The Company That Invented Its Market
You run an American semiconductor company. Your first great product, a memory chip, created an industry. You are now losing that industry to foreign competitors.
Do you fight to stay in the business that defines your company, or leave it?
The situation
It is 1985. Your company was founded in the late 1960s, and its early memory chips helped create the semiconductor memory business. For years memory was your largest product and your engine of manufacturing learning: memory chips are made in enormous volumes, and making them well teaches a company how to improve its processes.
In the early 1980s, Japanese manufacturers entered the memory market with large investments, high quality and aggressive pricing. Their share has risen rapidly. Prices have fallen steeply, and in the downturn of the mid-1980s they have collapsed. Your memory business is losing money. Several American competitors have already left or gone under.
What you know
You also make microprocessors, the chips that run computers. A few years ago, one of your microprocessors was chosen for a personal computer from the largest computer company in the world, and that design has become the industry standard. Clone makers are now building compatible machines, all of which need your processors. Microprocessors earn far better margins than memory, and your next generation is close to production.
Manufacturing capacity is scarce. Your middle managers allocate production according to the margin each product earns per wafer, and for some time that rule has been sending more capacity to microprocessors and less to memory, without any formal decision at the top.
What you don’t know
You do not know how long the memory price war will last, or whether trade negotiations between your government and Japan will change the terms of competition. You do not know whether the personal computer market will keep growing, or whether your processor will remain its standard. You do not know whether a company that stops making high-volume memory can keep its manufacturing skills sharp.
The voices in the room
Some of your most senior people argue that memory is the company. It is what you invented, what your engineers are proudest of and what drives your manufacturing. Others argue that you are fighting a war you cannot win against rivals with more capital and patience, while a far better business is growing under your nose. Leaving would mean closing plants and laying off thousands of people.
What happened
The company was Intel. In his book Only the Paranoid Survive (1996), Andy Grove, then Intel’s president, described a conversation with Gordon Moore, its chief executive, in the middle of 1985. Grove asked what a new chief executive would do if the board threw them out and brought in someone else. Moore answered that the newcomer would get Intel out of memories. Grove’s response, as he recounts it, was to ask why the two of them should not walk out the door, come back in and do it themselves.
Intel left the commodity memory business in 1985, keeping specialised memory products, such as EPROMs, for a time. The transition was painful: it closed plants and laid off more than 7,000 employees, nearly a third of its workforce. It concentrated on microprocessors. The 386, launched in that period, and its successors made Intel the dominant chip maker of the personal computer era.
Judging the decision apart from the outcome
This is one of the rare cases where both the process and the outcome were good. The process deserves study on its own. The leaders separated the economic question from the question of identity by imagining a successor with no history to defend, a direct technique against loss aversion and sunk costs.
The organisational scholar Robert Burgelman, who studied the episode closely, found that Intel’s exit had been partly prepared from below. Middle managers allocating capacity by margin had been moving the company towards microprocessors for some time. Top management’s decision ratified and completed a shift that the organisation’s internal rules had already started. That is dispersed knowledge at work: the people closest to the numbers saw the change first.
A strong process does not guarantee success. Had the personal computer market stalled, the exit would have looked like a retreat into a niche. But the reasoning, leaving a business where the company had no lasting advantage for one where it had a structural position, was sound given what was known.
What a good process would have weighed
The options of fighting, seeking protection or deferring each had a case. Fighting required capital and a cost advantage the company lacked. Trade protection might change prices but not the underlying competitive position. Deferring the choice would preserve options but bleed money and attention. The exit traded a defining identity for a structural advantage, and the advantage was visible: Intel’s processor was already the standard.
Sources and limitations
The conversation between Grove and Moore is Grove’s own account, written a decade later; it is a participant’s recollection, not a contemporary record, and is paraphrased here. Burgelman’s study is the main scholarly source on the internal dynamics. The layoffs figure is as reported by NPR. Other financial figures are omitted because consistent figures were not verified.
The reveal
Intel, 1985
Leave the commodity memory business and concentrate on microprocessors, keeping a few specialised memory products.
Paid off
The company left the commodity memory business in 1985, cut deeply, including layoffs of more than 7,000 people, nearly a third of its workforce, and became the dominant maker of microprocessors for personal computers.
Strong
Judged on what could be known at the time, separately from the outcome. See moral luck.
The forces pulling at the people deciding
-
Memory was not just a product line; it was what the company had been. Leaving felt like losing a self, which made an economically clear choice emotionally very hard.
-
The decisive move was a thought experiment. The two leaders asked what a newly appointed chief executive, with no attachment to the past, would do. The answer was obvious once the question was asked.
-
Middle managers allocating scarce manufacturing capacity by profit margin had already been shifting it away from memory towards microprocessors. The organisation had partly decided before its leaders did.
-
Years of investment in memory technology and plants pulled towards continuing. Treating them as sunk was necessary to see the choice clearly.
Signals that were visible at the time
- Memory prices had collapsed and the company was losing money on memory.
- Its microprocessors were the standard inside the leading personal computer design, with a major new generation on the way.
- Manufacturing capacity was already being allocated by margin, and memory was losing out internally.
- Foreign competitors were investing at a scale the company could not easily match.
When identity and economics disagree, ask what a successor with no attachment to the past would do.
Follow the thread
Connections
Where this comes from
Sources
- Andrew S. Grove, Only the Paranoid Survive (Currency Doubleday, 1996), chapter 5.
- Robert A. Burgelman, “Fading Memories. A Process Theory of Strategic Business Exit in Dynamic Environments,” Administrative Science Quarterly 39, no. 1 (1994), 24–56.
- NPR, “Intel Legends Moore And Grove. Making It Last” (2012). Source for the layoffs figure.
Teaching this case? The teaching note has learning objectives, a 60-minute session plan, discussion questions and the key facts on one page.
Shrikant Kumar. “The Company That Invented Its Market.” Decide case, Telos. https://shrikantk.com/decide/the-memory-maker/