Part IThe Story
Thirty-Five Cents
In September 2002, shares of a small Nasdaq software company called Opsware changed hands for about thirty-five cents. At that price the whole business was worth roughly $28 million, which was some $30 million less than the cash sitting in its bank account. The market was making a precise statement: it believed the people running Opsware would burn through the money before they built anything worth owning.
The chief executive was Ben Horowitz, then thirty-six. Eighteen months earlier his company, then called Loudcloud, had gone public at $6 a share in what BusinessWeek called "the IPO from hell." Three months earlier he had sold the only business Loudcloud actually had, the one that produced all of its revenue, and kept a piece of internal software that nobody outside the building had ever bought. Nasdaq had written to warn that the stock would be delisted if it stayed below a dollar. His largest customer, which supplied nearly all of the remaining revenue, was unhappy enough to threaten cancellation.
Five years later Hewlett-Packard bought Opsware for about $1.6 billion in cash. Two years after that, Horowitz and his longtime partner Marc Andreessen opened a venture capital firm with $300 million and an unusual premise: that founders should be helped to become chief executives rather than replaced by them. By early 2026 Andreessen Horowitz, known as a16z, reported roughly $90 billion under management and had raised another $15 billion in a single January.
Horowitz is best known for writing about the stretch in between. His blog posts and his 2014 book, The Hard Thing About Hard Things, described layoffs, near-bankruptcies, and the private misery of running a failing company with a frankness that management literature had mostly avoided. This profile follows that career in order: the Berkeley childhood, the Netscape years, the Loudcloud near-death, the Opsware recovery, and the construction of one of the largest venture firms in the world.
By the Numbers
From Loudcloud to a16z
$6.00Loudcloud's IPO price per share, March 2001
$0.35Opsware's share price after the 2002 sale to EDS
$63.5MPrice EDS paid for Loudcloud's hosting business, 2002
$14.25Per-share price HP paid for Opsware in 2007, about $1.6B in total
$300MSize of Andreessen Horowitz's first fund, 2009
$15BNew capital a16z announced in January 2026
A Berkeley Kid Between Two Worlds
Benjamin Abraham Horowitz was born in London in 1966 and grew up in Berkeley, California. His mother was Elissa Krauthamer. His father, David Horowitz, was one of the more combustible figures in American political writing: a co-editor of the radical magazine Ramparts in the 1960s who broke with the New Left in the 1970s and spent the rest of his life as a conservative activist and author. David's own parents had been members of the Communist Party. Horowitz opens The Hard Thing About Hard Things with a family story that captures the lineage: at a barbecue, his father mentioned that he had once handed out communist literature in the Queensbridge housing projects of New York, at the age of eleven.
Berkeley in the late 1970s and early 1980s did not expect a strong math student to play football. Horowitz joined the Berkeley High School team anyway, with no youth-league experience. He later described the first team meeting, at which head coach Chico Mendoza listed the ways a player could fail to take the game seriously and closed each one with the same instruction to hand in his equipment and leave. Horowitz called it his first lesson in leadership, and said he liked it at once.
He was, he wrote, the only player on the team who was also on the top academic track in mathematics, which meant he spent his days moving between groups that saw the world very differently. A new Run-D.M.C. album shook the locker room and went unnoticed in calculus; the Strategic Defense Initiative outraged young scientists and meant nothing at practice. Horowitz credited that double vision with a habit he leaned on later as a chief executive: when the facts seemed to dictate a particular ending, he went looking for another plausible reading of the same facts, because an alternative story was often enough to keep a frightened workforce from giving up.
He went east for college, earning a bachelor's degree in computer science from Columbia University in 1988. In the summer of 1986 he met Felicia Wiley on a double date arranged through a high school teammate; they married in 1988. He then took a master's in computer science at UCLA, finishing in 1990.
Silicon Graphics, NetLabs, and a Lesson About Priorities
During a summer in college Horowitz worked as an engineer at Silicon Graphics, the company whose workstations produced the computer graphics in films such as Terminator 2. He loved it and returned full-time after graduate school in 1990. About a year later a former SGI marketing head recruited him to a startup called NetLabs. He came to regard the move as a mistake. The company had been handed by its venture investors to a "professional management team" that, in his account, understood little about the product and kept changing direction. It was the first time, he wrote, that he understood why it mattered for founders to run their own companies. The observation would become the organizing idea of his venture firm nearly two decades later.
Home life was hard at the same time. The Horowitzes had three young children, and their second daughter, Mariah, had been diagnosed with autism. In the book Horowitz recounts a sweltering afternoon when his father, visiting a house that had no air conditioning, asked him whether he knew what was cheap. Flowers, his father said. And what was expensive? Divorce. Horowitz wrote that the joke made him see he had been trying to do everything at once and was about to fail at the thing that mattered most. He quit NetLabs the next day and took a job at Lotus Development, where he held senior product marketing roles.
It was at Lotus that a coworker showed him Mosaic, the graphical web browser built by students at the University of Illinois, including Marc Andreessen. Mosaic's team had since left to found Netscape Communications with Jim Clark, the founder of Silicon Graphics. Horowitz applied.
Netscape and the Lead Bullets
Horowitz joined Netscape in July 1995, a month before its famous IPO, and was put in charge of the Enterprise Web Server line: a standard server that listed for $1,200 and a secure version, using Netscape's new SSL protocol, for $5,000. When he arrived, two engineers worked on the product.
Then Microsoft shipped a web server of its own that matched Netscape's features, ran about five times faster, and was free. Horowitz's instinct was to escape the fight. He and executive Mike Homer drafted a plan of partnerships and acquisitions to wrap the server in enough extra functionality to sell. His engineering counterpart, Bill Turpin, a veteran of battling Microsoft at Borland, told him that no deal would fix a product that was five times slower; the team would have to use a lot of lead bullets instead of looking for a silver one. Netscape put its engineers on performance. By Horowitz's account the server eventually outran Microsoft's and grew into a $400 million business. He turned the phrase into a rule and used it again a decade later.
As director of product management he grew frustrated with how little value most of his product managers were adding. Having absorbed Andy Grove's High Output Management, he concluded that he had never told them what the job was, and in 1996 wrote a short internal document called "Good Product Manager/Bad Product Manager." Its best-known line is that a good product manager is the CEO of the product. He later said the team's performance improved almost immediately, and that the experience convinced him to invest heavily in training at every company he ran.
From 1997 to 1998 Horowitz was vice president of Netscape's Directory and Security product line. The chief executive, Jim Barksdale, liked to say that the company took care of the people, the products, and the profits, in that order. Horowitz borrowed the line for a chapter title. After AOL agreed to buy Netscape in late 1998, he became vice president and general manager of AOL's e-commerce platform division, where he oversaw the Shop@AOL service.
Loudcloud
In September 1999, the day after Andreessen stepped down as AOL's chief technology officer, the two men founded Loudcloud with two other Netscape and AOL veterans: Tim Howes, a co-inventor of the LDAP directory protocol, and In Sik Rhee, who had designed the Kiva application server. Horowitz was chief executive; Andreessen was chairman. The idea was to run the entire infrastructure behind other companies' websites, the servers, storage, networks, and software, as a service. In later language it was an early cloud computing company, years before Amazon Web Services.
The founders put in about $8 million, mostly Andreessen's, and Benchmark Capital followed with $15 million. Horowitz later wrote that at the first meeting with Benchmark's full partnership, senior partner David Beirne asked him, in front of his cofounders, when he was going to get a real CEO. He could hardly breathe, he wrote, partly because he suspected Beirne was right: he had never designed a large organization and did not know the executives and customers a professional CEO would bring. Whether he could learn the job before the board lost patience tortured him for months.
Loudcloud grew fast. By Horowitz's account it went from zero to about six hundred employees in less than eighteen months, signing customers that ranged from venture-backed startups to large companies. Then the dot-com market collapsed, and the startups that made up much of its customer base began to die. Private investors stopped funding infrastructure companies. Horowitz concluded that the only way to raise enough money to survive was to go public.
That required a board fit for a public company, and Horowitz did not want a stranger on it. The person he trusted most, the former Intuit chief executive Bill Campbell, had told him at the start of their relationship that he no longer joined boards. Horowitz called him anyway, to explain the situation. In Horowitz's telling, Campbell said he did not go on boards but could hear that Horowitz needed him, and joined. Campbell became the adviser Horowitz called in every later crisis.
The offering was a humiliation that saved the company. Loudcloud had to reverse-split its stock to get the price high enough to list, which infuriated employees who had been doing arithmetic with their share counts. Horowitz wrote that his mother-in-law, attending the all-company meeting, asked his wife why everybody hated Ben so much. On March 9, 2001, Loudcloud priced 25 million shares at $6, below its original target range, and raised $150 million. For the fiscal year that had just ended it reported revenue of $15.5 million and an operating loss of $164.8 million.
— Ben Horowitz, The Struggle, 2012The Struggle is not failure, but it causes failure. Especially if you are weak. Always if you are weak.
Spiral
The IPO bought time, not safety. When Loudcloud lowered its revenue forecast, the stock fell from $6 to about $2. In the third quarter of 2001 the company was assembling a strong quarter when the September 11 attacks threw its largest pending deal, with the British government, into doubt; a sales director persuaded officials to restore the money and Loudcloud made its numbers. Two weeks later its biggest competitor, Exodus Communications, filed for bankruptcy. Horowitz noted that Exodus had been valued at around $50 billion a little more than a year earlier and had raised $800 million on a "fully funded plan" only nine months before.
He began looking for a way out. A proposed merger with Data Return, a similar hosting company, died when his head of business development pointed out that combining two troubled businesses would only double the trouble. Looking at Data Return's numbers, Horowitz wrote, made it plain that Loudcloud's own story would probably not end well either.
His reasoning ran as follows. Loudcloud's infrastructure costs were so high that even firing every employee would not save it without a much faster sales ramp, and the market was not going to supply one. When he walked Campbell through the logic, Campbell answered with a single word, "Spiral," and Horowitz understood that his adviser agreed. What Loudcloud did have was the software it had written to automate its own data centers, which it called Opsware. The only survivable path was to sell the hosting business and become a software company whose first customer would be the buyer.
On June 17, 2002, Electronic Data Systems announced that it would acquire Loudcloud's managed hosting business for $63.5 million in cash and license Opsware under a three-year agreement with at least $52 million in minimum payments. About 140 employees and all fifty hosting clients moved to EDS. Loudcloud renamed itself Opsware Inc. During the negotiations Campbell had also advised Horowitz to quietly prepare the company for bankruptcy with his general counsel, and Horowitz read in his tone that the fallback might become the plan.
When the deal was signed, Campbell gave the advice Horowitz later called the foundation for rebuilding the company. Horowitz had planned to fly to New York for the announcement. Campbell told him to send Andreessen instead, stay home, and make sure every employee knew within minutes whether they were working for Opsware, working for EDS, or looking for a job. The people who stayed, Campbell's reasoning went, would judge the new company by how it treated the people who left.
Opsware at the Bottom
The market hated the transaction. Opsware's stock lost about 80 percent of its value and traded at thirty-five cents, a market capitalization near $28 million against roughly $60 million in cash. The new company had about eighty people. Horowitz gathered them off-site, told them he understood if they thought the plan was crazy, issued everyone new stock grants, and asked anyone who intended to quit to do so that day so the company would know who it could count on. Two people left. Of the rest, he wrote, all but two were still there when the company was sold five years later.
The Nasdaq letter came next. The board debated a reverse split or a buyback. Horowitz argued that the company simply needed to explain itself: a strong team, about $60 million in the bank, a $20 million-a-year contract with EDS, and real intellectual property. Unless he was the worst chief executive of all time, he told investors, that had to be worth more than $30 million. The stock climbed back above a dollar.
The EDS contract then nearly killed the company it had just saved. The software had been built to run Loudcloud's own data centers, with parts of the code tied to specific machines in the building, and deploying it inside EDS went badly. EDS, which accounted for about 90 percent of Opsware's revenue, gave the company sixty days to fix the problems or lose the contract. Opsware's team learned that the executive who controlled the relationship badly wanted to keep an asset-tracking product made by a small company called Tangram, which he was under pressure to abandon. Over the objections of his own staff, who thought the technology hard to integrate, Horowitz bought Tangram and bundled it for EDS. The contract held.
He also had to build a real enterprise sales force, and he hired Mark Cranney to lead it. Andreessen objected that Cranney did not look or sound like a head of sales and had gone to a weak school. Horowitz agreed with each point and said that if Cranney did not have those flaws, he would not be joining a company trading at thirty-five cents; he would be running IBM. Andreessen agreed to the hire.
Years later, when the competitor BladeLogic began consistently beating Opsware in large deals, Horowitz's senior people proposed every alternative to a head-on fight: a lightweight product, an acquisition, a shift to service providers. He told each of them that there were no silver bullets for this, only lead bullets. After roughly nine months of hard product work Opsware regained the lead. Horowitz guessed that without that effort the company would have been worth about a tenth of its eventual price.
Knowing the Price
By 2007 Opsware had grown to more than $100 million in annual revenue and several hundred enterprise customers. Horowitz had a standing rule that as long as the company was number one in a big market, it would not sell. Virtualization made him doubt whether the rule still applied: it multiplied the number of servers customers had to manage, which helped Opsware, but it also threatened to change the architecture underneath the product. After weeks of arguing with himself he asked his direct reports what they thought. All but one favored a sale.
He and John O'Farrell, his head of business development, decided the right price was $14 a share, about $1.6 billion. The board thought the figure was unrealistically high. Horowitz told every potential acquirer that Opsware would entertain nothing below $14, and for more than a month nobody called. Then BMC Software offered $13.25. Horowitz held firm, and two days later BMC came back at $14. When he told the other suitors he intended to accept, HP offered $13.50 to test whether he was bluffing. He replied that as a public company chief executive he could not take a lower bid. HP raised its offer to $14.25.
On July 23, 2007, HP announced it would acquire Opsware through a cash tender offer at $14.25 a share, an enterprise value of about $1.6 billion. The deal closed that September, and Horowitz became vice president and general manager of HP's Business Technology Optimization software unit. He wrote afterwards that he was physically sick for a period after selling what had taken eight years to build, and then concluded it was the smartest decision of his career.
Some Experience Required
The next idea arrived as an instant message. Horowitz wrote to Andreessen that they ought to start a venture capital firm whose motto for general partners would be "some experience required," meaning experience founding and running companies. Andreessen replied that he had been thinking the same thing. The two had been active angel investors, and Horowitz kept returning to Beirne's question from 1999. Why did founders have to prove they could run the companies they had created, instead of being helped to do it?
They studied the industry and found that returns were concentrated in a handful of firms, because the best entrepreneurs chose firms with the best track records. A new firm needed to change the basis of competition. Their thesis was that technical founders were the best people to run technology companies, citing Intel, Hewlett-Packard, Amazon, Apple, Google, and Facebook, but that those founders lacked two things professional executives had: the skills of running a large organization, and a network of executives, customers, partners, and reporters. The firm would try to supply the network directly. For the model they looked to Hollywood's Creative Artists Agency, the talent agency built by Michael Ovitz, whose staff worked on behalf of all of its clients rather than for individual agents.
Andreessen Horowitz launched in the summer of 2009 with a $300 million first fund. Within months it put about $50 million, a sixth of the fund, into the Silver Lake-led group that bought a majority of Skype from eBay at a valuation of $2.75 billion. Critics called the deal reckless, citing unresolved intellectual property litigation and competition from Google and Apple. In May 2011, eighteen months later, Microsoft agreed to buy Skype for $8.5 billion.
The firm raised a $650 million second fund in 2010 and grew quickly from there. In 2012 it invested $100 million in GitHub, its largest check to that point; Microsoft bought the company for $7.5 billion in 2018. It was an early backer of Instagram, led Coinbase's 2013 round, and invested in Airbnb and Facebook. Horowitz personally backed Lyft over Uber, later telling CNBC he preferred the character and culture of the Lyft founders, and sat on its board until 2020. He also served on the boards of Okta and of Nicira, the networking company VMware bought for about $1.26 billion in 2012.
Writing Down the Hard Things
The firm's marketing strategy was, to an unusual degree, its partners' writing, and Horowitz's blog became the best-known part of it. His April 2011 post "Peacetime CEO/Wartime CEO" used Larry Page's return as Google's chief executive to argue that companies facing an existential threat need a different kind of leader: directive, intolerant of drift, willing to break the rules that serve a company in good times. Horowitz wrote that most management books describe peacetime techniques, and that his own greatest management discovery during the Loudcloud-to-Opsware transition was how differently the two conditions had to be run.
— Ben Horowitz, Peacetime CEO/Wartime CEO, 2011Peacetime CEO knows that proper protocol leads to winning. Wartime CEO violates protocol in order to win.
In June 2012 he published "The Struggle," a short, incantatory list of what it feels like when a company is failing and its founder cannot quit. Few pieces of startup writing have circulated as widely. Its closing lines say that every great entrepreneur, from Steve Jobs to Mark Zuckerberg, went through it, that it does not guarantee survival, and that it is where greatness comes from.
In March 2014 HarperBusiness published The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers. The book combined the Loudcloud and Opsware history with practical chapters on layoffs, demoting friends, one-on-one meetings, hiring executives, and selling a company. Each chapter opened with rap lyrics from artists such as Nas and DMX, a reflection of Horowitz's long devotion to hip-hop that some readers loved and others found jarring. It became a New York Times bestseller. Horowitz pledged his portion of the proceeds to the American Jewish World Service to support women's rights in developing countries.
His second book, What You Do Is Who You Are: How to Create Your Business Culture, followed in October 2019. It built its argument around four unlikely case studies: Toussaint Louverture's leadership of the Haitian Revolution, the samurai code of bushido, the empire of Genghis Khan, and Shaka Senghor, who ran a prison gang and later became an author and criminal justice reform advocate. The thesis was that a culture is the sum of what people actually do, not the values printed on a wall. Some reviewers praised its concreteness; a Publishers Weekly review called the historical parallels strained and tone-deaf.
In August 2018 Horowitz and his chief of staff, Chris Lyons, announced the firm's Cultural Leadership Fund. Its limited partners were exclusively Black cultural leaders, among them Nas, Shonda Rhimes, Quincy Jones, and Kevin Durant. It would invest alongside a16z and donate its fees and carried interest to nonprofits helping Black Americans enter the technology industry. Lyons later said the fund had raised more than $60 million across its first three vehicles.
— Ben Horowitz, The Hard Thing About Hard Things, 2014Every time you make the hard, correct decision you become a bit more courageous and every time you make the easy, wrong decision you become a bit more cowardly.
Little Tech and Big Funds
As the firm grew, it moved into politics and policy. In July 2024 Horowitz and Andreessen published "The Little Tech Agenda," arguing that federal policy on crypto, artificial intelligence, and taxation favored large incumbents over startups. On their podcast they announced that each would donate to Donald Trump's campaign. "For little tech, we think Donald Trump is actually the right choice," Horowitz said, as reported by Axios. In October 2024 he told a16z employees that he and Felicia, citing a long friendship with Kamala Harris, would make a significant donation to groups supporting her campaign, while the firm's position stayed tied to its policy agenda.
His father died in April 2025 at eighty-six, after a long illness. Horowitz posted a tribute and shared a family obituary.
The firm kept scaling. In January 2026 a16z announced $15 billion in new capital across several funds, at a time when it reported about $90 billion in assets under management; in August it enlarged its growth fund to $8.5 billion and launched a $1.1 billion fund for AI hardware. What had started as a two-founder firm with a thesis about helping technical founders had become one of the largest private investment managers in technology, with dedicated funds for crypto, biotechnology, games, infrastructure, and defense-oriented "American Dynamism" companies.
Horowitz's reputation, however, still rests mostly on the account he gave of 2001 and 2002. The HP sale and the a16z fund sizes are the scoreboard. What entrepreneurs cite is the description of what it felt like to be the thirty-six-year-old chief executive of a thirty-five-cent stock, and the list of specific things he did next.
Part IIThe Playbook
Horowitz's principles come from a narrow and unusual stretch of experience: running a public company that nearly died twice, then building an investment firm around the lessons. They are less about strategy in the abstract than about what a chief executive does on the specific day when the numbers are bad, the team is scared, and there is no obviously right move. Each is drawn from his own accounts, checked against the record.
The principles
- 1.Name the war you are in.
- 2.Reach for lead bullets, not silver ones.
- 3.Tell the people who stay the truth within minutes.
- 4.Put people ahead of products and profits.
- 5.Train before you judge.
- 6.Sell the business that works against you.
- 7.Explain the company when the market misreads it.
- 8.Hire for strength rather than for lack of weakness.
- 9.Decide your price before anyone makes an offer.
- 10.Build the institution you wish had backed you.
- 11.Define culture as behavior, not slogans.
Principle 1
Name the war you are in.
The peacetime and wartime distinction is Horowitz's best-known idea, and its practical value lies in the diagnosis rather than the labels. In peacetime, a company with a big lead in a growing market should broaden its bets, encourage creativity, and follow protocol. In wartime, facing an existential threat, it has one shot and needs strict alignment behind a single objective. The damage comes when leaders apply the wrong mode: running a consensus process while the company bleeds, or barking orders at a team that is winning.
Horowitz's own switch came during the Loudcloud-to-Opsware transition, when he found that the collaborative habits he had absorbed from management books were making things worse. He also warned that wartime is a condition, not a personality, and that a leader stuck in permanent emergency mode will exhaust an organization that no longer needs it.
Tactic: At each quarterly review, write one sentence stating whether the company is in peacetime or wartime and why, and check whether your meeting cadence, decision rights, and tolerance for side projects match that sentence.
Principle 2
Reach for lead bullets, not silver ones.
Twice in his career Horowitz faced a competitor whose product was simply better, first Microsoft's free web server at Netscape and later BladeLogic at Opsware. Both times the organization's first instinct was to find a clever way around the fight: a partnership, an acquisition, a new market segment, a lighter product. Bill Turpin's line about lead bullets gave Horowitz a way to refuse those escapes.
The logic is diagnostic. If customers are buying in your category but choosing someone else, you have a product problem, not a market problem, and strategic maneuvers will not fix it. At Netscape the fix was raw performance work; at Opsware it was about nine months of hard engineering. Pivoting is sometimes right, but only when the market itself is the problem.
Tactic: Before approving any strategic workaround to a competitive loss, ask the team to state in writing whether customers are rejecting the category or rejecting your product. If it is your product, fund the engineering fix first.
Principle 3
Tell the people who stay the truth within minutes.
Bill Campbell's advice on the day of the EDS deal is the clearest example of a principle Horowitz returned to often: the message in a layoff or restructuring is really aimed at the people who remain. They will watch how their departing colleagues are treated and decide from that whether to trust the company again. Delay turns uncertainty into rumor, and rumor into resignation.
Horowitz's own procedures reflect this. Before a layoff, the chief executive addresses the whole company to explain why; managers, not human resources, tell their own people; and it happens fast, because a company that knows cuts are coming stops working. He wrote that Opsware survived multiple consecutive layoffs that should have broken its culture because it treated the people who left with respect.
Tactic: When a restructuring is decided, set the internal announcement for the same day the news could leak, and prepare manager scripts in advance so every affected employee hears it from their own manager first.
Principle 4
Put people ahead of products and profits.
Horowitz took the ordering from Jim Barksdale and argued that it was harder than it sounded. Taking care of the people, in his usage, means making the company a good place to work: where important work gets noticed, where the best workers are not passed over by the best politicians, and where process has not choked out the joy. He argued that this was the most difficult of the three and the one that made the others possible.
He credited the practices in that chapter, rather than any strategic insight, with getting Loudcloud through the dot-com crash, the EDS transition, and the Nasdaq warning. A company in trouble keeps its talent only if people believe the place is worth fighting for. The claim is conditional: a good workplace does not guarantee survival, but a bad one makes survival through a crisis very unlikely.
Tactic: Add a standing item to your executive meeting on who did important work that went unrecognized this month, and make sure someone senior acknowledges it before the next meeting.
Principle 5
Train before you judge.
"Good Product Manager/Bad Product Manager" began as a frustrated manager's attempt to write down what he expected. Horowitz later said the effect was immediate: product managers he had thought hopeless became effective once the job was defined. He drew a general lesson from it and put heavy investment into training at Loudcloud, which he credited with much of the company's eventual success.
His argument, in the a16z post "Why Startups Should Train Their People," is that most performance problems are really unclear expectations, and that startups skip training because they feel too busy. He called being too busy to train the moral equivalent of being too hungry to eat. The cost of writing a basic training document is a few hours; the cost of a team that does not know what good looks like compounds every week.
Tactic: For every role you manage, write a one-page document contrasting what a strong performer and a weak performer in that role actually do, and review it with each new hire in their first week.
Principle 6
Sell the business that works against you.
Loudcloud's hosting business produced all of its revenue and employed nearly all of its staff, and it was still going to kill the company. Its fixed infrastructure costs meant that no plausible amount of cutting would save it. Horowitz's decision to sell that business to EDS and keep the internal automation software was a bet that the less visible asset was the real one.
The move required separating what the company did from what it owned. Loudcloud's identity was hosting; its most defensible asset was Opsware, the software built to make hosting cheaper. By selling the business with the unfixable cost structure to a buyer who could afford it, and turning that buyer into the first software customer, he converted a liability into both cash and revenue.
Tactic: List every asset your company owns separately from the businesses it runs, and ask which of them a larger company would pay for and license back to you.
Principle 7
Explain the company when the market misreads it.
When Opsware's market value fell below its cash, the board considered financial engineering: a reverse split or a buyback. Horowitz chose to tell the story instead, laying out the team, the cash, the EDS contract, and the intellectual property, and inviting investors to do the arithmetic. The stock recovered above a dollar without a reverse split.
The principle rests on his high-school observation that the same facts support more than one narrative. A market that has written a company off is running one story; a chief executive who can credibly present another, grounded in verifiable numbers, can change the outcome. The same skill kept employees from walking out during the worst months. The narrative only works when the numbers underneath it are true.
Tactic: When your valuation or reputation diverges sharply from the facts, write a one-page explanation built only on figures an outsider can verify, and deliver it in person to the ten investors or customers whose opinion moves the rest.
Principle 8
Hire for strength rather than for lack of weakness.
The Mark Cranney hire shows how Horowitz evaluated executives. Andreessen's objections were accurate, and Horowitz conceded every one. His point was that a candidate without those flaws would not have been available to a thirty-five-cent company, and that Cranney's genius at sales was the thing Opsware needed most. Hiring processes that screen for the absence of weaknesses tend to produce well-rounded mediocrity.
The discipline is to decide first which strength the role requires right now and to weight it heavily, then to take on the job of managing the fit. Horowitz wrote that he worked with Cranney on cultural fit for years, and that the team eventually agreed he was the best person for the job even if not everyone became comfortable with him.
Tactic: Before interviewing for a senior role, write down the one or two strengths that would make the hire world-class, and require every interviewer to score those first and separately from any concerns.
Principle 9
Decide your price before anyone makes an offer.
Horowitz and O'Farrell set $14 a share as Opsware's price before any serious bid arrived, and he told every suitor the number in advance. When BMC offered $13.25 he held; when HP tested him at $13.50 he held again. The final price of $14.25 came from a buyer who was convinced he would walk.
The underlying decision came first: whether to sell at all. Horowitz tested it by asking his direct reports whether they had the energy for another major product overhaul, and by examining how virtualization would change the market. Only once he knew he was willing to sell did the number carry weight. A price named without a real willingness to walk away is a bluff.
Tactic: Before opening any sale, financing, or partnership negotiation, write down your walk-away number and the reasons behind it, get your board to endorse it, and state it to counterparties early.
Principle 10
Build the institution you wish had backed you.
Andreessen Horowitz was designed around a specific wound: a venture partner asking a founding CEO when he would get a real one. Horowitz and Andreessen built a firm to do the opposite, supplying founders with the executive network and operational help they lacked so that they could stay in charge. They borrowed the service model of a Hollywood talent agency and, early on, required general partners to have founded or run companies.
The advantage was positioning as much as service. In an industry where the top firms were entrenched by track record, a newcomer could compete only by changing what entrepreneurs valued. A clear, public thesis about founders gave a16z a way to be different, and the partners' writing spread it for free.
Tactic: Write down the three worst experiences you had as a customer of your own industry, and test whether a company built to eliminate them would attract customers the incumbents cannot.
Principle 11
Define culture as behavior, not slogans.
What You Do Is Who You Are argues that culture is what people do when no one is instructing them, and that leaders shape it through specific, memorable actions and rules rather than value statements. Horowitz's case studies, from Toussaint Louverture's rules for his officers to the samurai code, are about behaviors made concrete enough to enforce.
The practical implication is that a leader's own conduct under pressure is the strongest cultural signal available. The Opsware off-site, where Horowitz asked anyone planning to quit to do it that day and handed out fresh stock grants, set a norm of candor more effectively than any poster could. Fostering culture in this view means designing a few behaviors so distinctive that people ask why they exist.
Tactic: Pick one behavior you want everyone in the company to practice, turn it into a rule specific enough that violations are obvious, and enforce it publicly the first time it is broken, including by a senior person.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIMaxims
- The founder's disadvantage is teachable. Skills and networks can be supplied; the founder's knowledge of the product cannot be hired in.
- Cash in the bank is an argument. When a company trades below its cash, the chief executive's job is to make the market do the arithmetic.
- A reverse split is a confession. Horowitz resisted financial engineering because it signaled weakness to exactly the people whose confidence he needed.
- Your biggest customer can be your biggest risk. EDS saved Opsware and then nearly ended it; concentration cuts both ways.
- Most crises are survivable by people who stay in the room. Horowitz's recurring theme is that there is always a move for a company still alive to make it.
- Frankness scales. A blog post about misery did more to build a venture brand than any list of returns.
- Borrow models from other industries. A talent agency, a football coach, and a Haitian general all supplied frameworks for running technology companies.
- Advisers earn trust in the worst week. Campbell mattered because he was the call Horowitz could make when his life's work was on the line.
- Ownership of a decision includes its aftermath. Selling Opsware was right, and Horowitz still wrote plainly about how sick it made him.
In Their Own Words
The story of any company is the story of its people. And the story of any CEO is the story of how they handle the struggle.— Ben Horowitz
In peacetime, leaders must maximize and broaden the current opportunity. As a result, peacetime leaders employ techniques to encourage broad-based creativity and contribution across a diverse set of possible objectives. In wartime, by contrast, the company typically has a single bullet in the chamber and must, at all costs, hit the target. The company's survival in wartime depends upon strict adherence and alignment to the mission.— Ben Horowitz
The hard thing isn't setting up an organizational chart. The hard thing is getting people to communicate within the organization that you just designed.— Ben Horowitz
The only thing that prepares you to run a company is running a company.— Ben Horowitz
Take care of the people, the products, and the profits—in that order.— Ben Horowitz
The right thing to do is to recruit the best people you can and build the best products you can. If you do that, the profits will follow.— Ben Horowitz
When you are building a company, you must believe there is an answer and you cannot pay attention to your odds of finding it. You just have to find it. It matters not whether your chances are nine in ten or one in a thousand; your task is the same.— Ben Horowitz
The story of any company is the story of its people. If you get the people right, everything else follows.— Ben Horowitz
In my experience as CEO, I found that the most important decisions tested my courage far more than my intelligence.— Ben Horowitz
The primary thing that any technology startup must do is build a product that's at least 10 times better at doing something than the current prevailing way of doing that thing.— Ben Horowitz
Sometimes an organization doesn't need a solution; it just needs clarity.— Ben Horowitz
The key to high-quality decision making is not knowledge. It is whether our knowledge is true or false, and the best way to make it more true is to turn it into a prediction and test it.— Ben Horowitz
As a startup CEO, I slept like a baby. I woke up every 2 hours and cried.— Ben Horowitz
The trouble with innovation is that truly innovative ideas often look like bad ideas at the time.— Ben Horowitz
The venture capital business is a 100% game of outliers—it's extreme exceptions. If you're not finding a way to do something that nobody else can do, you're not going to make money in this business.— Ben Horowitz
The most important thing I learned as an entrepreneur was that the technology industry rewards people who can see the future and build it.— Ben Horowitz
In the technology business, you have to be willing to be wrong. The key is to be wrong quickly and cheaply.— Ben Horowitz
The best entrepreneurs are not the best visionaries. The best entrepreneurs are incredible salespeople. They know how to tell an amazing story that will convince talent and investors to join in on the journey.— Ben Horowitz
Culture is not what you say, it's what you do. And more specifically, it's what you do when nobody's watching.— Ben Horowitz
The culture of a company is not what the CEO says it is. It's what the CEO does when nobody's looking.— Ben Horowitz
Your company's culture is its behavior under stress. When the pressure is on, what does your organization do? That's your culture.— Ben Horowitz
The right culture with the wrong people is still the wrong culture.— Ben Horowitz
The most important skill that a CEO can have is the ability to learn. Everything else can be learned, but if you can't learn, you can't adapt, and if you can't adapt, you can't survive.— Ben Horowitz
Experience is what you got when you didn't get what you wanted.— Ben Horowitz
The difference between successful people and really successful people is that really successful people say no to almost everything.— Ben Horowitz
Embrace the struggle. It's the only way to get better.— Ben Horowitz
