a16z Podcast: Searching for Mobile’s Own PageRank

Episode Summary

Executive Summary: The discussion argues mobile is still in an early, pre-Pagerank/pre-Netscape phase: app stores don’t scale as discovery tools, and the app model lacks the open linking structure that made the web usable. The speakers debate whether this is a solvable UX problem or a platform limitation, and explore futures like deep linking, data-centric apps, cloud-backed mobile, and wearable-driven interaction shifts.

Main Topics: Mobile as a pre-Pagerank discovery problem: Benedict Evans compares app stores to Yahoo in 1995: they list content but do not help users discover the right apps or experiences at scale. Mobile as a pre-Netscape / pre-web architecture: The speakers argue smartphones lack a browser-like universal layer for linking, navigation, and open standards, making app interactions fragmented across apps, cards, URLs, and messaging. Deep linking, notifications, and platform control: They discuss Google’s deep linking efforts, iBeacon-style context awareness, and how OS/platform owners like Apple, Google, and app ecosystems may control discovery and access. The future of apps as data views rather than installed code: A more radical model is proposed where apps operate on shared local data stores and file structures, turning services into synchronized caches rather than isolated databases. Cloud vs device: Apple and Google diverge: Apple is portrayed as pushing innovation into integrated hardware/software and local sensors, while Google pulls capabilities into cloud services such as Google Now and Play services. The meaning of 'install' may disappear: The speakers speculate that in a few years 'installing an app' may no longer mean downloading native code, but merely enabling service access across the platform stack. A counterintuitive future beyond mobility: Balaji Srinivasan adds that a meaningful part of the future may be 'sessile'—VR/Oculus users sitting still with powerful computers—reviving desktops, laptops, and server farms.

Key Arguments: App stores are poor discovery mechanisms because they scale like directories, not search or ranking systems. Smartphones are not just small PCs; they are communications platforms with built-in social and sensor capabilities that shift power to the device/OS layer. The lack of a browser-like, universal linking model on mobile prevents the open, composable experience the web enabled. Deep linking helps, but it is a partial workaround because users still need the app installed. Context-aware experiences (for example, surfacing restaurant reviews automatically near a venue) will replace manual app-hopping. A future mobile architecture may center on local data stores where apps read/write shared personal data and services sync from the cloud as caches. Apple may eventually 'kill the native app' by choosing platform integration over the current app-download model. Performance objections against web-based UX may matter less over time as hardware and software efficiency improve. Google’s and Apple’s platform strategies are diverging: Google centralizes in cloud services, while Apple emphasizes integrated hardware/software and sensors. The mobile future may include more than mobile devices; VR and other stationary computing experiences may become a large category.

Data Points: App store scale analogy: "a million apps" - Benedict compares app-store browsing to browsing too many items to be useful, like Yahoo in 1995. Web scale analogy: "100,000 websites" - Used to illustrate why directory-style browsing does not scale for discovery. Mobile platform timeline: "20 years ago" - The current mobile discovery problem is likened to the early internet era. Future horizon: "five years, 10 years" - Benedict estimates the data-centric mobile architecture shift could happen within this timeframe.

Pivotal Quotes: ""It's like looking at Yahoo in 95."" — Benedict Evans: Describing app stores as weak discovery tools that do not scale well. ""What would install mean? You know, what would app actually mean?"" — Benedict Evans: Questioning whether the current notion of installing native apps will still make sense in the future. ""The reason that we only had the web was all that there was was because it was so hard to do anything else."" — Benedict Evans: Arguing that desktop web dominance was driven by historical constraints, not inevitability.

Implications: The mobile stack may shift from installed apps toward context-aware, linkable, data-centric services controlled by OS and cloud providers. Discovery, distribution, and even the meaning of an app could be redefined, with major consequences for developers and platform power.