Three platform transitions. Same two-beat pattern — an infrastructure moment that makes the shift technically possible, followed by a product moment that makes it culturally inevitable. The gap between the two compresses each time. So does the window for strategic positioning.
WAVE 1
Mainframe → PC
IBM ($8.2B loss, workforce halved). DEC (acquired at fraction of peak). Wang Labs (bankrupt, August 1992).
INFRASTRUCTURE
Intel 8080 (1974)
PRODUCT
Apple II (1977)
3 yrs
GAP
25 years
WAVE DURATION
WAVE 2
Client-Server → Cloud
Ballmer-era Microsoft (15yr stagnation at $300B). Blockbuster (bankrupt, 2010). Oracle (decade-long rebuild).
INFRASTRUCTURE
AWS EC2 (2006)
PRODUCT
iPhone + Netflix (2007)
<1 yr
GAP
15 years
WAVE DURATION
WAVE 3
SaaS → Agentic
In progress — $285B erased in Week 1. The sorting has begun.
INFRASTRUCTURE
Model Context Protocol (Nov 2024)
PRODUCT
Skills + Cowork (Jan 2026)
14 mo
GAP
5–7 yrs?
PROJECTED
ACCELERATING
Three Outcomes, Every Wave
Gerstners
Recognize the shift early. Accept revenue cannibalization. Rebuild around the new architecture.
Gerstner at IBM (1993) · Nadella at Microsoft (2014) · Netflix over Blockbuster
DECs
Execute too slowly. Become acquisition targets at a fraction of peak valuation.
DEC → Compaq ($9.6B) · Ballmer-era MSFT ($300B for 15 years)
Wangs
Fail to transition at all. Go to zero.
Wang Labs (bankrupt 1992) · Blockbuster (bankrupt 2010)