Sending SMS that arrives: segments, sender names and A2P registration

An SMS usually fails to arrive for one of three unrelated reasons: it split into more billed segments than anyone priced for, its sender name is one the destination country blocks outright, or its campaign was never registered for A2P 10DLC at the throughput it actually needs. Each cause sits in a different part of the sending stack - pricing, regulation, carrier throughput - which is why the two of them rarely get fixed by the same person, and why most pages about any one of them never mention the other two.

This guide walks all three in the order they actually bite: what a message costs before you send it, whether its sender name survives the destination country, whether the campaign behind it is registered for the volume you plan to send, and - underneath all three - whether the number itself is even readable. Every section hands off to a calculator that turns the explanation into a number for your own case, rather than leaving you to re-derive it from a paragraph.

Why does a text message cost more than you expect?Copy Anchor Link

A text message is billed by the segment, not by the character, and how much a segment holds depends on its encoding. In the GSM-7 alphabet, a single message holds 160 characters; split across several segments, each one holds 153, because six characters of every extra segment are spent on a header that lets the handset reassemble the parts. One character outside GSM-7 - a curly apostrophe pasted from a document, an emoji, an accented letter the alphabet does not cover - switches the whole message to Unicode (UCS-2), where a single segment holds only 70 characters, and a multi-segment one holds 67. This is how carriers actually bill, not a guess at character length, and it is the one fact in this guide that needs no outside citation: it is how the GSM 03.38 and 3GPP standards define a segment, not a claim any vendor makes about its own pricing.

The practical effect is that a 161-character GSM-7 message, or a 71-character Unicode one, is billed as two segments, not "a bit more than one." Our SMS length calculator counts a pasted message the way a carrier does and names the exact character that forced Unicode, if one did. Once you know the segment count, the SMS segment and cost estimator turns it into what one message costs at a wholesale price you pay, what a monthly volume costs, and - if you resell the number or the campaign - the margin between what you pay per segment and what you charge for it.

Which countries block an unregistered or alphanumeric sender name?Copy Anchor Link

An alphanumeric sender ID is a brand name, not a phone number, shown as who a text message came from - and whether a destination country allows one at all, requires it to be pre-registered, or restricts a link inside the message body are three separate questions that each country's own regulator or the vendor actually carrying the traffic answers differently. No regulator or vendor publishes one global table, and the few per-country pages that exist disagree often enough, and change enforcement dates often enough, that a page claiming to be current needs a date on every row, not just at the top.

That is 10 of the 16 countries our sender ID rules by country tool covers; the full table includes Canada, Italy, the Netherlands, Singapore, the UAE and Saudi Arabia, each dated the same way. A country not on that list has not been checked by the tool at all - which is a different claim from the source confirming no restriction - so the tool says that explicitly rather than treating silence as permission.

Two mistakes are common enough to name. The first is treating "generic sender IDs are prohibited" - a name like INFO, SMS or NOTICE - as a quirk of one country, when it is a standing rule on nearly every vendor page behind this table. The second is reading a single national row as the whole truth for every carrier inside that country: several of the sources above carve out exceptions by network that one row cannot show, so confirm with the vendor actually sending your traffic before a campaign launches, not after it is blocked.

How do you test a one-time code message before it ships?Copy Anchor Link

A one-time code (OTP) message is judged on three things that have nothing to do with whether the code itself is correct: how many segments the text is billed as, whether a link in the body is something the destination country's carriers restrict, and whether the brand name has to be in the body at all - which depends entirely on whether that country's alphanumeric sender ID can carry it. In a country with no alpha sender ID, like the United States, the message body is the only place the brand can appear. In a country where the sender ID already shows it, like Germany, the same brand repeated in the body is harmless but unnecessary.

Our OTP message tester previews a template against all three checks at once, reading the same sender ID table used above so the two tools never disagree about what a given country's rule is. It also flags a code outside the usual 4-to-8-digit range as unusual - a template often has more than one number in it ("expires in 10 minutes" as well as the code), and the real code is normally the longest run of digits, not the first one a person reads. None of this proves a message is actually delivered, or delivered before the person waiting for it gives up; only a live send through your carrier can show that.

What does A2P 10DLC throughput actually mean for a campaign?Copy Anchor Link

A2P 10DLC is the registration regime US carriers require before a 10-digit long code can send application-to-person traffic, and the two numbers that actually decide what a campaign can send are a daily message cap and a per-second throughput, both set mostly by a Trust Score. A Trust Score is a 0-to-100 score a brand earns from the carriers' and The Campaign Registry's (TCR) own vetting, and it is the single input behind both figures below (JustCall, read 13 October 2026).

Sourced to Bandwidth's T-Mobile 10DLC page (updated 13 August 2026) and Twilio's T-Mobile daily message limits page (updated 18 June 2026) for the daily cap, and to LeadConnector, HighLevel, Salesmessage and Twilio's message throughput page (updated 9 September 2026) for the throughput, all read 13 October 2026. Three things in that table are easy to get wrong:

• AT&T and Verizon publish no daily cap of their own. Every source we read discusses a daily cap for T-Mobile only; the other two carriers are described as throttled by per-second throughput alone. Treat the blank as "not documented," not "unlimited."
• A Sole Proprietor brand ignores the Trust Score entirely. Its daily cap (1,000 messages on T-Mobile, per Bandwidth's own exception) and its throughput (2.25 messages/second total, 0.5 for MMS per carrier) are fixed regardless of what score the brand carries - raising it does nothing here, unlike every other entity type.
• Not every source agrees. TSG Global's page, dated 24 March 2023, gives T-Mobile a flat 1 message/second regardless of band, where the three 2026-dated sources above scale T-Mobile with the Trust Score the same as the other two carriers. Nobody has reconciled the two; use the more recently dated figure and confirm your account's own approved throughput with your provider before planning a send around either one.

Registration itself is priced per brand (one-time) and per campaign (monthly), and no single schedule covers every provider - each one publishes its own pass-through figure, and they differ by more than a factor of ten at the high end.

Sourced to TCR's own fee PDF, Telnyx, Bandwidth and HighLevel, all read 13 October 2026. HighLevel's one-time figure bundles three separate line items most other providers itemise, so it is not a fair line for a line comparison against the other three - reading it as "the" one-time cost makes every other provider look artificially cheap. Our A2P 10DLC throughput and registration cost planner takes a Trust Score, a brand type and a campaign type and returns the daily cap, the per-carrier throughput and the fee range that actually applies, with every disagreement between sources shown rather than averaged away.

If you sell numbers to your own customers rather than registering your own brand, the ISV side of this is a different process again; see Twilio alternatives for reselling phone numbers for how a reseller registers a customer's brand and campaign.

Why does the phone number itself have to be right before any of this?Copy Anchor Link

Every check above assumes the number you are sending to is actually readable: a malformed number, or a national number read against the wrong country, breaks a segment count, a sender ID lookup and an A2P plan equally, before any of them get a chance to matter. Numbering-plan metadata - the same data Google's own libphonenumber project ships, and the one libphonenumber-js metadata set every tool below is built on - can say a number's shape matches a real, assigned range in its country's plan and what kind of line that range was assigned to. It cannot say the number is in service, or who carries it today after a porting. Both projects state this limit themselves, and every tool below repeats it rather than letting "valid" imply more than it means.

• The phone number validator checks one number, names its country and type, and shows it in every common format.
• The phone number formatter reshapes a number into E.164, international, national and tel: form without checking whether it is a real, assigned range.
• The bulk phone number validator runs a whole pasted list or CSV through the same check at once and returns a downloadable table - the natural first step before porting a customer list in, or loading it into a sending campaign.
• The phone number type lookup answers only whether a range was assigned as mobile, landline, toll-free or premium rate - not, its own page says up front, whether the specific line is live or who its current carrier is; that would need a live carrier query, which none of these four make.

If you still need a number to send or receive through - a different question from checking the shape of one you already have - see our comparison of phone number APIs for receiving SMS for providers and prices; it does not repeat the A2P or sender-ID detail above, only the number itself.

What does porting a number in actually require?Copy Anchor Link

Porting a number means proving two things to the carrier you are leaving: that you are who you say you are, and that the number is really yours to take. What counts as proof - and how long it should take - differs sharply by country, and two of the seven we checked - Germany and South Africa - have no regulator-published interval at all, only a vendor's own figure.

This is the national floor, not any one carrier's own form - a carrier can legally ask for more than its country's rule requires, and even a wholesale carrier that skips the paperwork at signup still expects you to produce it if asked later. The single most common reason a US port is rejected is an address on the LOA that does not exactly match the losing carrier's Customer Service Record, not a missing signature. Our number portability requirements checklist carries the full table, country by country, with the one document to tick off and the timing its own source publishes.

None of this is a feature ResellerHQ's own phone number product offers today: it rents numbers that receive SMS from upstream providers for fixed periods, it does not send, and it does not support porting an existing number in - so the sender-ID and A2P sections above do not apply to it either. If you need a number that can both send and receive under your own registered campaign, the vendors compared in phone number APIs for receiving SMS and Twilio alternatives for reselling phone numbers are the ones to evaluate against everything above.