Ericsson single-sourced its radio chips from one Philips fab, and lost 7 million phones when it burned

According to case accounts of the incident, on the evening of March 17, 2000, a lightning strike hit a power line in Albuquerque, New Mexico, touching off a fire at a Philips semiconductor plant.³ It was, by every honest account, a small fire. Ericsson's own company history says the blaze was minor enough that staff extinguished it before the fire brigade even arrived.¹ The academic case write-ups hedge that accounts vary, but agree it was brought under control in minutes.³ Eight trays of silicon wafers got wet and sooty.³ Within eighteen months, one of the two companies that depended on that plant would no longer make phones under its own name.

The story you've probably heard is that a ten-minute fire cost Ericsson the handset business. That's a good story, and it's mostly wrong. The fire didn't lose Ericsson anything the same fire didn't also throw at Nokia. What lost Ericsson the phone business was a decision made long before the lightning struck — and a second decision made in the days after, when the smoke had already cleared.

One fab, no backup, and a whole product line hanging on it: the fire only mattered because Ericsson had built a supply chain with a single door

Ericsson had streamlined. To simplify its supply chain, it had made Philips the sole provider for the radio-frequency chips at the heart of its phones.⁷ Single-sourcing is a real strategy with real benefits — better prices, tighter integration, one throat to choke. It also means you have built a system with exactly one point of failure. When the fire hit, no alternate vendor had been qualified, and qualifying a new chip supplier is not a phone call; it is months of testing and redesign. So production nearly halted.⁷ Despite Ericsson scaling up chipset manufacturing in Sweden 'in record time,' by its own annual report's phrasing, the resulting shortfall was estimated at 7 million phones.¹ Seven million phones is not a fire's worth of damage. It is a strategy's worth.

Here is the fact that dismantles the whole 'act of God' framing: Nokia depended on the very same plant. The two companies together took roughly 40 percent of Albuquerque's chip output.⁴ Same fab, same fire, same week, same missing chips. If the fire were the cause, both would have burned. Only one did. The difference was never the disaster. It was what each company had built to absorb one.

Ericsson took Philips at its word — and lost the days that mattered: the second failure wasn't the supply chain; it was the org chart moving too slowly to matter

Philips told Ericsson production would be back on track within about a week. Ericsson accepted that and took no action.⁵ It kept the problem from senior executives and assumed the disruption would be short-lived.² That assumption is the whole ballgame. When you single-source, your one remaining defense is speed — the ability to notice trouble and throw the full weight of the company at it before the gap becomes a hole. Ericsson's hierarchy did the opposite: the news traveled slowly upward, and by the time it reached people who could act, the smart moves were already gone.

Nokia's team had a crisis plan in place and ran an aggressive, multipronged response.⁴ It immediately redesigned its circuit board to cut dependence on Philips and re-engineered phones to accept American and Japanese chips.¹⁵ And it didn't wait on reassurances — management went directly to Philips' Amsterdam headquarters to demand extra resources.¹ The results diverged completely: Nokia kept its line largely unaffected,⁴ while Ericsson swallowed a loss. Philips itself, the company where the fire actually happened, absorbed only about $40 million in lost revenue.² The party that got hurt most was the one furthest from the flames.

How much did the fire cost? Depends what you're counting: the numbers swing wildly because they measure different things, and only one of them is really about the fire

Beware anyone who quotes you a single tidy 'cost of the fire.' The figures range enormously because they measure different things. One case-study figure -- repeated across several secondary write-ups rather than confirmed in any primary disclosure -- puts the direct fire-related loss at more than $400 million in annual earnings.⁵⁶ Ericsson's full mobile-phone division loss for 2000 has been cited at 'nearly $1.68 billion' in one retrospective,⁷ and, via Yossi Sheffi's book, at $2.34 billion for the division at year-end -- a secondary-sourced figure that doesn't fully reconcile with the $1.68 billion estimate or with Ericsson's later group-wide numbers.² Those larger numbers are the whole division's annual result — and that division was also absorbing well-documented product delays and cancellations that had nothing to do with New Mexico. Ericsson's own consumer-products chief told Reuters the company would cancel the T36 model and refocus on the R520.⁷ The fire is a chapter in that year's story. It is not the whole book.

Wasn't Ericsson already losing anyway?: the honest counter is that the handset business was sliding before the smoke — which is exactly the point

The fair objection cuts against the popular story and in favor of ours: Ericsson was already in trouble. Nokia had overtaken it in turnover by 2000, and Ericsson's handset models were suffering delays and cancellations that owed nothing to any fire.⁷ So maybe the fire is a footnote and the phones were doomed regardless. But that concession is the whole thesis, not a refutation of it. A company already stretched thin is precisely the one that cannot afford a supply chain with a single door and a hierarchy too slow to open it in time. The fire didn't create Ericsson's fragility. It exposed it — and a competitor standing in the same fire proved the fragility was a choice, not fate.

The end came in stages, not in a blaze. In October 2000 the board launched a 'Back to Business' program that relocated manufacturing out of Sweden and the US toward Eastern Europe.¹ A year later, in October 2001, Ericsson folded its standalone handset unit into the Sony Ericsson joint venture — a decision driven by many compounding factors, of which the Philips fire was a significant contributor rather than the sole cause.⁸ For all of 2001, LM Ericsson reported its first annual loss, nearly $2 billion.⁸ The handset business Ericsson had built on its own ended quietly, absorbed into a partnership.

The most durable myth about Ericsson's phones is that a freak accident took them down — a bolt of lightning, a ten-minute fire, bad luck. It's comforting because it absolves the strategy. But the fire was the exam, not the verdict. It asked both companies the same question: what happens when your single supplier goes dark? Nokia had studied. Ericsson had bet that the question would never be asked. Build a system with one door, and you are wagering the whole business that nothing ever burns in front of it. Something always does.