6 questions to go. Your result appears here.
What the record shows
Three questions the record says matter most: why you are separating the business, which part you are letting go, and what the rest of the company still needs from it.
Selling to raise cash is where the regret is: three of four went badly
Of 22 completed separations with a known outcome, three went badly for the company that made them, and all three were made to raise cash: Pan Am, Palm and LEGO. Pan Am sold its Pacific routes to fund its own survival; Palm licensed out its operating system for short-term money; LEGO sold its theme parks to pay down debt.
Of the 18 separations made to focus on the core, to unlock value or because investors or regulators pushed for them, none went badly and seven paid off as intended.
The one cash separation that did not go badly was The Washington Post Company, which sold its newspaper and continued as an education company, Graham Holdings.
Separating something the rest of the company relies on is the riskiest kind
Where the rest of the company relied on the business being separated, for technology, customers, a brand, supply or its future growth, four of eight separations went badly or were called off: Netflix, Gap, Pan Am and Palm. Where the business stood on its own, none of six did.
It is not a rule. Two separations of a linked business paid off as intended: PepsiCo and General Electric. PepsiCo, letting its restaurant chains go, kept a supply contract to keep pouring Pepsi in them.
Let go of the past, not the future
Where the business separated was the slower-growing, older or lower-margin part, four of eleven separations paid off as intended and none went badly (Netflix called its separation off). Where it was the faster-growing or more promising part, none of five paid off, and Gap and Pan Am went badly or were called off.
The pattern behind the separations that paid off is a company shedding what it had been, to concentrate on what it was becoming: PepsiCo, Novartis, Hilton and Merck.
The largest separations are the ones that get called off
Three separations in the record were announced and then called off: Netflix, Gap and Toshiba. All three were splits of half the company or more. Netflix canceled Qwikster 23 days after announcing it; Gap kept Old Navy and booked $189 million in charges; Toshiba's plan failed to win the required shareholder majority.
A large separation is a long public commitment. Customers, shareholders and the cost of untangling shared operations all get a vote before it is done.
What did not separate outcomes: the company's condition
Companies under strain when they decided went badly or called it off in three of nine cases; healthy companies in two of seven. Sales went badly more often than spin-offs (two of five, against one of 17), but every sale that went badly was a sale to raise cash.
Why the company is separating the business, and which business it is, matter more than the company's health or the form of the deal.
Warning signs, in the sources' words
- The case for the separation is about the cash it raises. “It sold its best business to fund its own survival.” (Pan Am: Pacific Division (1985))
- The business you would let go is one the rest will need. “Once the OS was split off and licensed to rivals, Palm had to buy its future back and never regained the fused advantage.” (Palm: PalmSource (2003))
- Costs, systems or customers are shared more deeply than the plan assumes. “The shared cost base fused to Old Navy made separation too costly and complex to execute.” (Gap: Old Navy (2019))
- You can imagine wanting it back. “The family bought the same parks operator back fourteen years after selling it.” (LEGO: LEGOLAND theme parks (2005))
Questions for the board before it decides
If we did not need the money, would we still separate this business?
Three of four separations made to raise cash went badly. Of 18 made for other reasons, none did.
What does the business we keep get from this one today, and how will it get that after the separation?
Where the rest of the company relied on the business, four of eight separations went badly or were called off. Where it stood on its own, none of six did.
Five years from now, which of the two businesses will be growing faster?
Four of eleven separations of the slower part paid off; none of five of the faster part did.
What would it cost to buy this business back, and what would make us want to?
LEGO sold its parks in 2005 to pay down debt, then bought the operator back fourteen years later.
Which shared costs, systems and contracts have we not yet priced into the separation?
Gap called off its Old Navy spin-off and booked $189 million in charges: the shared cost base made the separation too costly and complex to execute.
Who has to agree before this is done, and what would make each of them say no?
All three separations in the record that were called off were splits of half the company or more.
What it is not
A forecast. The record holds 30 separations researched from published sources: company filings and statements, and reporting. Of those, 22 were completed with a known outcome, 3 were called off, and 5 are too recent to call. That is too few for percentages, so the test shows you the cases instead. Well-documented separations lean toward the famous ones, and where the sources do not say, for example, whether the rest of the company relied on the business, that case is left out of that question rather than guessed.