The shark that never appeared
In 1974, Steven Spielberg was filming Jaws off Martha's Vineyard with a mechanical shark that would not work. Saltwater wrecked the hydraulics. Days went by without a usable shot of the creature the entire film was about.
He had two options: wait for the machine, or make the film without it.
He made it without it. The shark became a yellow barrel moving across the water, a two-note figure in the score, and the faces of people watching something the audience could not see. When the shark finally appears, it is nearly the end of the film.
Jaws became the highest-grossing film of its time and changed how horror is made. The effect everyone remembers is the one Spielberg never planned.
His first instinct was to show the shark. Losing that instinct was the best thing that happened to the film.
Source: Jaws production history — Wikipedia#Production)
The restaurant that closed for half the year
Ferran Adrià's elBulli, on the Catalan coast, held three Michelin stars and was named the best restaurant in the world five times. It was also shut for six months of every year.
Not for a holiday. During the closed months the team worked in a Barcelona laboratory doing nothing but experimenting — no service, no guests, no requirement that anything they made be edible or sellable. Then they reopened with a completely new menu.
Two million people a year applied for roughly eight thousand seats.
A restaurant's first instinct is to serve food. Adrià banned that instinct for half of every year, and that is where the food came from.
Source: elBulli — Wikipedia
Amazon writes the press release first
Before Amazon builds a product, someone writes the press release announcing it. Not a specification, not a business case — a press release, dated in the future, written for the customer, describing the finished thing as though it already exists and people already love it.
Then the team works backwards from that document to figure out what has to be built.
If the press release is not convincing, the product does not get made. The argument is that it is much cheaper to find out an idea is weak while it is still a page of text.
Most teams start from what they can build and hope someone wants it. This starts from what someone wants and works back to what to build.
Source: Working Backwards — Amazon
The protein bar that started as a chocolate bar
Every nutrition company works on the same problem: how do you make a protein bar taste good?
The founders of Fulfil, in Ireland, asked the question from the other end. How do you make a chocolate bar more nutritious?
Same destination — something people enjoy eating that is also good for them. But starting from confectionery rather than nutrition produced a different product. It looks and tastes like a chocolate bar, with the protein built in, rather than a health product trying to taste acceptable.
Hershey bought a minority stake in 2019. In 2022, six years after the company was founded, Ferrero — the owner of Nutella and Kinder — acquired the business outside North America, reportedly for more than €150 million.
Two teams can aim at the same result and arrive somewhere completely different, depending on which end they start from.
Source: Ferrero to acquire Fulfil Nutrition
Dyson and the bag
In 1978, James Dyson was vacuuming his house when the machine lost suction — again. He opened it up and looked at the bag.
Every vacuum cleaner ever built had one. Nobody in the industry questioned it. The bag was simply what a vacuum was.
Dyson asked what the machine would look like without one. It took five years and 5,127 prototypes to find out, each failure teaching him a little more about how air actually moves. The result was the cyclone vacuum, and eventually one of the most valuable companies in Britain.
The bag was never essential. It was only ever a default, and nobody had checked which one it was.
Source: Failure is the best medicine — BBC Future
LEGO sells the theme parks
By 2003 LEGO had expanded into theme parks, clothing, video games and television. It lost roughly $300 million in a single year and was carrying around $800 million of debt.
The new chief executive, Jørgen Vig Knudstorp — appointed in October 2004, the first person from outside the founding family to run the company — reached an uncomfortable conclusion. The company had forgotten what made it different in the first place. The obvious move was to diversify further and find new revenue. He did the opposite.
He sold the theme parks, closed the retail and publishing operations, and roughly halved the number of individual pieces LEGO produced. The brick colours, which had started with the yellow, red and blue Piet Mondrian used, had grown to more than fifty.
Everything that remained was built around the interlocking brick. LEGO is now the largest toy company in the world.
Rescuing the company meant dismantling most of it, and being honest about which part was actually holding the rest up.
Source: How Lego clicked — The Guardian
Fujifilm looks at what it already had
Digital photography destroyed the market for camera film. Kodak, which had dominated it, filed for bankruptcy in 2012.
Fujifilm was in the same business facing the same collapse. It did not try to defend film, and it did not simply race into digital cameras where it had no advantage. It asked a harder question: if the product disappears completely, what do we actually know how to do?
The answer had nothing to do with photography. Film is largely collagen. Decades of work had gone into antioxidant chemistry to stop ultraviolet light fading printed photographs — the same ultraviolet damage that ages human skin. The expertise had been built for one purpose and was worth a great deal somewhere nobody would have looked for it.
Fujifilm launched Astalift, a skincare line, in 2007. Healthcare and materials are now core to the company.
They did not copy anyone. They asked what would survive if the product vanished, and found the company was never really about film.
Source: How Fujifilm survived — CNA Insider
Why does the robot stop?
A robot on a Toyota production line kept shutting down. The obvious fix was to replace the blown fuse and restart it.
Instead, someone asked why five times.
Why did it stop? A fuse blew. Why did the fuse blow? The circuit was overloaded. Why was it overloaded? A bearing was not being lubricated properly. Why not? The oil pump was not circulating enough oil. Why not? The pump intake was clogged with metal shavings.
They fitted a filter. The problem never came back.
Replacing the fuse would also have worked — until the following week, and the week after that.
Every first answer is a symptom. The question is how many times you are willing to ask before you stop.
Source: Five whys — Toyota Production System