Consent, cadence and carrier rules
SMS has two costs.Most brands count one.
The carrier charges a cent or two a message and everyone budgets for that. The second cost is the subscribers who leave, and it is usually several times larger. SMS earns its place in our ecommerce practice when it is run on that arithmetic instead of on open rates.
★★★★★ 4.9 across 312 client reviews
Why SMS is not just faster email
A send you cannottake back
Email and SMS get bundled together on most agency websites, usually as one line item. They are different disciplines and the reason is arithmetic rather than taste.
An email send has effectively no marginal cost. If a campaign underperforms you have lost the opportunity and nothing else. An SMS send costs money to deliver and costs subscribers to deliver, and neither of those comes back.
Run those two together and a send has a floor. There is a revenue per message below which the send destroys more value than it creates, and unlike email that floor is not zero. The calculator below works out where yours sits.
Frequency is the lever, and it moves both numbers at once.
Most ecommerce brands send about two messages a month per subscriber. The higher performers run four to six without harming retention, but only when the messaging is segmented and relevant. When opt-outs climb past 3.5 percent per send it is almost always cadence, and 61 percent of people who unsubscribe say the reason was too many messages. Opt-outs also spike after the second or third message to a new subscriber, which points at the sign-up promise rather than the messages.
The other thing worth knowing before you plan around it: the 93 to 98 percent open rate that sells this channel is measuring message expansion on a lock screen, not someone reading. It is a structural property of the format. Click rate, opt-out rate and revenue per send are the numbers that actually predict whether the programme is healthy.
None of this argues against SMS. It argues for treating it as its own channel with its own economics, sitting alongside email rather than inside it.
Free, no email required
What a sendactually costs you
This prices a single SMS send against the same send on email, per recipient, with both costs counted. It also works out the revenue per message below which sending is destroying value rather than creating it.
Your programme
From your SMS platform's campaign reporting.
Revenue per send is the number your platform reports per message, not per delivered or per click. Published ecommerce averages sit near $0.71 for SMS and above $0.12 for email. The email column assumes the same cadence for a like-for-like comparison and a 0.25% unsubscribe rate, the midpoint of the published 0.1 to 0.3% band. Contribution margin comes from the map on the hub.
Per send, both costs counted
Updates as you type. Nothing is sent anywhere.
What you keep Carrier cost Value of the subscribers lost
A lost subscriber is valued at the contribution they would have produced over the following twelve months at your current cadence and revenue per send. That is a conservative way to price it, since it ignores the orders they would have placed through other channels. Bars are scaled to the same axis so the two channels are directly comparable.
Where the list comes from
Consent capture,by form type
A phone number is a harder ask than an email address, so the form matters more here than anywhere else on the site. These are published 2026 conversion ranges, and the ordering is consistent across every dataset we could find.
| Form type | Typical conversion | What it is good for |
|---|---|---|
| Second-page trigger | Around 29% | The highest-converting trigger in the published data, and almost nobody uses it. The visitor has shown intent by clicking through, which is exactly when asking is reasonable. |
| Cart abandonment | Around 17% | Catches a specific person at a specific moment with something concrete to offer them. Feeds straight into the highest-value flow you will run. |
| Gamified, spin to win | 8–15% | Two to three times a standard discount form, and the usual way brands capture email and phone in one interaction. Dual capture lands SMS opt-in around 8 to 12%. |
| Multi-step | Email first, phone second | The structural fix for SMS capture. Asking for the number on a second step, after the email is already given, preserves the overall signup rate instead of suppressing it. |
| Standard discount popup | 3–5% | The default most stores run. It works, it is unremarkable, and it is usually the thing being compared against when someone says their popup converts fine. |
| Exit intent | 2–4% | Looks weak next to the others and is not. These visitors were already leaving, so the comparison is against zero rather than against a click-triggered form. |
Two rules that matter more for SMS than for email. Never ask for the phone number in the same step as the email, because it costs you subscribers on both. And set the expectation at sign-up, since opt-outs spike on the second and third message when the arriving cadence does not match what was promised. The form lives on the store, so this work and the conversion work are the same project.
The part most agencies cannot explain
Consent rulesand carrier gatekeeping
Two separate systems govern whether your message is allowed and whether it arrives. Getting them confused is common, and only one of them carries litigation risk.
Consent, under the TCPA
The regulatory headlines of the past two years have mostly been about a rule that never took effect. What did not change is the part that generates lawsuits.
- In forcePrior express written consent before any marketing text sent by an automated system. This standard was reinstated by the FCC in August 2025 and is not in doubt.
- In forceSince April 2025, opt-out requests must be honoured when made through any reasonable method, not only your keyword.
- In forceDo Not Call Registry protections extend to texts, and carriers operate their own blocking requirements alongside.
- VacatedThe one-to-one consent rule. The Eleventh Circuit struck it down in January 2025 in Insurance Marketing Coalition v. FCC, holding the agency exceeded its authority. It never took effect.
- DelayedThe revocation-all provision, which would make a single opt-out apply across every channel and message type from one sender. Pushed again in January 2026, now to 31 January 2027.
One practical point that matters more than any of the above: a transactional message needs only prior express consent, but adding a promotional line to it converts it into a marketing message and moves it to the higher written standard. Keep the two cleanly separated.
Not legal advice. This is a plain-language summary of a fast-moving area, current as of mid-2026. Litigation exposure under the TCPA has not softened because the rules around the edges moved, and consent provability is what a plaintiff's lawyer actually attacks. Have your own counsel review your capture language and your records.
Delivery, under 10DLC
Separate from the law entirely. US carriers require every business sending on a standard ten-digit number to register the brand and each campaign, and unregistered traffic is filtered rather than delivered.
- BrandA one-time registration fee in the $4 to $50 range, plus monthly campaign fees of roughly $1.50 to $10, passed through from the carriers and the registry.
- ApprovalOne benchmark study found 55% of business profile rejections traced to five documentation issues, all preventable before submission rather than appealed after.
- ThroughputRegistration tier determines how fast you can send. A large list on a low tier means a "flash sale" arriving over several hours.
- FilteringCarriers filter on content as well as registration. Certain categories and link shorteners draw scrutiny regardless of how clean your consent is.
The failure mode here is quiet. Filtered messages do not bounce visibly the way email does, so a programme can lose a chunk of its audience without anything in the dashboard looking wrong. Registration status and delivery rate by carrier are the two things worth checking before blaming the copy.
Worth saying plainly. If an agency is pitching you SMS and cannot explain the difference between a consent problem and a carrier filtering problem, they will diagnose the wrong one. The fixes have nothing in common.
You've got questions
We've got
straight answers
How often should we send?
Most ecommerce brands land around two messages a month per subscriber. Higher performers run four to six without damaging retention, but the qualifier is doing the work in that sentence: only when the messages are segmented and genuinely relevant.
The opt-out rate tells you where your ceiling is, not a benchmark. Under 1.5 percent per send you have room. Past 3.5 percent you are burning the list faster than you are building it, and the calculator above will show you what that costs.
Is SMS better than email?
Per recipient, usually yes by a wide margin. In total, usually no, because the email list is several times larger and costs nothing to reach. They are answering different questions and the honest answer is that you want both.
The useful framing is that SMS is for things that are time-bound, where arriving in three minutes rather than three hours changes the outcome. Cart recovery, back in stock, a genuine deadline. Everything else is an email.
Can we text the phone numbers we already have from checkout?
Not for marketing, no. A number given to complete an order carries consent for messages about that order. Sending a promotion to it converts the message into marketing and requires the higher written consent standard, which was not obtained at checkout.
This is the single most common way a brand walks into exposure, usually with the best intentions during a launch. Build the consent properly at the form and start smaller. This is not legal advice and your own counsel should review the capture language.
Our messages are not arriving. Is that a consent problem?
Probably not. Non-delivery is almost always carrier filtering, which is a separate system from the law. Check 10DLC registration status first, then throughput tier, then whether the content or the link shortener is drawing scrutiny.
It is worth knowing that filtered messages fail quietly rather than bouncing visibly, so a programme can lose a real share of its audience while every number in the dashboard still looks normal.
What about the one-to-one consent rule?
It does not exist. The FCC adopted it in late 2023, the Eleventh Circuit vacated it in January 2025 on the grounds that the agency had exceeded its statutory authority, and the FCC reinstated the previous prior express written consent standard in August 2025.
If someone is still building your compliance approach around that rule, they stopped reading in 2024. What genuinely matters is unchanged: you need documented, provable consent per subscriber, with the source, timestamp and the exact language they agreed to.
How long does it take to get an SMS programme live?
Registration is the gate, and it usually takes days rather than weeks if the paperwork is right first time. Since a majority of rejections come down to a handful of preventable documentation issues, getting the submission right matters more than getting it in fast.
Capture forms and the first flows, welcome and cart recovery, go live alongside. Campaign sending starts small and deliberately, because the early opt-out rate on a new list is the most informative number you will get and it is worth reading before scaling volume.
The rest of the system
The fifth of five
SMS is the smallest channel here and the one with the sharpest edges. It is worth running properly and it is not worth running alone, which is the whole reason the ecommerce practice is built as one system against one contribution number.
Next step
Thirty minutes.
Bring your opt-out rate.
We will cost a send properly, check whether your capture form is suppressing signups, and tell you whether your delivery problem is a consent problem or a carrier one. They are not the same and the fixes share nothing.
No obligation · Usually booked 1-2 days out