What the book argues

Christensen's disk-drive research - chosen because the industry's generations turn over fast enough to study like fruit flies - shows the pattern with brutal clarity: incumbents win almost every sustaining battle (making good products better for existing customers) and lose almost every disruptive one (simpler, cheaper products that start in markets too small to matter). The failure is structural: the disruptor's early market is unattractive by every metric a well-run incumbent uses to allocate resources.

The mechanism is the value network: companies are held captive by the customers and cost structures they serve. Disruptive products underperform on the mainstream metric (capacity, in disk drives) but win on a different one (size, ruggedness, price) valued by a fringe market - then improve faster than the mainstream's needs grow, until one day they are good enough, and the incumbent's advantage evaporates in a product generation. Excavators, steel minimills, and motorcycles trace the same curve.

The prescriptions are organizational, not exhortative: you cannot disrupt yourself from the core P&L, because resource allocation will always starve the small opportunity. Spin out an autonomous unit sized to be excited by small wins, let it find the market by discovery-driven planning (plans for learning, not execution), and accept that the right early market is unknowable in advance. Agnostic marketing - watching what customers do with the product, not asking them what they want - completes the toolkit.