telecom · 1987–1999 · Motorola / Iridium LLC

Iridium's satellite phone constellation

How a program that delivered everything it promised, on time and on technical spec, still went bankrupt nine months after launch.

11 min read · 6 sources cited

// background

In 1987, three Motorola engineers — Bary Bertiger, Raymond Leopold, and Ken Peterson — sketched a constellation of low-earth-orbit satellites that could deliver voice service to a handheld phone anywhere on the planet. The system needed 77 satellites in its original design (after the element iridium, atomic number 77); it eventually flew with 66, but the name stuck. The technical premise was that ground-based cellular coverage would never reach much of the world's surface — oceans, deserts, polar regions, and most of the developing world — and a satellite system with cross-linked birds could fill that gap.

Motorola spun the program out as Iridium LLC in 1991 with a consortium of international partners and raised roughly $5 billion in equity and debt to build, launch, and operate the constellation. The technical program was, by its own internal measures, a remarkable success. First satellite launched May 1997. Constellation completed May 1998. Commercial service began November 1, 1998. Sixty-six satellites in 11 orbital planes, with cross-links between them so a call could be routed across the constellation and dropped to whichever ground station was closest to the recipient. No comparable system had ever been built.

The commercial launch was a disaster. The handsets weighed nearly a pound, cost about $3,000, and didn't work indoors or under tree cover. Service was billed at roughly $3–7 per minute. Iridium's business plan had been written in the late 1980s and locked through the early 1990s, when terrestrial cellular was still patchy, expensive, and analogue. By 1998 the cellular market had transformed: GSM had rolled out across Europe and much of Asia, U.S. carriers were aggressively building out digital networks, roaming agreements were proliferating, and the price of a cellular minute had collapsed. The market Iridium had been designed to serve had largely disappeared while the system was being built.

Iridium signed up roughly 10,000 subscribers against a business plan that needed about 500,000 to be solvent. The company filed for Chapter 11 bankruptcy on August 13, 1999, less than nine months after commercial launch — at the time the largest tech bankruptcy in U.S. history. The constellation was nearly de-orbited. Iridium Satellite LLC bought the assets out of bankruptcy in 2000 for $25 million, roughly 0.5% of the build cost, and the system has operated profitably as a niche service provider ever since.

// the decisions

1. Whether to freeze the system requirements in the early 1990s

By 1992 the technical architecture had to be locked to begin satellite manufacturing. Each satellite had a multi-year build cycle and the constellation needed dozens of them. The handset specification, the air interface, the billing model, and the price-per-minute economics were all derived from a market analysis Motorola had done in 1990. Cellular was already changing, but no one inside the program could say with confidence how fast.

options on the table

A.Freeze the spec and execute against the 1990 market analysis to hit the launch date.
B.Build in architectural flexibility — software-defined air interface, swappable handset radio modules, configurable billing — at higher cost and schedule risk.
C.Stage the program: launch a smaller proof-of-concept constellation first (8–12 satellites in selected planes), watch how the cellular market actually evolved, then commit the full build.

what they actually did

Motorola and the Iridium consortium froze the spec. The program was scoped and financed as a single integrated build to a fixed configuration; the partners' contracts and the $5B financing structure depended on the spec being stable enough to underwrite. There was no architectural off-ramp for incorporating major changes once construction began.

consequence

Between 1992 and 1998, U.S. cellular subscriptions grew from roughly 11M to 69M and average per-minute pricing fell by more than 50%. GSM coverage went from a regional curiosity to near-universal across Europe. By the time Iridium turned on service, the gap its handset was designed to fill had narrowed dramatically. The HBR teaching case explicitly identifies the 1992 spec freeze as the decision that determined the commercial outcome — long before the marketing problems of 1998–1999.

// what to take away

01Iridium delivered nearly every technical milestone on its original plan. The program-management failure wasn't execution — it was that the system requirements were locked against a market forecast that didn't survive the build cycle. 'On time, on spec, on budget' is not the same as 'on target'.
02Multi-year hardware programs implicitly take market-timing positions. The right place to surface that bet is at the architectural-flexibility decision, years before launch — not at the marketing-launch decision, weeks before commercial availability.
03The handset's physics envelope and the marketed use case were misaligned. Even if the cellular market hadn't shifted, a $3,000 single-mode handset that didn't work indoors was a poor fit for the business-traveller segment Iridium chased. 'The product works' and 'the product works where the customer is' are different statements.
04Bankruptcy didn't destroy the asset; it repriced it. Iridium Satellite LLC's profitable operation of the same constellation at 0.5% of the build cost is a clean case study in why sunk-cost reasoning blocks otherwise rational decisions about infrastructure.
05Iridium's failure scared investors away from satellite-broadband bets for nearly two decades. Starlink, OneWeb, and Iridium NEXT are all post-2015. The case is part of why aerospace-program PMs are now expected to discuss market-evolution risk explicitly in their funding pitches.

// timeline

1987Motorola engineers conceive the constellation; original 77-satellite design names the program after element iridium.
1991Iridium LLC formed as a separate consortium with international partners.
1992System architecture and handset spec frozen for manufacturing; financing structure executed against the spec.
May 5, 1997First Iridium satellites launched.
May 17, 199866-satellite constellation operational; cross-links live.
Nov 1, 1998Commercial service begins. Handsets ~$3,000; per-minute rates $3–7.
Q2 1999Subscriber count stalls near 10,000 against a plan needing 500,000+ for solvency.
Aug 13, 1999Iridium LLC files Chapter 11 — at the time, the largest tech bankruptcy in U.S. history.
Mar 2000Motorola informs the FCC that without a buyer the constellation will be de-orbited.
Dec 2000Iridium Satellite LLC buys the assets for ~$25M, anchored by a U.S. Department of Defense gateway contract.
2017–2019Iridium NEXT replacement constellation launched; the original commercial concept finally finds a sustainable form.

// sources

Iridium LLC (HBS Case 9-601-040) — Harvard Business School, 2001
Iridium World Communications, Form 10-K, fiscal year 1998 — U.S. Securities and Exchange Commission, 1999
Iridium LLC, Voluntary Petition for Chapter 11 (Case No. 99-45005, S.D.N.Y.) — U.S. Bankruptcy Court, Southern District of New York, 1999
Iridium's $5 Billion Wrong Number — The Wall Street Journal, 1999
Iridium, the Bankrupt Bet on Satellite Phones, Is Reinvented — The New York Times, 2000
Why Iridium Failed: A Retrospective — Roger Lanctot, Strategy Analytics, 2014

Practice this kind of decision

The simulator runs scenarios that exercise these same lessons under time pressure. Pick a chapter that exercises scope + risk.

3. The bankruptcy filing and the asset sale

August 1999. Iridium has roughly 10,000 subscribers, $1.5B of debt service due, and a daily operating cost (mostly satellite operations and ground stations) of about $7M. Motorola is the largest creditor and operator. The board's choice is to attempt a financial restructuring, or to file Chapter 11. Some advisors argue that a controlled de-orbit of the constellation might be the only viable end-state given that the operating costs alone exceed plausible revenue.

options on the table

A.Out-of-court restructuring with bondholders — would have required Motorola to commit additional capital.
B.Chapter 11 with intent to reorganize and continue operating the constellation.
C.Chapter 7 liquidation, controlled de-orbit of the satellites, asset sale of the spectrum and ground stations.

what they actually did

Iridium LLC filed Chapter 11 on August 13, 1999. Through 1999 and 2000 the company struggled to find a viable buyer; Motorola publicly notified the FCC in March 2000 that without a buyer the constellation would be de-orbited. In late 2000, Iridium Satellite LLC — a new investor group backed in part by a Department of Defense gateway commitment — bought the assets for approximately $25M.

consequence

The DoD's commitment to use Iridium for global communications anchored a much smaller, sustainable subscriber base. Iridium Satellite LLC operated the existing constellation profitably from 2001 onwards on a cost basis less than 1% of the original program. The replacement constellation, Iridium NEXT, launched 2017–2019, was financed conventionally and is now the operational system.

lesson

A program that fails commercially can still leave behind infrastructure that has value — but only if someone is willing to write down the build cost completely. Iridium's second life is the canonical example of why sunk costs are sunk: the constellation became viable as a business the moment its purchase price was disconnected from its build cost. PMs who can't get their leadership to write down sunk costs end up unable to make the right operating decisions about already-built systems.