Add SMS when your email program already works and you sell something people need to hear about fast or buy more than once: drops, restocks, replenishment, short sales. Collect explicit consent, keep texts for short time-sensitive messages and let email carry everything else. Then judge SMS on the revenue it adds against a holdout, not on what the dashboard credits it with.
When SMS is worth adding
Email costs close to nothing per send. SMS costs money on every message, so it only earns its place when it drives orders email wouldn’t. That usually requires most of these:
- Email is already built. Welcome, abandoned checkout, browse abandonment, post-purchase and winback flows are live, and deliverability is healthy. A second channel on top of a weak first one doubles the mess. If those flows aren’t in place, start with the Klaviyo flows ranked by revenue impact.
- Timing matters to your customers. Limited drops, restocks, low-inventory products and short promotions. Texts get read quickly, which only pays when a delay costs a sale.
- Customers reorder. Consumables and collectible ranges give you repeated moments worth a text.
- Contribution per order covers the cost. Low-margin, low-AOV stores need far more incremental orders per send.
Run the break-even math before signing a contract. A hypothetical example: a campaign to 20,000 subscribers at $0.02 per message costs $400. If each order leaves $40 of contribution after product, shipping and fees, the send needs 10 incremental orders to break even. Not 10 attributed orders; 10 orders that wouldn’t have happened through email or ads anyway.
Wait if your list is a few hundred people, your product is a once-a-decade purchase, or your only reason to text is another discount.
Consent and compliance basics
Get this right before anything else. This isn’t legal advice, and your opt-in language should be reviewed by counsel. In the US, the working rules are:
- Treat marketing texts as needing prior express written consent. A checkbox, form submission or keyword text from the customer can capture it, but the disclosure has to appear at the moment they opt in.
- Consent can’t be a condition of purchase. A phone number collected for delivery or order updates is not permission to send promotions.
- Carriers require approved sending numbers. Toll-free verification, 10DLC registration or a short code application. Your SMS platform usually guides the process, but approval takes time, so start early.
- Honor opt-outs promptly. STOP and its common variants must work automatically, and any clear request like “stop texting me” should be treated the same way.
- Sending hours are restricted. Federal and state telemarketing rules limit early-morning and late-night contact based on the recipient’s local time, and some states set narrower windows. Set quiet hours in the platform rather than relying on people to remember.
Before your first send:
- Opt-in language names the brand, says the texts are recurring automated marketing messages and states consent isn’t a condition of purchase
- Message frequency, “Msg & data rates may apply,” HELP and STOP instructions, and links to SMS terms and privacy policy appear at opt-in
- Checkout SMS box is unchecked by default and separate from email consent
- Consent records stored: phone number, timestamp, source and the language shown
- Welcome text confirms the subscription and repeats how to opt out
- Sending number registered and approved
- Quiet hours set by recipient time zone
- Every marketing text identifies the brand
Other countries set their own rules. The UK, EU, Canada and Australia all have consent and sender-identification requirements, so check each market before texting it.
Growing an SMS list
The best subscribers come from people already opting in to hear from you. In rough order of quality:
- Two-step sign-up forms. Email on step one, phone on step two. You keep your email capture rate and add phones from visitors who just raised their hand.
- Back-in-stock and drop alerts. “Text me when it’s back” captures people with a specific product in mind.
- Checkout consent. A separate, unchecked SMS box at checkout catches buyers at peak intent.
- Tap-to-text on mobile. A button opens the phone’s messaging app with a keyword prefilled; the customer sends it and gets a confirmation reply.
- Email-to-SMS invitations. Ask engaged email subscribers to join for early access or restock alerts.
- Offline prompts. Keywords or QR codes on packaging inserts and in-store signage.
Watch the incentive. If email sign-up already earns 10% off, don’t stack a second discount for the phone number. Offer one code for both, or SMS-only value like early access, which costs no margin. And never buy lists or text numbers pulled from order data.
Which messages belong on SMS vs email
The rule I use: if the message is short, time-sensitive and has one action, it can go on SMS. If it needs images, explanation or browsing, it belongs in email.
| Message | Channel | Why |
|---|---|---|
| Back-in-stock alert | SMS | Inventory can sell out before an email gets opened |
| Limited drop or launch | SMS, with email | The first hours decide sell-through |
| Last hours of a sale | SMS | The deadline is the message |
| Replenishment reminder | SMS for consumables | One tap to reorder |
| Abandoned checkout | Email first, SMS as follow-up | Email shows the cart; SMS nudges |
| Shipping and delivery updates | SMS if the customer opted in | Useful, and builds trust in the channel |
| Welcome series and brand story | Needs space and visuals | |
| Product education and comparisons | Too long for a text | |
| New collection or newsletter | Built for browsing | |
| Review requests | Email, optional SMS reminder | The request needs context |
Keep texts tight: brand name first, one link, under 160 characters where possible. An emoji switches the encoding and cuts a segment to 70 characters, which can double or triple the cost of a message. Use MMS only when the image earns its higher price.
Frequency and unsubscribe management
A STOP closes that channel unless the subscriber chooses to opt back in, so treat every opt-out as a cost. My usual starting point is flows plus two to four campaigns a month, raised only when the data supports it and never beyond the frequency you disclosed at opt-in.
For every send, track:
- Opt-out rate against your running average
- Click rate and revenue per recipient
- Replies that are complaints
A send with opt-outs well above your average tells you the message type, timing or audience was wrong. Move that type back to email.
Segment campaigns. Send most of them to subscribers who clicked or ordered in the last 60 to 90 days, and save the full list for the biggest moments of the year. In Klaviyo, turn on Smart Sending so profiles don’t receive a second text within the window you set.
Coordinating SMS with email flows
Treat the two channels as one conversation. Inside Klaviyo:
- Branch on consent. Add a conditional split on SMS consent so non-subscribers stay on the email-only path.
- Stagger, don’t duplicate. The same message on both channels at the same minute trains people to ignore one.
- Exit on purchase. Use flow filters so anyone who orders leaves the flow before the next message.
- Suppress on engagement. If someone clicked the email, the follow-up text is usually unnecessary.
An example abandoned checkout sequence. The timings are starting points to test, not benchmarks:
| Step | Timing | Channel | Who receives it |
|---|---|---|---|
| 1 | 1 hour | Email with cart contents and images | Everyone |
| 2 | 3-4 hours | SMS with a direct checkout link | SMS subscribers, no order, didn’t click step 1 |
| 3 | 24 hours | Email with reviews, shipping and returns info | No order |
| 4 | 48-72 hours | Email or SMS with an offer, only if margin allows | No order, first-time buyers |
Campaign days work the same way. Send the email in the morning, then text only the subscribers who didn’t click, later in the day and inside quiet hours. Post-purchase, text shipping updates, email the review request and time replenishment texts to how fast the product runs out. That timing ties into the wider plan in how to increase repeat purchase rate.
Measuring incremental revenue
Here’s the trap. Klaviyo credits an order to the most recent message the customer interacted with inside its attribution window. When a subscriber gets an email and a text about the same sale and taps the text, the order moves from email to SMS in reporting. Total revenue barely changes, yet SMS looks like a new revenue stream. Ad platforms may claim the same order too.
The fix is a holdout test:
- Randomly assign 10-15% of SMS subscribers to a holdout, using a random profile property, not engagement.
- Keep everything else equal. The holdout gets the same email program and transactional texts, but no marketing SMS.
- Run it for six to eight weeks, covering at least one promotion.
- Compare total store revenue per subscriber from order data across all channels, not Klaviyo-attributed revenue.
- Subtract costs. Incremental revenue times contribution margin, minus message costs and the platform fee.
A hypothetical example: 18,000 subscribers get SMS and 2,000 are held out. Over eight weeks, revenue per subscriber is $14.00 in the SMS group and $13.20 in the holdout. That $0.80 gap across 18,000 people is $14,400 of incremental revenue. At a 45% contribution margin, that’s $6,480 of contribution, set against what the texts cost. In a scenario like this, the attributed SMS revenue on the dashboard can be far larger than the measured lift.
Small holdouts are noisy, so read the result as a range and re-run it when the program changes. Setting up this kind of measurement alongside the flows is a core part of my lifecycle marketing work.
Get it built
If you’re weighing SMS, or already paying for it without knowing what it adds, the Growth Audit reviews your email and SMS program, consent setup and measurement. It’s $1,500 fixed and credited if we continue. See pricing or get in touch.