Use SMS instead of email when your message is truly time-critical, short enough to act on without opening anything else, and valuable enough that the cost of the send is comfortably repaid by the action it triggers. Everything else belongs in email, which is cheaper, roomier and far more tolerant of being ignored for a day.
The channel decision is really a decision about the job you are asking one message to do, and most teams answer it by asking which channel is available instead.
That is why so many SMS programmes end up feeling like email programmes that cost you more.
Why the question usually gets answered wrongly
Channel choice gets made by org chart, not by intent
In most businesses the person who owns the message owns a channel, so your message ends up wherever its owner happens to sit.
Email teams write emails (because that is what they are measured on).
Your SMS budget usually belongs to someone else, often in a different function entirely, and the two rarely sit in the same planning conversation.
So the channel gets chosen before anyone has asked what your customer is supposed to do next, which is the only question that should decide it.
Performance comparisons between the two are close to meaningless
You will see SMS compared to email on engagement rates, and SMS wins that comparison every single time.
It wins because it interrupts, because the message is short, and because your SMS list is usually a fraction of the size of your email list (made up of people who opted in more recently, and more deliberately).
Comparing them on rates tells you almost nothing about which one should carry a given message. What would it tell you to know that a channel you use twelve times a year outperforms one you use two hundred times?
The cost asymmetry is the part people skip
Email costs you almost nothing at the margin, so a mediocre email is cheap and a brilliant one is cheap.
SMS costs you per message, every time, which means every send has to clear a commercial bar before it clears a creative one.
That single difference should change how you plan.
A channel with a per-message cost cannot be your default, and it cannot absorb the volume of a programme designed around a near-free one. So what is your cost per SMS, and what does a send have to earn to justify it?
What is really going on for the customer
SMS borrows attention that was never offered to you
An email waits.
A text interrupts, and it interrupts in the same place your customer receives messages from their family, their bank and their GP.
That proximity is the entire value of the channel, and it is also why your permission there is thinner than you think.
People who happily receive fifty emails a year from you will not accept fifty texts, and they will not warn you before they go.
Have a look at your own phone. How many brands are in there, and how many of them do you still tolerate?
Fatigue arrives faster and the exit is more final
Email disengagement is gradual, which gives you time to notice it and respond.
SMS fatigue tends to arrive as an opt-out instead, and that opt-out is close to permanent, because you have very little left to re-permission them with. How would you win one of those customers back?
That makes SMS an unusually unforgiving channel to experiment in, and it is why I would rather under-use it for six months than teach your list to ignore it.
The message being short is a feature, not a limitation
Anything that needs explaining, comparing or browsing is being asked to do a job the format cannot do, and sending it anyway produces a text that is really just a link with an apology in front of it (which your customer has seen many times before).
If your customer has to leave the message and go away and think, you have written an email and posted it through the wrong slot.
Where each one actually belongs
SMS earns its place on time and consequence
The cases where SMS clearly beats email share two features: the window for acting is short, and there is a real consequence to missing it.
Delivery and appointment updates, a payment that has failed on a live subscription, a security or verification code, a service disruption, a tightly time-boxed offer to a high-intent group.
In each one your customer is measurably worse off if the message waits until they next open their inbox.
That is the whole test, and it is worth applying message by message rather than programme by programme.
Is your customer materially worse off if this sits unread until tomorrow? Would they thank you for the interruption, or resent it? And if the honest answer is that it could have waited, why is it going by text?
Email earns its place on everything that needs room
Anything requiring context, choice, comparison, several links or a considered decision belongs in email, which covers most of what your CRM programme actually sends.
Email also carries your relationship between the urgent moments, and that is the part quietly building the permission your SMS programme then spends.
Channels like WhatsApp has changed the shape of the question
Your choice is no longer only SMS or email, because richer messaging channels now sit between the two, offering images, buttons and real two-way conversation, such as WhatsApp.
They come with a different consent model and a different cost structure (conversation-based rather than per-message), and the regulatory furniture is catching up, with unsubscribe controls now appearing natively in RCS in the UK.
None of that changes your underlying decision.
It means the answer to “which channel” now has more than two options, while the question behind it is still what the message is for.
What getting this wrong actually costs
You spend real money proving something you could have assumed
An SMS programme carrying messages your email could have carried is a direct, recurring cost with no incremental revenue attached to it.
Nobody reports it that way, because SMS gets measured on its own engagement rates and those rates look excellent.
The number nobody produces is what those same customers would have done without the text.
You burn the channel you were saving for the important messages
Every non-urgent text spends a little of the permission you will need on the day something urgent actually happens.
A pattern I see repeatedly in CRM and email marketing reviews is a business that added SMS for promotions, watched opt-outs climb over two quarters, then found the channel too depleted to be useful for service messages (which is where it would have paid for itself).
The customer reads your internal structure again
When the same customer gets an email and a text about the same thing within an hour, they are experiencing two teams who did not talk to each other.
What I would do
Decide by message intent, and write it down
The first thing I would do is take every recurring message in your programme and assign it a channel on one criterion: does this need acting on before your customer next opens their inbox?
Anything that does is a candidate for SMS.
Everything else stays in email until someone can argue otherwise with a number attached.
Give SMS a job description rather than a share of volume
I would define, in one sentence, what SMS is for in your business, then hold every proposed send against it.
Businesses that treat SMS as another channel to hit the target with always end up over-sending it, because no rule exists that would ever say no.
Suppress across channels before you add volume anywhere
Before adding a single new SMS to your plan, I would make sure a customer who has just received a text does not get the same message by email an hour later.
Cross-channel suppression is unglamorous, it usually already exists in tools you own (so there is nothing to buy), and it removes more customer irritation per hour of effort than almost anything else on your list this quarter.
It does depend on your data and how your channels are set up. Every time I have looked at a programme running both channels without a priority rule between them, though, there has been at least one message pair that contradicted itself.
Three actions
- Review: List every automated and recurring message you send, and mark each one as either time-critical (the customer is worse off if it waits) or not. Note how many of your current SMS sends fail that test, and how many emails would pass it.
- Decide: Write one sentence defining what SMS is for in your business, agree it with whoever owns the budget, and set a rule that any send outside that definition needs an explicit case (not simply a slot in the calendar).
- Test: Take one journey where you currently send both an email and a text, enforce a priority rule for six to eight weeks so only one goes, and measure the effect on completion of that journey and on SMS opt-out rate, rather than on the performance of either message.
If you are not sure your programme could answer that question message by message right now, that is the gap worth closing before you add another channel to it.
Find out with the free CRM Priority Check, it takes about five minutes >