Home / Blog

Careers

Your career is a startup

Reid Hoffman set out to become a philosophy professor. He co-founded LinkedIn instead, after a product job at Apple and a startup of his own that failed. In The Start-Up of You, he and Ben Casnocha argue that a career works like a startup that never ships a final version, and that the backup plan has to exist before the first plan breaks.

Chen-Yang Su · ·  

Main takeaways
  • Hoffman and Casnocha call a career that stays open to revision permanent beta. Hoffman’s own plan moved from academia to product management to founding companies, and each step grew out of the one before.
  • They describe a competitive advantage as three gears that only turn together. Your assets are one gear and your aspirations a second. The third is what the market needs.
  • In Mark Granovetter’s survey of 54 people who found a job through a personal contact, only 9 had seen that contact often. The other 45 heard from people they saw occasionally or rarely.
  • Before putting his savings into his first startup, Hoffman asked his father whether he could move into the guest bedroom if it failed. That was his Plan Z. His Plan B, PayPal, grew out of walks with Peter Thiel and Max Levchin.
  • Casnocha judges a risk by its worst case. He is open to a risk whose worst case he could survive, and wary of one that could leave him bankrupt with his reputation destroyed.

In the summer of 2004, Sean Parker was raising money for Facebook, then a website for college students run by the 20-year-old Mark Zuckerberg. Parker later wrote in a tribute in TIME that Reid Hoffman was the first person they called. Hoffman had launched LinkedIn the year before. He was impressed by Facebook, and he also judged that leading its first investment while running another social network would be a conflict, according to his profile at the Academy of Achievement. He arranged a meeting with his friend Peter Thiel instead. Thiel agreed to lend Facebook $500,000, which would convert into a 10.2% stake.

Hoffman and Thiel knew each other from Stanford. In the New York Review of Books, Anne Diebel describes the undergraduate Thiel, who “stayed up late sparring with Reid Hoffman, a progressive from Berkeley he had met in a philosophy class.” Hoffman later called his choice to step back from Facebook’s round “the most expensive decision” of his business career. Thiel’s half-million dollars reached Facebook through a friendship that began in a philosophy class.

How Facebook met its first investor in 2004 Four cartoon people stand in a row. On the left, Sean Parker holds a phone next to Mark Zuckerberg. A dotted line labelled first call runs from Parker to Reid Hoffman, who runs LinkedIn. A thought bubble over Hoffman says leading this round would clash with LinkedIn. A blue arrow labelled introduction runs from Hoffman to Peter Thiel, with the note friends from a Stanford philosophy class. A money bag travels from Thiel back to Facebook along the ground, labelled a $500,000 loan that converts to a 10.2% stake. Sean Parker Mark Zuckerberg FACEBOOK, 2004 Reid Hoffman RUNS LINKEDIN Peter Thiel INVESTOR First call Introduction friends from a Stanford philosophy class $ $500,000 loan, converting to a 10.2% stake

Thiel agrees to lend $500,000, which converts into a 10.2% stake. Hoffman later called stepping back “the most expensive decision” of his business career.

Figure 1. Press the steps. The sequence and numbers are from Parker’s TIME tribute, Hoffman’s Academy of Achievement profile and Kirkpatrick’s account. The thought bubble paraphrases Hoffman’s reasoning and is not a quotation. The people are cartoons.

Eight years later, Hoffman and Ben Casnocha published The Start-Up of You. The book argues that a person should manage a career the way a founder manages a startup. Hoffman’s own career supplies many of its examples.

There is no final version

Software companies release beta versions of products they expect to keep changing once people use them. Hoffman and Casnocha borrow the idea for careers.

Permanent beta

Permanent beta is Hoffman and Casnocha’s name for treating a career as a product that is never finished. In a podcast conversation about the book, Hoffman describes it as being “an infinite learner.” The direction can hold for years while the route changes with what each step reveals.

Hoffman’s route changed several times. He graduated from Stanford in 1990 in symbolic systems, a program that combines philosophy with computer science, and earned a master’s in philosophy at Oxford in 1993. He meant to become a professor and a public intellectual. At Oxford he concluded that academic philosophy would reach too few people, and he went looking for “a more direct impact on society,” in the words of his Academy of Achievement profile. In 1994 he joined Apple to work on eWorld, an early online service.

Reid Hoffman, release notes

PHIL CS symbolic systems

Version 1 · 1990

Stanford

  • Graduates in symbolic systems, which combines philosophy with computer science.
  • Plan A is to become a professor and public intellectual.
  • Keeps a friendship with Peter Thiel from a philosophy class.

Version 2 · 1993

Oxford

  • Earns a master’s in philosophy.
  • Concludes that academic philosophy would reach too few people, and drops Plan A.

Version 3 · 1994

Apple, then Fujitsu

  • Works in user experience on eWorld, an early online service.
  • Tests product management on the side by writing product documents for a director to review.
  • The director later gives him a reference for a product-management job at Fujitsu.

Version 4 · 1997

SocialNet

  • Co-founds one of the first online networks and puts his savings into it.
  • Plan Z is his father’s guest bedroom.

Version 5 · 2000

PayPal

  • Plan B, built through walks with Thiel and Max Levchin and a seat on PayPal’s board.
  • Shuts SocialNet, returns investors’ capital and joins PayPal as chief operating officer.

Version 6 · 2003

LinkedIn

  • Co-founds LinkedIn in December 2002 with two former SocialNet colleagues. It goes live on 5 May 2003.
  • A year later he sends Facebook’s founders to Thiel.
Figure 2. Hoffman’s career as a startup’s version history. Press a version to read its release notes. Dates and events are from his Academy of Achievement profile, Wikipedia and his own account in “Planning to Adapt”. The version numbers and drawings are mine.

Each version kept something from the one before. At Apple he tested product management while still in his old job, and two product jobs came before he co-founded SocialNet in 1997. When SocialNet stalled, what he had learned from it was what Thiel and Max Levchin wanted on the board of the company that became PayPal. Each move in Hoffman’s career was an experiment run from the one before, with a fallback in place before he needed it.

Permanent beta has a failure mode of its own. A startup that rebuilds its product after every complaint never ships anything. Hoffman stayed with each plan long enough for it to produce evidence, three years in SocialNet’s case.

Three gears have to turn together

In a podcast episode on the book, Casnocha and Hoffman describe “the three gears to competitive advantage.” Casnocha’s warning about them is short. “Any one of them is insufficient.”

The first gear is your assets. Hard assets are things like “the money in your bank account.” Soft assets are, in Casnocha’s words, “your skills, your network, your personal brand,” and Hoffman adds that “soft assets tend to be the things that will compound better.” The second gear is your aspirations and values, the work you want to be doing in ten or twenty years. The third is market realities, meaning where the demand is and who else is competing to meet it.

Casnocha’s example of a missing gear is nursing. California has “a huge demand for nurses,” so the market gear turns, and a person who hates nursing will probably not last in the job. The reverse failure is a passion that nobody will pay for. Sara Blakely is one of their examples of the gears meshing. She spent years selling fax machines door to door, and the selling became an asset when she started Spanx.

The three gears of a competitive advantage Three meshed gears labelled assets, what you have; aspirations, what you want; and market, what people need. A wire runs from the last gear to a light bulb. When all three gears are present they turn and the bulb is lit. When any one gear is missing it is drawn as a dashed outline, the other gears stop and the bulb goes dark, labelled stalled. ASSETSwhat you have ASPIRATIONSwhat you want MARKETwhat people need Turning

All three gears turn. In Casnocha’s telling, Sara Blakely’s years of selling fax machines door to door became an asset when she started Spanx.

Figure 3. Press a button to take out a gear. The three gears are Hoffman and Casnocha’s metaphor, and the examples are theirs, from “Develop Your Competitive Edge”. The machine and the light bulb are decoration.

A hypothetical biology student who has also learned to program holds two assets that are common on their own. Together they can let the student automate a lab’s image analysis. A lab that needs that done may choose this student over a stronger biologist or a stronger programmer. The test is whether the student solves one group’s problem better than the other options that group has.

Job leads travel through acquaintances

Hoffman’s introduction worked because he and Thiel were close friends. Most career information moves through looser links. In 1973, Mark Granovetter, now a sociologist at Stanford, published The strength of weak ties. For his doctoral research he had surveyed recent white-collar job changers in a Boston suburb. He asked those who found the job through a contact how often they had seen that person around the time the tip came.

Of 54 people who found a job through a contact, how many had seen that contact at least twice a week?

How often 54 job changers had seen the contact who told them about their job Fifty-four small figures, one per person, sorted into three groups. Nine, in blue, had seen their contact often, at least twice a week. Thirty, in gold, had seen the contact occasionally, less than twice a week but more than once a year. Fifteen, in grey, had seen the contact rarely, once a year or less. 54 PEOPLE WHO FOUND A JOB THROUGH A CONTACT 9 Often at least twice a week 30 Occasionally under twice a week, over once a year 15 Rarely once a year or less

Only 9 of the 54 had seen their contact often. The other 45 heard about the job from someone they saw occasionally or rarely.

Figure 4. Make a guess to sort the 54. Counts are from Granovetter (1973), who reported them as 16.7%, 55.6% and 27.8% of 54 recent white-collar job changers in a Boston suburb. One figure stands for one person.

Only 9 of the 54 had seen their contact often, meaning at least twice a week. Thirty had seen the contact occasionally, and 15 once a year or less. The contact was often “an old college friend or a former workmate or employer,” someone kept up with only sporadically. “Chance meetings or mutual friends operated to reactivate such ties,” Granovetter wrote. His explanation was that close friends move in the same circles you do, so they tend to hear about the same openings. An acquaintance moves in a different circle and hears about different ones. He called these looser connections weak ties.

In 2022, an experiment on more than 20 million LinkedIn members tested the idea at scale, as I described in How to get luckier. Ties with a moderate number of mutual connections led to the most new jobs, and MIT News put the turning point at about 10 mutual connections.

What the weak-ties evidence covers

Granovetter’s figures come from 54 people in one Boston suburb around 1970, and his “often” group is only nine people. The LinkedIn study is far larger and randomized, and it covers one platform, where a new connection costs a click. Both describe the routes that job information takes. Neither shows that collecting acquaintances in bulk improves a career.

Plan A, Plan B and Plan Z

Hoffman and Casnocha’s planning method has three parts. In a podcast episode called “Planning to Adapt,” Casnocha defines Plan A as “what you’re doing right now.” Plan B is where you might pivot if something changes, ideally close enough to Plan A that your skills carry over. Plan Z is “the plan that you do if everything else falls apart.”

In 1997, Hoffman’s Plan A was, in his words, “Make SocialNet successful.” He was about to put his savings into the company. Before he did, he asked his father whether he could move into the guest bedroom if SocialNet failed and left him broke. He promised to get a job and pay rent until he could move out. His father said yes. “I think it’s good that you’re taking risk in order to make something happen,” Hoffman recalls him saying. “I like seeing you, so sure.”

His Plan B came from conversations. He “started going on walks with Peter Thiel and Max Levchin,” he says, “and kind of telling them what I’d learned.” They asked him to join the board of the company they were putting together, which became PayPal. By 2000, SocialNet was clearly not going to become what Hoffman wanted. He shut it down, returned his investors’ original capital and joined PayPal full-time as chief operating officer.

Hoffman's Plans A, B and Z around 1997 to 2000, drawn as a circus act Two trapezes hang from the ceiling. The left one is Plan A, SocialNet in 1997. The right one is Plan B, PayPal, where Hoffman sat on the board before becoming chief operating officer. A dotted swing path joins them. Below both is a safety net with a small bed on it, labelled Plan Z, his father's guest bedroom. A cartoon acrobat hangs from the Plan B trapeze, which is what happened. Buttons move the acrobat to Plan A or down into the net. PLAN A SocialNet, 1997 PLAN B PayPal, board then COO PLAN Zhis father’s guest bedroom

By 2000, SocialNet was not going to become what Hoffman wanted. He swung to PayPal, where he already sat on the board, and became its chief operating officer.

Figure 5. Press the buttons. The plans are Hoffman’s own, as he describes them in “Planning to Adapt”. The circus is a drawing, and only the middle button shows what happened.
A Plan Z depends on who you are

Hoffman’s Plan Z was a parent with a spare room who was willing to offer it. Casnocha’s examples of a Plan Z are a spare room or savings to draw on. Many students have neither, and some are the person their family relies on. A thinner safety net makes the same bet more dangerous, so the size of a sensible risk differs from one person to the next.

Risks you can recover from

Putting your savings into a startup sounds reckless. The guest bedroom changed what was at stake. If SocialNet failed, Hoffman would lose money and keep a place to live while he started again. He would also have learned how an online network grows, and he co-founded LinkedIn in December 2002 with two former SocialNet colleagues.

Casnocha described the test behind a bet like this in a 2013 interview. It starts from the worst outcome. “If I feel like the worst-case scenario is my company goes under, I go bankrupt and my reputation is destroyed,” he said, then he is wary of the risk. Hoffman and Casnocha call bets made this way intelligent risks. Figure 6 has five risks to sort.

Spending a semester in a lab outside your major.

Recoverable. The worst case is a semester of electives that leads nowhere, and you still find out whether you like lab work. A hypothetical case.

PRD

Writing product documents nobody asked for and asking a director to review them.

Recoverable. Hoffman at Apple. The worst case was a polite no, and he came away with a reference for his next job.

Putting your savings into a startup, with a parent’s spare room as the fallback.

Recoverable. Hoffman in 1997. The guest bedroom capped the worst case. Without one, the same bet moves toward the other answer.

LOAN $

Borrowing more than you could repay to fund an idea no customer has tried.

Hard to reverse. If the idea fails, the debt stays, and so does a credit record that follows you for years. A hypothetical case.

?

Listing a skill you do not have on a CV to get an interview.

Hard to reverse. If it comes out, the damage is to trust. Casnocha’s own worst case included a destroyed reputation. A hypothetical case.

Figure 6. Sort each card. Cards two and three are Hoffman’s own bets, and the other three are hypothetical. The verdicts apply Casnocha’s worst-case test from his 2013 interview.

The risks that are hard to reverse share one feature. Their worst case damages something slow to rebuild, such as a credit record or other people’s trust.

The next job starts inside the current one

At Apple, Hoffman worked in user experience, and he suspected product management would suit him better. He wrote sample product requirement documents on his own and took them to the director of product management, James Isaacs. As Hoffman tells it in “Planning to Adapt”, he asked Isaacs, “will you assemble a group and review and give me feedback on them?”

Isaacs liked the work. He became Hoffman’s reference for a product-management job at Fujitsu, the last job Hoffman held before he started SocialNet in 1997.

Further reading

careers startups networks risk