Interview
a16z Podcast | Searching for Mobile’s Own PageRank
- Mobile Discovery Paradox: The current mobile ecosystem resembles the "pre-PageRank" era of the web circa 1995 (Yahoo) and the "pre-Netscape" era, where no unified model exists for discovering or exploring content across diverse interfaces (apps, cards, web apps, URLs, messaging apps).
- Architectural Limitations: The native app model lacks the universal linking structure of the web (URLs); deep linking into apps often requires the app to already be installed, creating a "download-first" friction that prevents true content discoverability.
- Platform Power Shift: Smartphones have evolved into communication and social platforms in themselves (e.g., WhatsApp accessing phonebooks and photo libraries), inverting the desktop model where the browser was the primary platform and social apps were merely containers within it.
- OS Dominance: Operating systems (iOS, Android) currently hold the most power over discovery and interaction, with companies like Line or Facebook competing as secondary layers; this dynamic is shifting as wearables (Android Wear, Apple Watch, rumors of Apple Healthbook) move toward non-native, card-based, or XML-style interactions that do not require full app installation.
- Future Data Architecture (Urbit Model): A potential long-term shift (5–10 years) involves "local computing" where devices act as distributed data centers with internal directory structures (e.g.,
/contacts,/files), allowing apps to read/write shared local data automatically without custom APIs, reducing reliance on centralized cloud databases. - Hardware vs. Cloud Strategies: Apple is driving innovation through tight hardware-software integration (Bluetooth, fingerprint scanners), whereas Google is extracting value by lifting services into the cloud (Google Play Services, Google Now) and pursuing the Chrome OS vision, a model that mirrors Microsoft's historical lack of visibility into individual PC hardware.
- Android-Chrome Fusion: With Sundar Pichai overseeing both Android and Chrome, there is a plausible scenario where Android's Java-based APIs are rewritten as JavaScript APIs to create a unified platform, potentially rendering the concept of "installing" a native binary obsolete in favor of enabling cloud-based services.
- Performance Obsolescence: Technical arguments favoring native apps over the web regarding UX performance are becoming less relevant as resources improve (Moore's Law) and technologies like Google's V8 compiler advance, suggesting the "web" will eventually match native speed.
- Killing the Native App: Benedict Evans posits that Apple, rather than Google, will be the entity to fundamentally "kill" the native app model if necessary to strengthen the iOS platform's appeal and integration, despite Apple's current revenue reliance on the App Store.
- Sessile Computing Return: Future mobile computing will increasingly fuse with "sessile" experiences (stationary, high-power devices), suggesting a resurgence of laptops, desktops, and server farms (e.g., Oculus VR) where the distinction between mobile and stationary computing blurs.