters most.
How many winback texts should I send?
I use one per flow. Postscript shows two-message winbacks, so if you test a second text, cancel it for anyone who already ordered.
When should a winback text go out?
After the winback has started on email. Klaviyo says the winback's first delay should be a little longer than your average buying cycle, and Postscript says 30, 60 or 90 days after the last purchase are common triggers.
Should the winback text have a discount?
My habit is to include an offer in any SMS automation after the first text. Klaviyo's winback guide recommends a time-capped incentive. Test with and without one before you decide.
Can I text past customers who only got shipping updates by SMS?
Not for winback. Klaviyo says consent for transactional texts doesn't subscribe someone to marketing messages.
- How to create a winback flow, Klaviyo Help Center
- How to add an SMS action to a flow, Klaviyo Help Center
- Basics: SMS best practices, Klaviyo Help Center
- Basics: SMS compliance, Klaviyo Help Center
- Understanding text message quiet hours in flows, Klaviyo Help Center
- 7 SMS automations to boost revenue, Postscript
- Clean your SMS list with a win-back campaign, Postscript Help Center
- High SMS opt-out rate? Here are 6 tips to improve it, Attentive
- Messaging Principles and Best Practices, CTIA (PDF)