The standard advice for adding SMS to a product is straightforward. Sign up for Twilio, Vonage, or a similar aggregator, provision a number, and pay per message. It works. The API is well-documented, the onboarding is fast, and you can be sending within an hour. For a prototype or a low-volume use case, there is nothing wrong with this approach.
The problems surface when volume grows, when your use case requires a number your customers already know, or when per-message billing starts to feel like a tax on every interaction your product has with a user. At that point, the question worth asking is not “which aggregator should I use” but “do I need an aggregator at all”.
What you give up with a shared short code
A shared short code is a five- or six-digit number used by multiple businesses simultaneously. Recipients see the same number regardless of which business sent the message. There is no persistent identity. A patient who saved your clinic's number does not see that number. They see a generic short code shared with dozens of other senders. The relationship context that makes SMS valuable is stripped out.
Dedicated long codes are better, but they are still leased infrastructure. The number belongs to the provider. Pricing can change. The provider can shut down, be acquired, or change their terms. If you have built your customer communication around a number you do not own, you are one business decision away from a migration.
The Android SIM model
The approach is to take an Android phone with an active SIM, install a small agent app, and connect it to an API. The phone's native SMS capability does the sending. The number is your SIM's number, one you control, that your customers can save, and that replies come back to. The carrier relationship is yours directly.
Relayion provides the API layer, which handles queuing, delivery tracking, webhook dispatch for inbound messages, and device management. Your code sends an HTTP request and the SIM delivers the message. The infrastructure is yours in the most literal sense.
What changes with flat pricing
Per-message billing is an incentive misalignment. The provider profits when your users interact more. Conversational workflows, confirmation exchanges, and multi-step AI interactions are expensive not because they are technically costly but because each turn is a billable event. Flat monthly pricing per device removes that misalignment entirely. You can design your messaging workflows for user experience, not for minimising turn count.
There is a real trade-off here. You become responsible for device uptime. The phone needs to be plugged in and connected. That is operational overhead that an aggregator handles for you. For most businesses, mounting a device on a charger in a fixed location is not a significant burden. But it is worth being honest that the model trades managed infrastructure for direct control, and that trade-off is not right for every situation.
When it makes sense
The Android SIM model is the right choice when your customers recognise and trust a specific number, when your volume is high enough that per-message billing adds up, when your use case involves conversational exchanges that multiply message count, or when you want to avoid vendor lock-in on a channel that is central to how your business communicates. If none of those apply, an aggregator is probably the simpler path.
If some of them apply, it is worth understanding what you are actually trading away each month on your current provider's bill.