Acer's 1992 'fast-food' reengineering turned a loss-maker into a global top-7 PC firm

Acer had pioneered Taiwan's PC industry, but by 1990-92 it was a company that could not earn: since listing in 1988 growth had stalled into break-even or losses, and in 1991 it laid off about 300 people in Taiwan. Founder Stan Shih concluded assembly had become the lowest-value part of the business, and that Acer's problem was structural, not cyclical.

In 1992 Shih launched the first 'reengineering' (再造宏碁) built on two ideas. The first was the smile curve: value concentrates at research and branding, while assembly sits at the bottom of the curve — so Acer should abandon local assembly and concentrate on the high-value ends. The second was the fast-food model: Taiwan became the central kitchen producing standard components, shipped to 39 local assembly lines in 35 countries, with regional business units as independent 'masters' and headquarters as their servant.

The model worked with brutal speed: by 1995 Acer was making about 4 million PCs with revenue near $5B, the world's seventh-largest PC maker, with ROE above 30% for two straight years and per-employee revenue of $463,000. Shih's book on the reengineering ('Reengineering Acer') became a Chinese business classic.

By 1995: ~4M PCs, ~$5B revenue, global No.7 PC maker, 30%+ ROE two years running — from a company that could not make money

Moving value up the smile curve beat cutting cost: abandoning low-margin assembly and concentrating on R&D and brand raised the economics of every unit sold, not just trimmed losses.

The fast-food model outsourced the heavy step: shipping standard parts to local assemblers avoided import duties, currency swings and inventory risk while local partners took the market-specific pain.

Client-server organisation freed the front line: regional units as 'masters' made market calls fast, and headquarters serving instead of commanding was the organisational expression of the strategy.

When the value curve moves under you, do not optimise the dying position — redraw it: push activity to the high-value ends and make market-facing units the masters

Acer briefly became the world's third-largest PC brand in the late 1990s before the post-2000 PC wars, but the 1992 reengineering is still taught as the canonical Chinese-language turnaround. Shih's smile curve became the standard diagnosis of electronics-industry value distribution.