Netflix

By 2007, Netflix had a very good problem: a business that worked. DVDs arrived by mail, customers loved the convenience, and the company was profitable. Blockbuster, meanwhile, was the giant — thousands of stores, brand recognition Netflix could never buy.

Reed Hastings faced a decision most executives avoid: build a technology that could make his own core business obsolete, on purpose, before anyone forced him to.

Streaming in 2007 was nothing like it is now — slow internet, thin content libraries, and DVDs still made real money. Cannibalizing a profitable, working business to chase an unproven one is one of the hardest calls in business. Most boards say no. Many executives don't even bring the idea to the table, because on a spreadsheet, it looks like self-sabotage.

Blockbuster had the chance to buy Netflix outright in the early 2000s, for a fraction of what it would later be worth, and passed — not out of blindness, but because it genuinely didn't see mail-order DVDs, let alone streaming, as a serious threat. By 2010, Blockbuster had filed for bankruptcy, undone by the very disruption it once had the option to own.

Netflix kept building streaming through the 2000s, investing in it long before it was the obvious bet — racing against its own most profitable division. That's the uncomfortable part founders rarely want to hear: the biggest threat to a good business is often the next version of itself, and somebody is going to build it. The only real choice is whether that somebody is you.