10DLC registration without tears

If you send business text messages to United States numbers from an ordinary ten-digit phone number (a "10DLC"), you must register first. Since early February 2025, US carriers and the providers that connect to them have moved from throttling and surcharging unregistered 10DLC traffic to blocking it. There is no degraded path to launch on, and teams routinely discover this in the week they meant to go live.

The US has three main routes for business texting, each with its own registration.

Lead times vary by provider and by how complete your submission is; treat them as planning ranges, not promises. The rest of this piece is about 10DLC.

Brand. Who you are: legal entity name, tax identifier (EIN), address, website. Registered once and reused.

Campaign. What you send and why. Registered per use case, so a clinic sending appointment reminders and running a separate fundraising programme registers two campaigns under one brand.

Both go through The Campaign Registry (TCR), usually via your messaging provider rather than directly, and both carry one-off and recurring fees that providers pass through.

Nearly all of it is avoidable, and nearly all of it comes down to consistency.

• Sample messages that do not match the campaign type. If the use case says "account notifications" and a sample contains a discount code, that is a marketing campaign filed as a service one.
• A website that does not corroborate the brand. Reviewers look. If the site does not obviously belong to the registered entity, expect friction.
• No visible opt-in. You must show how consent is collected. A privacy policy that never mentions text messages is not evidence.
• Missing opt-out and help language. Every campaign needs a working STOP path and a HELP response.
• Name mismatches. The brand name must match the tax record exactly, not the trading name.

1. Use the exact legal name and EIN as they appear on tax records.
2. Pick the narrowest use case that covers the traffic. A "mixed" campaign invites scrutiny of every message in it. Two narrow campaigns clear faster than one broad one, and fail independently.
3. Write samples from real traffic. Four or five messages you will actually send, each with the brand name and opt-out wording. An idealised sample your production template does not match surfaces later, in an audit, when it costs more.
4. Publish the opt-in before you submit. The page or form where consent is collected must exist and say what people will receive, how often, and how to stop.
5. Describe the call to action. Explain in plain words how a person ends up on the list.

Vetting takes time, and a rejection usually means correcting and resubmitting, which restarts the clock. The failure mode is not the delay; it is discovering it in launch week. Register during development, not during rollout.

Brand vetting produces a score derived from the business's identity, web presence and history. Some carriers use it, together with the campaign type, to set how many messages per minute or per day you may send. Throughput is therefore not something you can buy at the moment you need it; it follows from a clean record. The cheap move is to behave well from the first message.

A dental group with twelve clinics wants to send appointment reminders and, separately, a quarterly newsletter with offers.

• One brand, under the group's legal name and EIN, with the website showing that name in the footer.
• Two campaigns: "Customer care — appointment reminders" and "Marketing — newsletter and offers". The reminder samples contain no offers; the marketing samples say so plainly.
• Two opt-ins: reminders consent captured at booking; marketing consent as a separate, unticked checkbox.
• Timeline: submitted six weeks before go-live, with one rejection (the website did not mention texting), fixed in two days.

Had they filed one "mixed" campaign, a single promotional sample would have put the reminders at risk too.

The instinct is to treat this as a form somebody fills in once. That costs a rewrite when the second country arrives. Model it instead: a brand record, campaign records each with a use case and consent basis, and a link from every outbound message to the campaign it was sent under. You should be able to answer "which campaign, and which consent, covered this message?" with a query, for any message ever sent.

10DLC is one example of a global pattern: identified senders, declared use cases, traffic that has to announce itself.

• India requires entity, sender header and message template registration on the DLT platform, with messages checked against registered templates.
• Singapore made sender ID registration mandatory from 31 January 2023; unregistered alphanumeric senders were labelled "Likely-SCAM".
• Australia launched a national SMS Sender ID Register, with registration of branded sender IDs required from 1 July 2026; unregistered ones are shown as "Unverified".
• The UAE and Saudi Arabia register sender IDs and mark promotional traffic — see getting messages into the Gulf.

If your data model already holds brand, use case and consent evidence, each new regime is a field mapping rather than a new concept.

• Do you submit brand and campaign registrations for me, and who sees the rejection reasons — me, or only you?
• Which fees are passed through, which are yours, and are any charged per month per campaign?
• What throughput will my campaign get on each major carrier, and what would raise it?
• If I move to another provider, can my registered brand and campaigns move with me?

Next in this track: Gulf sender IDs. For the consent records registration depends on, read consent architecture.