RIM paid NTP $612.5 million in 2006 to buy off a BlackBerry shutdown that was never ordered

In the winter of 2006, a federal judge in Virginia held in his hand the power to switch off every BlackBerry in the United States. Not the phone in your pocket — the whole service: the push email that had made the device an addiction for lawyers, bankers, and half of Capitol Hill. Judge James R. Spencer was weighing whether to enter an injunction against Research In Motion's U.S. operations, and the company facing him had run out of appeals.⁴ On March 3, RIM stopped fighting and wrote a check: $612.5 million, one lump sum, to a company that made nothing, sold nothing, and had never shipped a single device.³

The official story is that RIM lost a patent fight and paid up. The truer story is that RIM was not buying an admission of guilt — it explicitly admitted none — and it was not even buying a favorable read of the patents. It was buying one thing: the certainty that nobody could turn off the lights.

The company on the other side of the table made nothing: NTP owned patents, not products, and that is exactly what made it dangerous

NTP was a patent-holding company, built around wireless-email patents originally developed by an inventor named Thomas Campana Jr. with a patent attorney, Donald Stout. Campana died in 2004, two years before the case he started was settled.⁵ That detail matters more than it seems. A company that sells products can be countersued, out-competed, or squeezed on price — it has a business to protect. NTP had no business to protect. It had a legal right and a courtroom, and against a company whose entire service depended on continuing to operate in the U.S., a legal right to an injunction is a gun to the head.

The legal ground was not hypothetical. In 2004 the Federal Circuit largely affirmed that RIM's BlackBerry system infringed NTP's patents — even wrestling with the awkward fact that RIM's relay hardware sat in Canada, and holding that a patented 'system' could still be 'used' inside the United States.¹ In 2005 the same court denied RIM's petition for a full-panel rehearing and sent the case back down for injunction proceedings.² That remand is the whole engine of the 2006 crisis: appeals exhausted, the question was no longer whether RIM had infringed, but whether the judge would flip the switch.

Why a rational company overpays for certainty: the value of a settlement isn't the merits of the case — it's the size of the downside you're removing

Here is the calculation RIM was actually running, and it has nothing to do with whether the patents were good. By 2006 RIM was a large, fast-growing, highly profitable company with a subscriber base expanding by the month. A U.S. shutdown would not have been a line item — it would have been the removal of the product itself from its most important market, overnight, with no way to know for how long. Against that, $612.5 million was a very large but survivable one-time cost.³ The settlement was never framed as an existential event for RIM, and it wasn't one. It was insurance bought at the moment the house was already on fire.

The twist: the patents mostly fell apart afterward: the patent office gutted the very claims RIM had paid to license — years too late to help

This is the part that makes the story look, in hindsight, like a robbery. In the years following the 2006 settlement, the U.S. Patent and Trademark Office reexamined NTP's asserted patents and rejected or narrowed most of the key claims that had formed the basis of the infringement case.⁶ Read that again: the legal weapon that forced a $612.5 million payment was later disarmed by the government office that issued it in the first place. If RIM had held out, so the second-guess goes, it might have paid nothing at all.

But that reframes the decision as a question RIM could not have answered in time. The reexamination outcomes arrived years after the check cleared. In early 2006, the patents were valid, in force, and backed by an appeals court that had already ruled against RIM. You cannot settle a case using evidence that does not yet exist. The invalidation was real, and it was also useless to the man deciding in February whether he could afford to gamble the U.S. market on a bet that the patent office would eventually change its mind.

So was $612.5 million just a bad bet?: the honest counter is that RIM had leverage it may have left on the table

The fair objection is that RIM had more room than it played. The reexaminations were already underway; a company with RIM's cash and legal firepower might have pushed harder for a stay, or gambled that no judge would truly dark-out millions of government and enterprise users, some of whom relied on BlackBerry for official communications. There is a version of the counter-history where RIM waits six more months, the patent office moves, and the whole thing evaporates. That version is not crazy — it just wasn't knowable. The honest answer is that RIM paid a certainty premium, and certainty premiums always look overpriced from the far side, once the risk you insured against turns out not to have materialized. That is what insurance always looks like after the fire is out. The question is never whether you needed it afterward. It's whether you'd have slept at night without it.

BlackBerry never went dark. That is the fact most retellings get exactly backwards — there was no U.S. outage, because RIM paid to make sure there never would be one.³⁴ The $612.5 million bought a service that kept running, a market that stayed open, and an end to a threat that, at the moment it was live, was total. That the patents later crumbled is a fine footnote about the patent system. It is a lousy verdict on the decision. RIM didn't buy the patents. It bought the guarantee that a man in Virginia couldn't switch off its company on a Tuesday — and in early 2006, that was the only thing money could actually buy.