There is no app to install and no account to create. RCS is a protocol layered into the messaging app that is already on the phone, which is why it reaches people who would never download a business's own channel.
How RCS messaging works, step by step
- Capability check. Before the message is composed, the messaging app asks whether the recipient's number can receive RCS. This is a lookup against the RCS network, not a guess, and it is the step that decides everything after it.
- Route selection. If the answer is yes, the content goes over IP to an RCS backend. If it is no, the same message downgrades to SMS or MMS on the cellular channel.
- Delivery. The backend hands the message to the recipient's messaging app, which renders the rich content: cards, carousels, buttons, media at full resolution.
- Feedback. Delivery status, read receipts, and any button taps travel back the same way, which is the part SMS never had.
The check is per recipient and per conversation, not per campaign, so one send can be rich for some people and plain for others in the same moment.
The backend, and who runs it
Google's Jibe platform operates RCS for most of the world's carriers and for Google Messages, alongside a smaller number of carrier-run deployments. Apple Messages connects through supporting carriers rather than running its own RCS backend, which is why RCS on iPhone arrived carrier by carrier after iOS 18.1 rather than everywhere at once.
This concentration is why RCS reaches Android users even where a carrier never built its own implementation: the backend and the client both already exist. See what is the Google Jibe platform and how RCS routing works.
What changes for a business message
The path adds one actor and one guarantee. The message leaves the business, or its platform, as a verified RCS agent rather than as a phone number, and Google has approved that agent's name, logo, and checkmark before it can send to anyone. The recipient sees a brand rather than an unknown number, and the branding is rendered by the phone from the approved record, so it cannot be faked by putting a company name in message text.
The guarantee is fallback. A business send is one instruction, not two campaigns: rich where the handset can take it, SMS or MMS everywhere else, reported per channel so you can see which version each person received. See what is an RBM agent and how RCS falls back to SMS.
Why it needs data, and what happens without it
Because RCS is IP-based, it needs Wi-Fi or mobile data. That is also why it does not consume a texting allowance, and why it works on Wi-Fi in a basement with no cellular signal.
With no data at all, the message is not lost. It takes the SMS path instead, which is the same safety net that covers handsets and carriers without RCS. The trade is what the plain version cannot carry: no read receipts, no typing indicators, compressed images, and 160 characters per segment instead of 3,072. See does RCS use data or Wi-Fi.
- Transport: IP-based, not the cellular voice or SMS channel, which is why RCS needs a data connection.
- The capability check runs per recipient before the send, and it is what decides rich delivery against fallback.
- Backbone: Google's Jibe platform operates RCS for most carriers and for Google Messages; Apple Messages connects via supporting carriers.
- Business messages travel as a Google-verified RCS agent, so the brand identity is rendered by the phone rather than claimed in the text.
- Fallback to SMS or MMS is the universal safety net and the reason RCS never costs you reach.