The seesaw diagram from the slides captures a non-obvious truth: winning customers away from a dominant platform isn't as simple as building a better product. The incumbent has staying power (switching costs), complementary products (an entire ecosystem built around it), exchange value from its existing user base, and technological functionality. A new entrant has only technological functionality — one factor against four. This is why so many technically superior products have failed to displace dominant platforms.
Apple's Mac has had better reviews than Windows PCs for decades — more stable, better design, less malware. Yet Windows has maintained roughly 90%+ market share throughout that period. The Mac's technical superiority is vastly outweighed by Windows' network effects (more software written for Windows), switching costs (corporate IT infrastructure is Windows-standardized), and complementary products (enterprise software, corporate VPNs, business applications). A new entrant needs to offer improvements so dramatic that they overcome all four incumbent advantages simultaneously — which is an extraordinarily high bar.
The slide's formulation — "products have features, platforms have communities" — is more than a marketing tagline. It captures a fundamentally different business model and competitive dynamic. A product competes on features: whoever has the best feature set wins. A platform competes on ecosystem size and developer participation: whoever attracts the most complementary goods wins. This distinction changes every strategic calculation.
Roblox (Instapoll #3) is a platform because users — not Roblox — create the games and experiences. Roblox provides the infrastructure, marketplace, and virtual currency (Robux), and takes a cut of transactions. The more developers build games on Roblox, the more players join. The more players join, the more developers build. This is a two-sided market with cross-side network effects. By contrast, a game like Call of Duty is a product — Activision builds all the content, and when it stops releasing new content, the value stops growing. Platforms scale fundamentally differently than products because third parties create the value.
Instapoll #5 — "What has happened to the digital camera market?" — is a story of brutal convergence. The smartphone didn't just compete with digital cameras; it made the dedicated digital camera category nearly irrelevant for most consumers. The iPhone launched in 2007 with a 2-megapixel camera. By 2012, smartphone cameras had improved enough that Kodak filed for bankruptcy. By 2025, the only remaining mass-market digital camera buyers are photography enthusiasts — everyone else uses their phone.
This is classic convergence: once-separate markets (phones and cameras) merged into a single device that serves both needs. The smartphone then enveloped GPS devices (eliminating Garmin's consumer business), portable music players (ending the standalone iPod), portable gaming devices (shrinking Nintendo's handheld market), alarm clocks, calculators, flashlights, and dozens of other single-purpose devices. Each envelopment destroyed an industry. The pattern: a platform (the smartphone OS) enables developers to replicate the functionality of dedicated single-purpose products, at near-zero marginal cost, on hardware consumers already carry everywhere.
Instapoll #4's answer is simple but profound: winner-take-all platform markets with network effects and zero marginal cost. Bill Gates (Windows platform), Mark Zuckerberg (social network), Jeff Bezos (marketplace platform), Larry Page and Sergey Brin (search platform), Eric Yuan (video conferencing platform). Every major tech billionaire built or dominated a platform market where network effects drove winner-take-all dynamics and software's zero marginal cost turned dominant market share into extraordinary profitability.
Contrast with industries without these properties: retail (Walmart), manufacturing, restaurants, construction. These industries can make people wealthy, but rarely produce billionaires from a single company because they lack zero marginal cost (each sale still requires real goods or labor) and winner-take-all dynamics (competitors can always open another store or factory). Software platforms, by contrast, can serve a billion users with essentially no incremental cost, and network effects tend to concentrate market share into one or two dominant players. That combination — scale without cost, plus structural monopoly tendency — is the billionaire-making formula.
Adjacent business models have more similarities than differences. The spectrum runs from one-sided (single user class) to fully multi-sided platforms.
| Market Type | Network Effect Type | Value Driver | Winner-Take-All Tendency |
|---|---|---|---|
| One-Sided e.g., Messaging, Waze |
Same-side exchange benefits — users benefit from interacting with each other | Size of single user community | Very high — users flock to biggest network |
| Two-Sided Marketplace e.g., Uber, Airbnb |
Cross-side exchange benefits — each side benefits from the other growing | Balance and quality of both sides | High — but geography can fragment markets |
| Platform e.g., iOS, Windows |
Both same-side (user community) and cross-side (developers + users) | Complementary goods ecosystem (apps, software) | Extremely high — dominant standards persist for decades |
Key insight: the more sides a market has, the more complex the network effects — but also the more defensible the platform becomes once it achieves critical mass on all sides.
Konana's model (2007) shows software stacked in layers — each layer depends on the one below it. Control the bottom layers and you control what runs on top.
Source: Konana, Prabhudev 2007
| Layer | What it does | Who controls it? | Strategic importance |
|---|---|---|---|
| Hardware | Physical compute, storage, networking | Intel, AMD, Nvidia, Qualcomm, Apple Silicon | Nvidia controls ~80–95% of AI training chips — enormous power |
| Operating System | Manages hardware, enables software to run | Microsoft (Windows), Apple (iOS/macOS), Google (Android/ChromeOS) | OS control = platform control = app store control = revenue |
| Middleware | Bridges OS and applications; enables portability | Browser vendors (Google Chrome, Apple Safari), cloud platforms | Whoever controls middleware can reduce OS relevance |
| Applications | Deliver value directly to users | Thousands of ISVs (Independent Software Vendors) | App ecosystems determine platform attractiveness to users |
| Strategy | How it works | Real-world example |
|---|---|---|
| Move Early | First mover advantage in network markets can be decisive — early users attract more users before competitors can reach critical mass | Yahoo! Auctions won in Japan before eBay arrived; Amazon built AWS before Microsoft Azure launched |
| Subsidize Adoption | Give the product away free (or below cost) to accelerate user growth and network effects | PayPal paid new users $10 to sign up; Zoom gave K-12 schools free unlimited access during COVID |
| Viral Promotion | Design the product so using it promotes itself — every transaction, message, or interaction creates awareness | Skype's "Powered by Skype" branding; WhatsApp's message footers; Uber's referral codes; Airbnb's social sharing |
| Social Proof | Leverage the influence of visible adoption — when people see peers using something, they're more likely to try it | Blue Apron used social sharing; Zoom's ubiquity during COVID created massive social proof across all demographics |
| Redefine / Expand the Market | Bring in new user categories who weren't part of the original market, growing the network's total addressable size | Nintendo Wii attracted non-gamers (seniors, families) into gaming; iPhone converged phone + camera + music + maps |
| Form Alliances | Partner with competitors in adjacent markets to create a combined network large enough to challenge a dominant player | Didi/Ola/GrabTaxi/Lyft alliance to counter Uber globally; NYCE bank ATM network to counter Citibank |
| Encourage Complementary Goods | Invest in helping third parties build products that make your platform more valuable | Oculus and Amazon Echo developer funds; Apple Swift Playgrounds for iOS development training |
| Backward Compatibility | Ensure new technology works with products built for the old standard, protecting existing ecosystem investment | Apple Rosetta for PowerPC-to-Intel transition; Samsung Pay supporting existing mag-stripe terminals |
The Zoom case study illustrates multiple strategies simultaneously: freemium (subsidize adoption) + friction-free signup (viral promotion) + social proof + K-12 giveaway (expand the market) + pandemic timing (convergence moment). No single strategy won — the combination did.