Casting turns a phone into the way people find something and a television into the way they watch it, which is how a large number of households actually behave. It also reaches televisions your apps do not.
Searching a catalogue with a directional pad and an on screen keyboard is genuinely unpleasant, and viewers avoid it wherever they can. What they do instead is find the thing on the device already in their hand, where typing is easy and browsing is fast, and then move it to the screen they actually want to watch on. Casting exists to serve that behaviour rather than to be a lesser alternative to a television app. It is worth designing for properly, because for a meaningful share of households it is not a fallback at all. It is the normal route, and a service that treats it as an afterthought feels broken to exactly the people who use it most.
A cast button where the viewer already is, rather than a separate mode they have to know exists. If it has to be discovered, most of your audience never will.
Pause, seek and volume from the device in their hand, which is a better remote control than the one that came with the television.
Casting works on televisions and streaming sticks where you have no app at all, which quietly widens your addressable audience without another build.
This is the part viewers misunderstand and it changes what they expect. When something is cast, the television fetches the stream itself rather than receiving it via the phone, which is why quality is not limited by the phone connection, why the battery is not drained, and why somebody can leave the app or take a call without playback stopping. It also means the television is doing the adapting, so the rendition it selects reflects what the television can receive rather than what the phone could. That is usually better, since the television is generally on a stronger connection. How that selection works is covered under adaptive bitrate streaming.
A cast stream is still a stream, which means it is checked the same way as any other. The viewer has to be entitled to the title, the territory still applies, and the session counts against the concurrent screen limit on their plan. That last point is worth stating plainly, because it is the question operators ask first and the one viewers occasionally hope the answer to is no. It also means casting cannot be used to bypass a quality cap, since the ceiling belongs to the plan rather than to the device. The enforcement behind all of that is described under conditional access, and the screen counting under device management.
Casting and television apps solve overlapping problems and neither makes the other unnecessary. An app gives you the lean back experience, a presence on the platform store, and a service that works when nobody has a phone to hand, which matters more than it sounds in a household where the television is shared. Casting gives you reach onto devices you have not built for and a far better search experience for everybody. The services that get this right offer both and make the handoff between them invisible, so somebody who cast last night and opens the television app tonight finds the same position waiting. That continuity comes from continue watching, and the apps themselves from TV apps.
Have questions about taking advantage of this limited-time offer? Check out the FAQ for answers.