Netflix says its custom video encoding method and use of VP9 and H.264/AVC High codecs are saving 36% bandwidth on Android and 19% on iOS for offline downloads
When Netflix introduced the ability to download content onto mobile devices for offline viewing earlier this week …
Context & Ripple Effects
This lands five days after Netflix switched on downloads for all streaming plans, and the company is already publishing efficiency numbers for that feature — evidence that offline viewing was engineered as a bandwidth problem from day one. It extends a pipeline Netflix started a year earlier, when it began re-encoding its entire catalog to save users up to 20% of data.
The split result — 36% on Android versus 19% on iOS — tracks the codecs named in the announcement: VP9 gains more traction where Android devices handle it, while the H.264/AVC High fallback carries more of the load on iPhones. The earlier cellular data usage setting showed Netflix giving users manual bitrate control; this move does the same work invisibly at the encode level.
First-order effects
- Subscribers downloading Netflix originals onto phones store the same titles in roughly a third less space on Android and about a fifth less on iOS, directly easing storage and data-cap pressure the company had already acknowledged with its bitrate-restriction setting.
- Netflix's own delivery costs per offline download drop immediately, since every title served through the new encode ladder consumes less bandwidth than the same file under the prior pipeline.
Second-order effects
- Rival streaming services face a widening efficiency gap on mobile: if Netflix serves comparable quality at lower bitrate, competitors must either invest in similar codec work or accept higher CDN bills and worse performance on constrained connections.
- The Android-iOS disparity puts pressure on Apple's codec roadmap — the platform where Netflix saves least is the one where better decode support would deliver the biggest user-visible win.
Third-order effects
- Encoding shifts from a fixed format decision to a continuously tuned system: within months Netflix extended this approach to AI-based encoding that matches compression to scene content, suggesting per-title and per-scene optimization becomes table stakes for large catalogs.
- If bandwidth-per-hour becomes a competitive metric, codec selection turns into strategic infrastructure — favoring platforms and services that control their full encode-and-deliver stack over those renting generic pipelines.
The trend: Streaming video is consolidating around services that treat compression engineering as a core differentiator, with each codec generation resetting the quality-per-bit frontier across mobile platforms.