Email vs SMS marketing ecommerce: which channel to add next

A second channel doesn't fix the first one

Brands ask about SMS when email has stopped feeling exciting. Rarely the same thing as email being finished. If three of the flows are missing and nobody has segmented the list since launch, the fault is in the build. And a text message is a faster way to make the same mistakes.

They aren't really competing anyway. They cost different things, they interrupt differently, and each one is consented to separately. That last bit is the part most comparisons skip.

One of them charges you by the message

Email is billed on how many people you could send to. SMS and WhatsApp are billed on how much you actually send. That isn't a detail. It's the difference between a channel that tolerates a weekly newsletter and one that punishes it.

So the brands doing well on SMS send less of it than you'd expect. They send at the moments where speed is worth paying for, and let email carry the rest.

Four moments where a text beats an email

We don't run SMS as a parallel newsletter. It goes on the handful of moments where the message has an expiry date on it, and the phone is the only place it'll be read in time.

• The cart, inside the first hour, when an email is still sitting unopened
• Back in stock, where the first person to reply gets the thing and everybody else doesn't
• Order and delivery updates, which are the only marketing messages people actively want
• The final day of something genuinely time-boxed, sent to a segment, not to everybody

What we won't do is send a text because the calendar looks empty this week. Every message you pay for teaches people whether your texts are worth reading, and a filler send costs you twice: once at the network, once at the next opt-out.

A service channel that happens to sell

WhatsApp is the one channel brands underrate, and they underrate it because they judge it as a broadcast channel. It's a conversation channel. The value is that somebody can reply and get an answer.

The friction is real. Templates go through approval before you can send them, the categories you send under are priced differently, and somebody has to be there when a customer writes back. A channel nobody's staffing is worse than no channel.

Where we've seen it work is on the messages a customer would have written to you anyway. Where's my order. Will this fit. Can I change the address. Answer those properly for a year and the occasional promotional message gets read when it turns up.

You can't move a list from one channel to another

The most expensive assumption in this whole subject is that a phone number on an order is an SMS list. It isn't. Sorting that out afterwards means going back and asking everybody twice.

• Collect the number and the marketing permission in the same place, on their own tick box
• Keep the email opt-in separate, so somebody can leave one channel without leaving both
• Record where and when each permission was given, because that's what you'd be asked for
• Give people a way out of the channel they're annoyed by, without walking away from everything

Design the sign-up form for the channels you mean to run in two years, not the one you run today. Retrofitting consent is the sort of job nobody budgets for and everybody ends up doing.

What we would add, and when

1. Finish the email programme: the flows that exist, the segments that decide who gets what, the list you can actually reach
2. Add SMS on two moments only, the cart and back in stock, and leave it there for a quarter
3. Read the two together, not separately, because the second channel changes what the first one earns
4. Add WhatsApp when somebody is answering it, and treat the first year of it as service

That order is deliberate, and it's slower than most people want. It's also the version where each channel gets judged on its own work instead of being handed the credit for a good quarter. If step one is the one you haven't finished, how we build email systems and lifecycle strategy are the pages to read instead of this one.