A2P 10DLC shows up on every SMS provider's onboarding checklist, and almost nobody explains what it stands for before asking you to register for it. Application-to-Person 10-Digit Long Code. Underneath the acronym is a real system, built by the wireless carriers, and getting it wrong is the single most common reason a store's texts stop sending.
What it is
Before 2021, businesses sent SMS through long codes, the same ten-digit numbers a person's cell phone uses, with no registration required. Spam volume through that channel grew fast enough that carriers built a registration system on top of it: every business sending application-generated texts now registers a Brand, then registers one or more Campaigns under that Brand, before a single message goes out at scale.
The Brand is the business entity. The Campaign describes the use case: marketing, cart recovery, order updates, whatever the message content is. Carriers use both pieces to decide how much trust to extend a sender, and trust translates into throughput, the number of messages per second a number can send before carriers start filtering it.
Where store owners get tripped up
The most common mistake isn't technical. It's structural. A Brand has to be registered under the legal entity sending the messages, tax ID and all. Stores that started with a DBA, switched holding companies, or brought on a new SMS platform without re-registering end up with a Brand tied to an entity that no longer matches what's running the sends.
That mismatch doesn't always cause an immediate failure. Messages can keep sending for months on a legacy registration before a carrier audit catches the gap, and by then a store has built months of campaign history on top of a foundation that won't hold up to scrutiny. Fixing it later means re-registering under the correct entity, which starts Brand vetting over from scratch, with no guarantee the new score matches what the old one earned, and no fast path back to the old throughput while the new registration works its way through carrier review.
This shows up over and over in agency-managed accounts. A brand rebrands, spins up a new LLC for tax or liability reasons, or gets acquired, and the SMS registration never gets touched because nobody flagged it as connected to the entity change. The number keeps working right up until it doesn't, and by the time it stops, the fix competes with whatever caused the entity change in the first place.
Campaign category matters just as much. A Campaign registered as "Customer Care" and used to send marketing promotions gets flagged, filtered, or shut down over time, because the content doesn't match what carriers were told to expect. The category isn't a formality. Carriers monitor content against declared use case on an ongoing basis.
Why the timeline catches people off guard
A clean submission, matching EIN, a straightforward use case, can clear secondary vetting in about a week. Almost nothing about a small e-commerce brand's paperwork is clean on the first try. A business name on the SMS platform that doesn't match the EIN registration, a P.O. box where carriers expect a street address, a use case description that reads as marketing when the campaign type says something else: any of these bumps a Brand into manual review.
Manual review runs on the Campaign Registry's queue, not the sender's. A rejected Brand can be corrected and resubmitted, but each round adds days on top of the last one. A store working through two or three rounds of correction can land four to six weeks out from a clean Brand and Campaign approval. Add a business restructuring in progress at the same time, a new LLC, a name change, an acquisition, and the timeline stretches further, since A2P paperwork and legal paperwork almost never land on the same calendar.
None of this shows up in a typical SMS platform's onboarding flow. A store signs up expecting to send texts that week and finds out the registration alone can run a month or longer.
TCPA, in short
A2P 10DLC governs whether carriers deliver a message. TCPA, the Telephone Consumer Protection Act, governs whether sending it was legal in the first place. The two run on separate tracks, and passing one says nothing about the other.
TCPA requires clear consent before marketing texts go out, consent that has to be specific to SMS rather than inherited from an email opt-in or a general terms-of-service checkbox. It also requires an easy opt-out, honored right away, and record-keeping that shows consent existed at the time a message was sent, not just a record showing consent was collected.
A store can be well registered under A2P 10DLC and still be violating TCPA, because one system checks whether a carrier will deliver a message and the other checks whether sending it was allowed at all. Registration and consent solve different problems, and passing one is not evidence the other exists. A store that gets an audit right on one front can still fail the other, and each failure carries its own consequence: a throttled number on one side, a legal complaint on the other.
Response rate vs. open rate, in this context
Every SMS platform reports open rate high because SMS gets read within minutes of arriving. That number says nothing about registration health or consent quality. A well-registered number sending to a list with real, documented consent still shows the same 90%+ open rate as a number registered under the wrong entity, right up until the carrier starts filtering messages or a TCPA complaint arrives.
Response rate tells a different story, because a shopper replies to a message she trusted as legitimate, not one that reads as spam. A number that's been flagged, throttled, or tied to spam complaints degrades on the carrier side before it shows up as a metric a store can see on its own dashboard. By the time open rate drops in a way anyone notices, the underlying registration or consent problem has existed for weeks.
What good registration looks like
Brand registered under the entity operating the sends, with a tax ID that matches. Campaign category that matches the message content, not a generic catch-all. Consent language that's SMS-specific and documented at the point of opt-in. Opt-out handled right away and confirmed back to the shopper.
None of it is complicated once it's done right. Most of the trouble comes from doing it once, at the start, under whatever entity existed at the time, and never revisiting it as the business structure changed underneath it. Checking the registration against the current legal entity and campaign content every few months catches the drift before a carrier does.
How we handle it
CART registers the Brand and Campaign on a store's behalf, under the entity that runs the sends, and keeps monitoring registration health after approval instead of treating it as a one-time task. Trust Score, throughput limits, deliverability, and campaign-to-content match all get checked on an ongoing basis, so a mismatch that would otherwise surface as filtered messages months later gets caught before it costs a store anything.
Most of the delay described above traces back to a submission with something off in it: a mismatched EIN, an address that doesn't match public records, a use case description carriers can't map to a clear category. A team that has run this process hundreds of times knows what carriers flag before a rejection comes back, not after. That's the difference between guessing at the form and submitting an application that's already been checked against the same criteria carriers check.
A compliance team that submits clean applications avoids the manual-review queue. For the typical case, a store's EIN and LLC information check out clean, and CART gets Brand and Campaign approval done in 72 hours, not because the carrier process moves faster, but because the submission never triggers the review that slows everything else down.
That's the difference between registering once and moving on, and treating registration as infrastructure that needs the same attention as everything else running in production.
Not sure your registration would survive a carrier audit?
Run a Revenue Proof →