Interview, Fireside Chat
a16z Podcast | The Topic That's Lasted the Entire History of Computing -- Bundling and Unbundling
- Core Thesis: The computing industry follows a perpetual cycle of bundling and unbundling driven by user context, screen constraints, and competitive pressures, rather than a simple pendulum swing.
- Trend in Western Markets (Unbundling):
- Smartphones favor unbundling due to limited screen real estate, shifting discovery dynamics, and the ease of switching between apps (e.g., Instagram and WhatsApp separating from Facebook).
- Apps are splitting from the web; features once contained in a single browser icon are now standalone modules.
- Motivation: Features added to a primary product (e.g., email in a calendar, photo editing in a messenger) rarely match the quality of dedicated standalone competitors, creating a "quagmire" where products become bloated yet fail to satisfy niche needs.
- Risk: Excessive unbundling creates "download shrapnel," where users install many apps but struggle to find the specific functionality they need or duplicate features across multiple applications.
- Contrast with China (Bundling/Aggregation):
- Chinese models (e.g., Baidu Maps) demonstrate that deep integration is viable by focusing on specific contexts (like location) rather than a "constellation of apps."
- Method: Apps aggregate diverse services (taxis, restaurant bookings, cinema tickets) based on user intent; users search for a goal (e.g., "restaurant"), and relevant options appear within the single context.
- Efficiency: This avoids the "infinite layers of aggregation" seen in the West (e.g., Uber, Lyft, Yelp, and TripAdvisor all carrying their own maps) and reduces the cognitive load on the user.
- User Behavior Statistics:
- Cognitive Limit: Humans utilize approximately 7 ± 2 apps at any given time, regardless of the millions available in stores.
- App Usage: Approximately 80% of smartphone time is spent within apps, while the remaining 20% is spent on the web; desktop patterns mirror this distribution.
- Long Tail: While most users use a single-digit number of apps daily, the app ecosystem's value lies in the long tail where niche apps serve specific professional workflows for small groups.
- Strategic Trade-offs for Product Managers:
- Bundled Approach: Easier to drive app installation and top-of-mind presence (e.g., Baidu Maps), but challenges arise in discovering specific features within the app.
- Unbundled Approach: Features are easier to discover independently, but the cost of persuading users to install multiple distinct apps is high.
- Metric Shift: Success should be measured by engagement and usage depth rather than raw download counts; focusing solely on downloads can lead to artificial behavior that degrades product utility.
- Monetization and Discovery:
- As apps mature, there is a risk of over-monetizing engagement, which often leads to poor design choices (e.g., aggressive prompts to download companion apps).
- Cost Dynamics: The marginal cost of creating software has collapsed by two orders of magnitude, but the cost of acquiring billions of users has not; success often relies on serving a niche deeply rather than scaling to the mass market.
- Future Outlook:
- The definition of "the web" is evolving into "the death of the URL," with embedded browsers and in-app navigation obscuring web metrics.
- Future product strategies must balance the "gravity" of existing user bases (which pulls R&D into monolithic apps) against the need for specialized feature discovery.
- The "7 ± 2" rule suggests that successful aggregation in the West will likely emerge by focusing on specific use-case contexts (social, location, productivity) that naturally bundle complementary services without requiring additional home screen icons.