The honest answer is that the open-rate numbers you will find for this question are not worth the time it takes to read them, and the decision turns on three things you can work out for your own store in an afternoon: what a send costs, what consent costs you at signup, and how fast the product sells out after it comes back.
Why the published benchmarks mislead
Search this question and you will find confident figures — email opens in the sixties, SMS opens in the high nineties. Both numbers are real measurements of something. Neither measures what you need.
An email open is a tracking pixel loading. Apple Mail Privacy Protection pre-loads that pixel whether or not a human looked at the message, which inflates the number for any list with a lot of Apple users, and deflates nothing for anyone else. It is not comparable across two stores, let alone across two channels.
An SMS "open" is not measured at all. There is no pixel in a text message. The 98% figure that circulates is a read-rate estimate from carrier research, not something your provider observed about your messages.
And both numbers come from lists that selected themselves. People who hand over a phone number for a restock alert are, by definition, keener than people who hand over an email address — the friction filtered them. A channel that attracts only the most motivated subscribers will always look better per message and may still produce fewer sales, because it has fewer subscribers.
The metric that survives all of this is units sold per restock, divided by the number of people who were waiting. Everything upstream of that is noise.
What actually differs between the two channels
Cost per send
Email is effectively free at any volume a Shopify store will reach. SMS is not — it is priced per message and per destination country, and an international waitlist costs materially more than a domestic one. This is the single biggest structural difference, and it is the one people skip.
Work out the break-even before choosing. If a restock notifies 400 people and your margin on the product is £8, SMS at 4p a message costs £16 to send and needs two extra sales to pay for itself. That is usually an easy bar. If the product is a £6 accessory with £2 of margin, the same send needs eight extra sales and the arithmetic stops working.
Consent, and what it costs at signup
Email consent for a restock alert is uncontroversial — the customer asked to be told about one product, and telling them is what they asked for. A phone number is a bigger ask, and in most jurisdictions marketing SMS needs express consent that is separate from the email opt-in, with the sender identified and a documented way to stop.
The practical consequence is signup rate. A widget that asks for a phone number collects fewer subscribers than one that asks for an email address. Whether the better per-message performance covers that gap depends entirely on how much narrower the funnel gets, which you can only find out by running both.
What the message can carry
An email holds the product image, the variant, the price, and a button. A text message holds about one line of useful text and a link, and the link is what does the work. For a product where the customer is choosing between variants, the email is meaningfully better at doing the choosing for them. For a single-variant item they already decided on, the extra content is decoration.
Timing
This is where SMS earns its cost. A text is read in minutes; an email may be read that evening. If the restock is thirty units against a waitlist of four hundred, the difference between minutes and hours decides who gets one. If you restock 500 units of a staple, nobody is racing and the speed advantage buys you nothing.
A rule that holds up
Use SMS when the restock is scarce relative to the waitlist, when the item is high-margin enough to absorb the per-message cost, and when your audience is domestic or concentrated in a few countries.
Use email when you restock in quantity, when margins are thin, when the catalogue is broad enough that subscribers are waiting on many different items, or when your waitlist spans a lot of countries.
Offer both and let the customer pick when you have products in both categories, which most stores do. The cost of offering both is one extra field on the widget; the benefit is that the people who want to be interrupted can ask to be, and the people who do not are still on the list.
WhatsApp, and why it is not a third column in the same table
In Brazil, India, much of Europe and most of Southeast Asia, WhatsApp is where people actually read messages, and offering it can lift signup rates in those markets more than any copy change on the widget. It is also a different animal operationally: templates need approval before you can send them, and the pricing model is conversation-based rather than per-message. Treat it as a market-specific channel you add deliberately, not as SMS with a different logo.
How to test this properly on your own store
Most channel tests answer the wrong question because they compare two different sets of people. Do this instead:
- Offer both channels on the same widget for a month, so subscribers self-select and you learn the real split. This alone tells you something useful: if 90% choose email when both are offered, the SMS decision is already made.
- On one restock, notify the whole list at once through whichever channel each person chose, and record sales by channel in the first hour, the first day, and the first week.
- Measure sell-through per subscriber, not opens and not clicks. Units sold divided by people notified, by channel.
- Subtract the send cost from the SMS side before comparing. This is the step everybody skips and it is the one that changes the answer.
- Repeat on a slow-moving product. The answer for a limited drop is not the answer for a restocked staple, and if you only test on one you will apply the wrong rule to the other.
Two restocks is usually enough to see which way it falls for your catalogue. It will not match anyone's published benchmark, and it does not need to.
The failure that costs more than the channel choice
Whichever channel you pick, the thing that reliably destroys a waitlist is notifying four hundred people about thirty units. Everyone who clicks through to a sold-out page has now been rejected twice on the same product, the second time by invitation. Do that once and your next restock email gets opened by far fewer people, in any channel.
Matching the size of the send to the size of the restock matters more than whether it arrives by text or by inbox.
Where to go next
Stok offers email, SMS and WhatsApp on the same widget, so the channel is the customer's choice rather than a decision you make for everyone. For sizing the send against the stock, see what to do when more people are waiting than you have stock, and for setup and timing generally, the complete back-in-stock guide.
