Back to blog

Email Subscriptions

SMS or Email? Mapping the Subscriber Lifecycle to the Right Channel

May 9, 2026 · 11 min read

In most subscription brands' marketing stacks, email and SMS are run by the same person, off the same calendar, with the same copy lightly trimmed for character count. The two channels get treated as interchangeable pipes for the same water. For one-off DTC purchases you can almost get away with that. For subscribers, people in an ongoing billing relationship with you, it quietly damages both channels at once.

Email is cheap, deep and skimmable; subscribers triage it on their own schedule. SMS is expensive per send, immediate and interruptive; it lands in the same thread as messages from their family. Those are not stylistic differences. They dictate which lifecycle moments each channel should own, and using the wrong one either gets ignored or burns consent you spent real money earning.

This post maps the subscriber lifecycle to the right channel, moment by moment — with the cost maths, the consent posture and the cadence guardrails we apply in the audits we run.

---

Two channels with opposite physics

Start with the economics, because they explain everything else. Once the platform fee is paid, the marginal cost of an email is a fraction of a penny. That cheapness is a feature. You can afford depth, education, imagery, a 600-word onboarding message that a subscriber skims in eight seconds and files for later. Email is pull-shaped: the subscriber decides when, and whether, to give it attention, and an unopened email costs both sides almost nothing.

SMS inverts every one of those properties. You pay per message, at typical UK platform rates of 3p to 5p per segment (US rates run a little lower), and a segment is only 160 GSM characters, dropping to 70 the moment a curly apostrophe or special character sneaks into the copy. More importantly, a text does not wait politely in a promotions tab. It buzzes the device in the subscriber's pocket and lands in the same thread as messages from their family.

An ignored email is free. An ignored text is a small withdrawal from a trust account you cannot easily top back up.

"SMS is not a louder email. It is a different contract with the subscriber, and the penalty for breaking it is the channel itself."

That gives you a clean test to apply before any SMS send: would the subscriber thank you for the interruption? For a flash sale, almost never. For "your card failed and your dog's food won't ship", almost always. Every row in the mapping below falls out of that one question.

---

The lifecycle-to-channel map

Every subscription, whatever the product, moves through the same set of moments: a welcome, a billing rhythm, an action window before each order, the occasional failed payment, deliveries that sometimes go wrong, and eventually a cancelation you will try to reverse. Here is where each channel belongs.

The lifecycle-to-channel map. Teal marks the default channel for each moment; amber marks conditional use. If a send does not fit a row, it probably should not exist.

A few rows deserve expansion. Welcome and onboarding belong to email because the job is education. A new subscriber needs to know how billing works, where the skip button lives, what happens when they go on holiday and when the first box ships. We have written a full guide to Klaviyo flows for subscription brands that covers the sequence in detail; the short version is that it is reading material, not an interruption.

Nobody wants a four-part onboarding journey arriving as text messages.

Pre-billing reminders default to email, with one exception worth making. If your order value sits north of roughly £70, an optional SMS the day before the charge is defensible, because a surprise charge at that size generates support tickets, refund requests and the occasional chargeback. Below that, the reminder email plus a clear portal does the job.

Win-back runs the opposite way to most people's instinct. Email goes first because you need room to address why the subscriber left, and SMS is held back for a single, expiring final offer at the end of the sequence. We cover the full structure in our win-back campaign guide. Promotions, meanwhile, are email's natural territory: a list will tolerate a weekly promotional email indefinitely, and a weekly promotional text for about a fortnight before the STOP replies begin.

---

Where SMS earns its keep: dunning and the customization window

A failed payment is the clearest case for SMS in the entire lifecycle. Most failures are involuntary: an expired card, a stale billing address, a soft decline from an over-cautious issuer. The subscriber wants the product; the bank is the obstacle. This is the rare moment where interrupting someone is a favour, because every hour the failure sits unresolved is an hour closer to a canceled subscription neither side wanted.

The sequence design that works is a channel switch, not a volume escalation. Open with email: it is cheap, it carries a full explanation and an update-card link, and it creates a paper trail. But after two ignored emails, a third email is just a quieter version of the first two. Swap that touch for a short text with a deep link straight to the card update form, and the interruption does the work the inbox could not. It is one of the highest-impact changes in the dunning optimization playbook.

The order customization window is the other moment SMS genuinely shines, because it is a time-boxed action with a real deadline. "Your box locks in 48 hours; tap to swap or skip" is exactly the shape of message a text was built for. Email works as the supporting act, but a message that needs action within two days cannot rely on a channel where attention routinely arrives three days late.

A box brand we worked with moved its customization nudge from a day-before email to a 48-hours-out text with a deep link straight into the swap screen. Skips and swaps went up, and the support tickets that read "cancel my order, I forgot to change it" largely disappeared. More skips sounds like lost revenue; in practice it is the opposite, because the alternative to an easy skip is rarely a kept order — it is a refund request or a cancelation.

Delivery issues follow the same logic. A parcel stuck at a carrier or an address problem is time-critical, personal and unambiguous, and a text about it reads as service rather than marketing.

---

Run the cost maths before you build the cadence

Cadence debates get much shorter when somebody opens a spreadsheet. Take an illustrative brand with 15,000 SMS-consented subscribers, paying a typical UK rate of roughly 4p per message segment, and compare a triggered program against a weekly blast.

Illustrative figures at typical UK platform rates. The triggered program touches the exact moments money is at stake; the blast spends sixteen times more on the sends least likely to be welcome.

The asymmetry is the point. A fully built triggered program costs about as much per year as a single decent paid-social test, and every message in it answers the "would they thank you?" question. The weekly blast costs an order of magnitude more, and it is also what drives opt-outs — and every opt-out removes someone from the dunning rescue texts that were quietly paying for the whole channel.

Watch segment counts, too. A message with a link, a brand name and a curly quote can tip past 160 characters and silently bill as two segments, doubling the cost of every send. Most platforms show the segment count before you confirm; read it every time.

---

Consent, quiet hours and keeping the channel alive

Consent is where channel discipline stops being a taste preference and becomes exposure. In the US, the TCPA requires express written consent for marketing texts and carries statutory damages per message, which is why claims in this space are a cottage industry. In the UK and EU, PECR and GDPR set the same broad posture: explicit opt-in, easy opt-out, records you can produce. We are not your lawyers and this is not legal advice — but the defaults below are also simply good manners.

Quiet hours deserve a special mention because they are the easiest rule to break by accident. The common US standard is no marketing texts before 8am or after 9pm in the recipient's local time, and some states are stricter still; a 6pm send from a UK office lands at 10am in California and 3am in Sydney. If your platform supports quiet-hour enforcement per recipient timezone, turn it on and forget about it.

Six guardrails that keep an SMS program compliant and, just as importantly, worth subscribing to.

One more rule we hold firmly: never send the same message on both channels on the same day. If an email needs an SMS echo to perform, the email is the problem, and duplicating it teaches subscribers that your texts contain nothing the inbox did not already have.

---

Email carries the relationship. SMS marks the moments.

That pair of sentences is the whole strategy, and it is the one we repeat most often when reviewing subscriber messaging. The healthiest SMS programs we see send noticeably fewer texts than their platforms would happily bill for, and the texts they do send get acted on at rates the inbox has not produced in years. Restraint is not a compliance tax. It is what keeps the channel valuable.

The pressure on this discipline only points one way. Inbox providers are summarising and triaging email with AI, which compresses the attention each message receives, while RCS and WhatsApp are arriving as richer, costlier interruption channels with the same physics as SMS. The mapping survives all of it, because it is built on the economics of attention rather than on any single channel's feature set.

A useful exercise this week: pull your last ninety days of sends and place each one on the matrix above. Anywhere SMS is doing email's job, you are paying for annoyance. Anywhere email sits alone on a time-boxed moment, you are losing actions a text would have caught in time. Both are recoverable, and in our experience the fix is usually flow configuration rather than a replatform.