The road not taken keeps costing after you reject it
IBM licensed DOS non-exclusively in 1980 and the license outlived the business it was written for. eBay wrote down $1.4 billion of Skype two years after buying it. Fifty-three rejected options on record, priced.
The cheapest decisions in the record are terms, not cheques
Fifty-three counterfactual decisions are on record and thirteen were reversed, the second-highest count here at 63% persistence. These are the decisions where the alternative is knowable, because somebody took it.
IBM took a non-exclusive license for DOS in 1980. The verdict is one sentence: the non-exclusive license stood and the article's stated share collapse and eventual exit followed from it. No capital was committed, no factory was built, and the decision reorganized an industry. The exclusive version of that clause was available at the table and was not worth arguing for on the day.
RadioShack shipped a proprietary machine in 1983 and market share collapsed as compatibility became the basis of value. Both are the same decision seen from either side, and in both the expensive part was a term rather than a number.
The ones that bought the wrong thing with the right money
eBay acquired Skype in September 2005, believing buyers and sellers would want to talk before they traded, and in doing so allocated capital and attention away from defending its own marketplace advantages. It wrote down $1.4 billion in 2007 and sold 65% in 2009 at a valuation of $2.75 billion, below what it had paid.
AMD sold its fabs in October 2008 while near insolvency, taking a balance-sheet reprieve in exchange for its most valuable physical assets. The fabless model stood, and the original exclusivity terms were later bought out for $703 million.
Both companies got the money question right in the moment and the terms question wrong for a decade.
What the ones that held actually decided
Meta acquired the nascent threats in 2012. The acquisitions stood, Instagram became a major revenue source and a court declined to break Meta up. Berkshire Hathaway paid in its own stock in 1993 and the record notes plainly that the cost kept growing as Berkshire's price rose — a decision that held and was expensive, which is the ordinary case rather than the exception.
Bayer closed the Monsanto deal in 2018 and kept the acquisition, continuing to pay escalating litigation costs rather than reversing it. That is on the held list too.
The useful reading of this set is not that some companies chose well. It is that the option they rejected went on having a price, and the ones who did badly are the ones who stopped tracking it the day the decision was made.
How to run this decision
Drawn from what separated the records that stuck from the ones that were undone. A step nobody in the record can be seen taking is not here.
- 01
Write the rejected option down, with the date and the reason
One page. What was available, what it would have cost, why it was declined, and who decided. Most companies cannot produce this document a year later, which means the decision cannot be reviewed, only re-argued.
- 02
Separate the number from the terms
IBM's DOS decision had no number in it. Exclusivity, territory, duration, termination and renewal rights are where the long-dated value sits, and they are approved by people optimizing the number.
- 03
Put a date on re-pricing the road not taken
Twelve months out, ask what the rejected option would cost now and what it would have returned. This is cheap, and it is the only mechanism that converts a decision into a learnable one.
- 04
Name what the decision spends besides money
eBay's Skype acquisition spent attention that was not available to defend the marketplace. Capital is tracked and attention is not, and in this record attention is the more frequently misallocated of the two.
- 05
Check whether you are choosing under duress, and label it
AMD sold its fabs near insolvency. Decisions made to survive are legitimate and they should be recorded as such, because they will later be cited as strategy by people who were not in the room.
- 06
Decide in advance what would make you take the other road
Bayer kept Monsanto through escalating litigation. Whether that was right is arguable; what is not arguable is that the condition for reversing it was never written down, so every subsequent quarter was a fresh debate rather than a test.
What it looks like when this is going wrong
- The decision has no number in it, so no one senior reviews the terms.IBM, 1980.
- Nobody can produce the rejected option a year later.The thirteen reversed counterfactuals on record.
- The acquisition spends attention the core business currently needs.eBay and Skype, 2005.
- A survival decision is being described as strategy.AMD, 2008.
- The condition for taking the other road was never written down.Bayer and Monsanto, 2018.