What RCS changed, and what it did not

RCS — Rich Communication Services — is the GSMA's replacement for SMS: a carrier-run messaging standard with rich cards, suggested replies, typing indicators and read receipts, delivered through the phone's default messaging app.

It spent roughly a decade as the technology that was about to matter.

Apple added RCS support in iOS 18, released in September 2024. That single decision moved the standard from "Android only, therefore ignorable in many markets" to something approaching universal on modern handsets.

The GSMA has continued to develop the standard since: Universal Profile 4.0 was finalised in March 2026, adding video calling initiated from a thread and richer text formatting, and Universal Profile 4.1 followed with a focus on secure messaging foundations. End-to-end encryption work has been published as part of the same programme.

It is still carrier-mediated. Reach depends on operator deployment, not just handset support. A market where operators have not launched RCS Business Messaging is a market where your rich message is a fallback SMS.

It is still A2P. Registration, sender verification and content rules apply. Rich formatting does not exempt a promotional message from being a promotional message.

The fallback is mandatory design work. Every RCS programme is really two programmes: the rich one and the SMS one it degrades to. If the fallback is an afterthought, a meaningful share of your audience receives the afterthought.

The honest framing is that RCS is a presentation upgrade on an existing channel, not a new channel. It uses the same address (a phone number), the same regulatory apparatus, and the same commercial relationships.

That makes the business case much easier to evaluate than a new platform would be. You are asking: is a richer render of the message I already send worth the incremental cost and the dual-path complexity?

For a delivery notification with a track-package button, frequently yes. For a one-line OTP, almost never — and OTPs are a large share of A2P volume, which is why the transition is gradual rather than sudden.

Whether verified sender identity in RCS becomes meaningful enough to move trust. SMS has no real sender authentication, and that is why it is the preferred channel of every phishing operation on earth. If RCS delivers a visibly verified brand, the value is not the carousel — it is that the recipient can tell it is you.

Reach in RCS is not one number, and conflating the three is where most confusion comes from.

Handset support. Does the device speak RCS at all. Since iOS 18 this is close to universal on modern phones.

Operator deployment. Has the recipient's carrier turned it on. This varies enormously by market and is the gate people forget.

Business messaging availability. Can a business send to that subscriber through the operator's RCS Business Messaging arrangement — a further, commercial gate on top of the technical one.

A vendor quoting "RCS reach" without saying which of the three it means is quoting the largest number available. Ask for the third, for your markets.

The feature most likely to justify the work is not the carousel. It is that the recipient can see who is writing.

SMS has no sender authentication. That is why it is the preferred channel of every phishing operation on earth, and why banks spend enormous sums teaching customers to distrust messages that look exactly like theirs.

A verified brand identity changes the economics of that problem. If your messages are the ones criminals imitate — banking, delivery, government — the value of a channel where impersonation is harder can exceed everything else on the feature list combined.

Nobody switches to RCS. They add it, run both, and shift the mix over years.

Budget for two content variants per message, two previews to review, two failure modes to monitor, and a capability lookup in the send path. Start with the messages where the rich format changes the outcome — anything with a next action — and leave one-time codes on SMS, probably permanently.

• iOS 18 removed the "Android only" objection; operator deployment is now the constraint.
• Reach is three gates, not one. Ask about the third.
• Verified sender identity is the underrated feature, especially for impersonated brands.
• Every RCS programme is two programmes. Design the fallback first.