If you only need to push a text from code you already run on AWS, Amazon SNS or AWS end-user messaging is a fair choice, and you may not need an alternative. If you're a software platform that texts for its own customers, you'll want something that also handles registration, opt-outs, a reply inbox and webhooks, and that's where a messaging platform like Ready earns its place.
The gap isn't about whether AWS can send a text. It can. The gap is everything around the send: who registers each customer's business with the carriers, where STOP replies go, how a support rep sees a conversation, and how your app hears about replies. On AWS those are parts you assemble. This post lists them honestly so you can decide which side of the line you're on. We make Ready, so weigh our view with that in mind.
What AWS gives you and what it leaves to you
AWS is a toolbox. You get a way to publish a message to a phone number, origination identities such as numbers, and the usual AWS controls for permissions and logging. For a team that already lives in IAM, queues and Lambda functions, that's comfortable.
What a platform product usually needs on top, in general terms:
- A registration flow per customer, so each business's texts are tied to its own brand and campaign.
- An opt-out store that every send checks, and that shows up in your own UI.
- A two-way inbox, so replies reach a person instead of a log stream.
- Webhooks or an events pipeline so your app reacts to replies and delivery changes.
- Compliance checks such as quiet hours, Do Not Call screening and consent records.
None of these are impossible on AWS. Each is a small project, and together they're a product you now maintain. Our post on build vs buy for SMS infrastructure walks through that tradeoff in more depth.
Registration is the part that bites first
US carriers expect A2P 10DLC registration for business texts. A platform with 200 customers has 200 businesses to register, each with its own brand details, use case and sample messages. Someone has to collect those details, submit them, watch for rejections and resubmit.
With Ready, each customer is a sub-account with its own numbers and its own carrier registration, and you can drive it through the API: POST /v1/registration/brand, POST /v1/registration/campaign and POST /v1/registration/confirm, with GET /v1/registration for status. Registration is one line on the bill, $25 a month per customer for the brand and its first campaign, and most approvals take 2 to 7 days. Ready registers businesses with an EIN only, so a sole proprietor without one can't register. Details are in the registration API docs and our 10DLC registration API guide.
If you're building on AWS, you design that intake and tracking yourself. That's real work, but it's work you control.
Opt-outs, quiet hours and consent
An opt-out has to stick everywhere. On Ready, an inbound STOP or UNSUBSCRIBE is honored, clear phrases like "please stop texting me" are caught too, and the number is blocked in the sub-account that received it. Before you send from another tool, or during a migration, check GET /v1/opt-outs/:phone for the per-number answer.
Quiet hours work on the recipient's area. Campaigns, drip campaigns and automations wait up to 24 hours for the window. A single API text is held only if the window opens within 6 hours. Otherwise the API answers 409 QUIET_HOURS_BLOCK and you send again once the window opens. A reply to someone who texted you in the last 24 hours goes out at any hour.
Consent works like this: when you set consent_attested on an API send, Ready records who sent, to how many, and when. That's your attestation, not a per-contact consent record. Ready doesn't store each recipient's opt-in time or wording, so keep that in your own form or CRM. Compliance stays the sender's responsibility either way.
The send itself
Here's the request shape. It's one call, with an idempotency key so a retry never sends twice:
curl -X POST https://api.tryready.com/v1/messages \
-H "X-API-Key: $READY_API_KEY" \
-H "Idempotency-Key: 6f1c2a9e-3b7d-4e0a-9c55-0d2b8a1e4f77" \
-H "Content-Type: application/json" \
-d '{"to": "+12125550199", "body": "Your order #1042 has shipped.", "consent_attested": true}' Leave from out and Ready picks the number: the one that contact already texts with, otherwise your number closest to their area code. Replies land in an inbox, and the inbound webhooks guide covers getting them into your app. The API quickstart covers the key, first send and webhook setup.
Side by side
| Need | Building it on AWS | With Ready |
|---|---|---|
| Send a text from code | Yes | Yes, POST /v1/messages |
| Per-customer registration | You design the intake and tracking | Sub-account per customer, registration by API or in the app |
| Opt-out handling | You store and enforce it | Built in, with GET /v1/opt-outs |
| Reply inbox for people | You build it | Built in |
| Quiet hours and DNC screening | You build them | Built in; screening is a per-account switch |
| iMessage | Not part of it | Blue line available |
| Deep control inside your own cloud | Excellent | Not the point of the product |
When AWS is the better call
Stay on AWS if your team already builds everything else itself, you want texting inside your own account's networking and billing, and you have engineers who will own registration and the opt-out store for the life of the product. Also stay if your volume is mostly alerts to your own staff, where a reply inbox and per-customer registration don't matter.
Move to a platform if you're hand-registering customers in spreadsheets, a support rep has asked where replies go, or an opt-out got missed because two systems disagreed. Those are the signs the glue code has become the product. Pricing is one Standard rate plus the carrier fee, itemized; see the pricing page for the figures, and note that screening is extra when switched on.
Frequently asked questions
Can I keep AWS for everything else and use Ready only for texting?
Yes. Ready is a REST API with an X-API-Key header, so it sits next to Lambda functions, queues and whatever else you run. Nothing requires you to move other workloads.
Does Ready replace my opt-out table?
For texts sent through Ready, the opt-out is enforced at send time and exposed through the API. If you also send from another tool, check the per-number endpoint before sending so the two don't disagree.
Is Ready right for sending one-time passcodes inside an AWS stack?
It can send them, but a pure alert or passcode workload with no inbox and no per-customer registration is the case where raw AWS messaging fits well. Choose on the shape of the product, not the logo.
Next step
Make a list of what you've built or would have to build around the send: registration, opt-outs, inbox, webhooks, quiet hours. If that list is short and you enjoy owning it, AWS is fine. If it's long, try one sub-account and one test send through the quickstart and see how much of the list disappears.