How to turn failure and setbacks into the next step

1 The bottom line

Treat a failure as information about your tactics, not a verdict on your aim. Founders who changed the path, admitted the error fast and kept going turned early losses into Walmart, Nike, Nvidia and IKEA.

2 Key takeaways

• Keep the aim and change the target. A failed route says little about whether the goal was wrong.
  • Disney's Kansas City company went bankrupt, and he decided only his target was wrong. He moved to Hollywood and sold Alice at once Walt Disney, The Walt Disney Company31:00.
  • Pixar dropped hardware and went all in on computer animation, its dream from the start Ed Catmull, Pixar28:16.
• Admit the mistake early and cut it. Pride and sunk cost keep a dead path alive.
  • Musk called Tesla's heavy automation "my mistake" and tore robots off the line Elon Musk, SpaceX and Tesla1:00:51.
  • Singleton closed Packard Bell's television business once Japanese rivals made it unprofitable Henry Singleton, Teledyne33:03.
• Answer a blow with "what is the next step?" A plan stops panic and keeps your team moving.
  • Knight asked what he knew, what the future held and what to do next. That led to the Nike line Phil Knight, Nike43:43.
• Make failures cheap, visible and instructive. Progress depends on how many tries you get and what you learn from each.
  • SpaceX posted its landing crashes in a video called "how not to land an orbital rocket." Boosters now fly 20 times each Elon Musk and the SpaceX team35:46.
• Stay in the game through the worst stretch. Many payoffs arrive years after the pain.
  • Nvidia's stock fell 80 percent in 2008 and Huang kept building CUDA. The payoff came with AlexNet Jensen Huang, Nvidia24:50.
  • Kelleher fought four and a half years of lawsuits before Southwest flew in 1971 Herb Kelleher, Southwest Airlines12:58.

There's nothing wrong with my aim I got to change my target.

Walt Disney, The Walt Disney Company

2.1 Separate the aim from the tactic

When a route fails, ask what you still want, then pick a new route to it.

• Name your aim, then let the target move. This is how Walt Disney, The Walt Disney Company31:00 recovered.
  • His Kansas City company was bankrupt. He slept in his office and ate cold beans.
  • He sold his camera for a train ticket to Hollywood in 1923 and sold his Alice's Wonderland idea to a distributor right away.
  • The first six Alice shorts paid $1,500 each, and the next six paid $1,800 each.
• Refuse to feel whipped, and restart in the same game. Sam Walton, Walmart11:10 lost his first store and heard from his lawyer that he was finished.
  • He said he had found Newport and the store once, so he could find another good town and another Ben Franklin.
  • He rebuilt from scratch at 32. That restart led to Bentonville and eventually Walmart.
• Return to the original dream after wrong detours. Ed Catmull, Pixar28:16 shows how.
  • Pixar lost money for years selling hardware it had no experience making or marketing.
  • It abandoned hardware and bet everything on computer animation.
  • The pivot cost more than a third of its employees.
• Hold conviction about the goal, not the exact path. Bill Gates, Paul Allen and Steve Ballmer, Microsoft2:44:00 hedged.
  • Microsoft publicly backed IBM's OS/2 while quietly building Windows and Mac applications.
  • The market rejected OS/2 in December 1987. Windows 3.0 in 1990 then doubled Windows's DOS penetration in six months, and Windows became Plan A.
  • Windows 1.0 and 2.0 had reached only 5% of the DOS installed base.
• After a loss, aim at the next platform, not the current one. The Microsoft applications group2:20:50 learned this from Multiplan.
  • Multiplan targeted older 8-bit machines and lost to Lotus 123, which was built only for the IBM PC.
  • The group's rule was to never leave itself open to the next generation of technology. That is why Excel debuted on the Mac.

2.2 Admit the error and cut the loss

The faster you say "that was wrong," the more time and cash you keep.

• Own the mistake in public and reverse it hard. Elon Musk, SpaceX and Tesla1:00:51 did this on the Model 3 line.
  • He had pushed heavy automation and it did not work. He announced the error and marked robots with orange spray paint to remove them.
  • He added a rule: automate only after requirements are questioned and parts deleted.
  • Robots were thrown into the parking lot, and a hole was cut in the building to remove equipment.
• Keep a low resistance to change. Sam Walton, Walmart37:30 taught this to the whole company.
  • He called it the RC factor. At 61 he reversed his view that computers were overhead after his lieutenants convinced him.
  • In 1979 Walmart spent $500 million on a computer system linking stores, warehouses and distribution centers.
• Do not let sunk cost keep a business alive. Henry Singleton, Teledyne33:03 practiced this.
  • Teledyne bought Packard Bell for $16.5 million in stock and assumed $5 million of debt. It was profitable for years.
  • When Japanese makers grew too competitive, Singleton closed the television operation.
• Name exactly why the product was wrong. Larry Page, Sergey Brin and Eric Schmidt, Google59:43 did.
  • Google Video was a search service for TV content, at first without a player, and it reviewed uploads for one to two days.
  • YouTube won with easy upload, easy watching, embedding and good search. Google bought the product that worked.
  • One Lazy Sunday upload raised YouTube traffic by 83%.
• Scrap your own standard when the market has picked another. Jensen Huang, Nvidia41:57 had about nine months of cash left.
  • Microsoft built triangle-based Direct 3D into Windows and Sega walked away from the NV2 in 1996. Huang told his co-founders to throw out the plan and compete on performance inside the standard.
  • Nvidia's chips were also designed very tight on memory costing about $200, while rivals paid about $50 for more the same history39:31.
• Write your mistakes down every year. Bill Gates, Microsoft1:04:54 kept a memo for this.
  • He updated "the 10 great mistakes of Microsoft" annually.
  • Many were markets Microsoft entered later than it should have.

2.3 Turn the blow into the next step

Ask what you know and what comes next, then use the attack or the loss as material.

• Ask three questions before reacting. Phil Knight, Nike43:43 learned his supplier was lining up other US distributors.
  • He asked what he knew, what the future held and what the next step was.
  • He stayed with Onitsuka long enough to scare off replacements, develop other supply and launch his own brand.
  • He found where Adidas made shoes outside Japan and visited those factories.
• Reframe the break for your team. When Onitsuka cut Blue Ribbon off early, Phil Knight, Nike46:47 called it independence day.
  • His team had slumped during a recession. He pointed to late deliveries and ignored design ideas that the break would end.
  • He reminded them they had posted $2 million in sales the year before that had nothing to do with Onitsuka. The room sat up straighter.
• Regard every problem as a possibility. Ingvar Kamprad, IKEA48:49 faced a dealers' boycott.
  • Sweden's furniture dealers' association pushed IKEA out of trade fairs, banned its price ads and told suppliers to stop selling to it.
  • IKEA asked loyal suppliers for altered designs, sourced in Poland, then designed its own furniture.
  • Flat packing followed, and the first flat-packed table, Max, came out of it. It cut damage, freight and factory costs the IKEA history58:00.
• Review the business to find where the money is really made. After being locked out of FedMart, Saul Price, Costco44:27 did this.
  • He and his son Robert saw that almost all the margin was at the warehouse level.
  • They built Price Club around the warehouse, selling to small businesses.
  • Price Club stocked about 3,000 high volume items, against roughly 50,000 at Walmart and Kmart.
• Fix the method when quality fails. Soichiro Honda, Honda19:30 made piston rings by hand.
  • Toyota sent back 47 of the 50 he sold.
  • He turned to interchangeability and automatic equipment that relatively uninitiated women workers could run. He later used that experience for motorcycles.
  • Morris Chang treated a TI transistor line with near zero yield as a process to optimize. It reached 20% in about four months Morris Chang, TSMC25:47.

2.4 Make failure cheap and learn from it

Run more tries, at a risk level you can survive, and keep the lesson from each.

• Test against reality instead of analyzing forever. Elon Musk, Gwynne Shotwell and the SpaceX team17:17 argue you cannot think your way to perfect solutions.
  • Complex systems show emergent behavior only when bolted together, so SpaceX invests in prototypes, not studies, and treats failures as data.
  • After a three-failure start, each Falcon flight became the next test, with fixes incorporated immediately.
• Build so that losing one try does not hurt. Elon Musk, SpaceX21:35 says a high production rate solves many ills.
  • Stainless steel for Starship is cheap and weldable in a tent, unlike carbon fiber that needs giant autoclaves.
  • With Raptor engines built at high volume, losing one in a test does not set the program back months.
• Match the risk level to the stakes. Elon Musk, SpaceX19:50 runs three programs side by side.
  • Dragon carries crew and can never fail. On Falcon 9, ascent cannot fail but some landings may. Starship has no one on board, so failure is instrumental.
  • The groups talk to each other, so lessons flow between them.
• Show your failures and extract the lesson. Elon Musk and the SpaceX team35:46 published drone ship crashes set to music.
  • It signaled that failure is fine if you learn from it.
  • Those landing failures iterated into reuse of boosters 20 times each, and a Falcon 9 launches every two to three days.
• Let failures grow with the company. Jeff Bezos, Amazon1:02:00 argued winners follow a long-tail return distribution.
  • In baseball the payoff caps at four runs, but in business one swing can score a thousand, so big winners pay for many experiments.
  • The Fire Phone failure fed straight into building Echo and Alexa.
  • Amazon has sold roughly 500 million Echo and Alexa powered devices, according to the host.
• Treat misses as a brute-force way to find the path. Jeff Bezos, Amazon4:04:16 kept backing up and turning another way.
  • Auctions, zShops, the Fire Phone, the Sotheby's partnership and pets.com stakes all failed.
  • The hosts call it a philosophically straight line but a tactically random set of dots.
• Learn the market by shipping, then ask customers for the fix. Jensen Huang, Nvidia51:40 did this with Riva 128.
  • The chip crashed on about two thirds of the Direct 3D blend modes. Nvidia asked developers to use only the modes that worked, and they agreed because performance outweighed everything else.
  • Nvidia sold 1 million Riva 128 units within four months in 1997.
  • By contrast, Facebook's Beacon was "way too clever" first. It exposed purchases, was made opt in, and was killed two years later Mark Zuckerberg, Meta2:54:00.

2.5 Keep going through the long night

Many founders won because they stayed in the fight while others quit or ran out of nerve.

• Keep investing while others retreat. Jeff Bezos, Amazon4:09:00 built through the dot-com crash.
  • Amazon laid off 1,300 people in 2001 and set a public goal of profitability by Q4 2001.
  • The hosts argue 2000 to 2007 was when it built wide moats, while funding came cheap and rivals went under.
  • Q4 2001 revenue was $1.1 billion with $5 million of GAAP net income, and the stock jumped 25% in a day.
• Stay with the bet when the stock collapses. Jensen Huang, Nvidia24:50 faced an 80 percent drop in 2008.
  • Most CEOs would have shopped the company. He kept building CUDA.
  • Nvidia did not regain its 20 billion dollar peak until 2016.
• Do not declare game over. Jensen Huang, Nvidia1:56:30 survived at least four near death moments.
  • The hosts credit part of it to luck and strategy, and part to his willingness to sit through the pain.
  • After cutting about 70% to roughly 35 people, Nvidia shipped Riva 128 in months instead of the 18 to 24 months rivals needed. It used emulation hardware and software that cost a third of its cash the Nvidia history45:18.
• Keep trying new things until it works. Thomas Edison1:00:30 said other inventors try a few things and quit.
  • He tested thousands of filament materials and sent agents to the West Indies, Central America and the Amazon.
  • Frank McGowan spent 15 months exploring the upper Amazon for him.
• Pay for the fight yourself if you must. Herb Kelleher, Southwest Airlines12:58 kept Southwest alive in court.
  • Braniff, Trans Texas and Continental sued and won in two courts, and legal costs ate the investors' $543,000.
  • Kelleher offered to defer his fees and pay court costs from his own pocket. He argued the case to the Texas Supreme Court and won.
  • Southwest began flying on June 18, 1971.
• Count the years behind a sudden success. Ray Kroc, McDonald's36:40 was 52 when he started.
  • People marveled at his sudden success after the chain exploded.
  • He had sold paper cups for 17 years and multimixers from 1938. He called 30 years "a long long night."
  • Steve Jobs' bad calls at NeXT from 1985 to 1997 were prerequisites to the steadiness he showed later the Jobs biography6:00.

2.6 Where they differ

Cutting fast and enduring long both worked, so the facts of each case decide.

• Cut the loss or stay the course. The sources split on when to quit.
  • Henry Singleton, Teledyne33:03 closed a business once rivals made it unprofitable.
  • Jensen Huang, Nvidia24:50 kept building CUDA through an 80 percent drop, and the payoff came years later with AlexNet.
  • One test fits both: ask whether the facts behind the bet changed, as with Singleton's rivals, or only the mood around it.
• Fail fast only where failure is survivable. Elon Musk, SpaceX19:50 blows up Starship because no one is on board.
  • Dragon carries crew and can never fail, so the same approach does not apply there.
• Treat pain as fuel or as a cost. The founders disagree on how they feel about it.
  • Jensen Huang, Nvidia17:40 says pain and suffering refine a company's character.
  • He also says that, knowing the pain, vulnerability and shame, he would not start Nvidia again his later interview1:21:02. The superpower is asking "how hard can it be."
  • Ingvar Kamprad, IKEA48:49 framed a boycott as "a dizzying chance" instead.
• Being early is its own failure. Sam Walton, Walmart1:54:30 launched Hypermart USA, but only three stores were built and they were too big.
  • After his death, smaller Supercenters worked, and grocery became 55% of revenue.
  • Jeff Bezos, Amazon31:26 instead waits years once a customer need is meaningful and durable. The Kindle took from 2003 to November 2007.