Amazon Web Services (AWS)
AWS is invisible infrastructure that makes the internet possible.
AWS is invisible infrastructure that makes the internet possible. The company provides compute (EC2), storage (S3), and networking - boring backend services that thousands of companies build on. Revenue is $90 billion annually, but the economic value enabled is orders of magnitude larger. Startups that couldn't afford $100,000 in server costs in 2006 now scale to millions of users for $1,000/month. Remove AWS and vast portions of the internet don't just slow down - they suffocate.
What's remarkable is how AWS captures value. The company reduced prices 100+ times since launch, building trust and growing the ecosystem rather than extracting maximum rents. The strategy works because AWS operates at the infrastructure layer - capturing small percentages of massive transaction volumes. But the December 2021 us-east-1 outage revealed the strategy's vulnerability: companies with multi-cloud failover stayed online while single-provider dependents went dark.
AWS also pioneered flexible pricing: On-Demand instances ($1.00/hour with no commitment) versus Reserved Instances ($0.50/hour with 3-year lock-in). COVID exposed the trade-off - companies with on-demand flexibility scaled down instantly when revenue dropped, while reserved-instance customers paid for unused capacity. The lesson: shared infrastructure beats proprietary infrastructure, but don't let shared infrastructure become a single point of failure. And flexibility costs more upfront but pays dividends when environments shift.
Key Leaders at Amazon Web Services (AWS)
Jeff Bezos
Founder
Issued 2002 API Mandate that created preconditions for radiation
Andy Jassy
CEO (former AWS leader)
Led AWS from 57-person team to global dominance
Key Facts
Amazon Web Services (AWS) Appears in 14 Chapters
Filed for bankruptcy (2012) after systematically eliminating portfolio diversity through 1990s-2000s. Divested Sterling Drug, photofinishing, Eastman Chemical, medical imaging, cameras, and graphic arts - transforming from diverse conglomerate into pure-play film company as digital eliminated film.
Eliminating diversity before crisis →Failure case for assuming current conditions (film dominance) were permanent. Divested diversifying businesses in pursuit of focus, then collapsed when digital photography cycle displaced film demand.
Assuming permanence →Cited alongside Blockbuster, Nokia, and BlackBerry as example of rigid membrane preventing adaptation when technology shifted. Membrane mechanisms preserved identity during growth but became barriers to necessary change.
Rigid membranes preventing change →Chapter's central cautionary tale. In 1976, Steve Sasson demonstrated first digital camera - 26 years before bankruptcy. Tragedy wasn't failure to sense; it was transduction failure. Three blockers prevented action: incentive (70% profits from film), infrastructure (sunk costs), identity (film company culture).
Organizational neuropathy →Dominant film photography company destroyed when digital photography disrupted technological paradigm. Example of technology disruption causing organizational extinction despite previous market dominance.
Technology disruption as extinction →Failed marginal value test with 80% film market share (2000). Despite inventing digital photography, stayed in dying film market, refusing to migrate to digital patch. By 2012, filed for bankruptcy.
Refusing to leave the patch →Exemplifies catastrophic failure of letting politics override fitness data. By 2005, film division had fitness score 2/10 - they knew: film sales declining 15% annually, digital growing 40%. But film was 'strategic,' 'heritage,' a 'cash cow.' Fed $500M annually plus $3B cumulative R&D into dying division. Bankruptcy 2012.
Politics vs fitness data →Failed to kill low-fitness variants despite clear data. Film fitness score 2/10 (2005): sales declining 15% annually, digital growing 40%. But leadership protected film's budget, starving digital. Measuring fitness perfectly but letting politics override data.
Measuring but not acting →Exemplifies how successful niche construction becomes trap. For century, constructed film photography ecosystem (cameras, film, paper, chemicals, retail). By 1990s held ~70% film market share. Invented digital photography (1975) but couldn't abandon constructed niche. Filed bankruptcy 2012.
Constructed niche as tomb →Cautionary example of misreading gradient speed. Detected digital photography gradient in 1990s but treated it as slow gradient (waited 15 years to commit) when it was actually medium-speed (5-7 years). By the time resources reallocated, competitors had won.
Misreading change speed →Founded 1888, dominated film for century but became case study in quorum-sensing failure. Despite inventing first digital camera (1975), dismissed digital as minority behavior. Between 2000-2003, multiple thresholds crossed but recognized too late. Filed bankruptcy 2012.
Missing the threshold →Failed regeneration example. Spent years (1990-2012) trying to regenerate film business when reinvention or exit required. Wasted time and capital attempting to revive dead business model rather than accepting root system was obsolete in digital era.
Regeneration vs reinvention →Example of corporate freezing - maintained only 2% digital investment while market shifted to digital photography. Metabolism stopped, company died. Counterpoint to WeWork's overheating: both temperature extremes are fatal.
The freezing death →AWS holds 32% market share versus Azure (23%) and Google Cloud (11%) - capital-intensive digital territory with ~12% defensive intensity ratio.
See cloud infrastructure competition →