Part IThe Story
You Folks Had All of the Ideas
The Venetian in Las Vegas is a building designed to make you forget where you are: canals under a painted sky, gondoliers working a chlorinated lagoon three stories above the Strip. In the spring of 2018 it was full of people who maintain the plumbing of the modern corporation. They were help-desk managers, change-approval administrators, and the configuration-management people who know which server talks to which database and why it stopped at 2:14 a.m. There were 18,000 of them, gathered for Knowledge, the annual conference of a company called ServiceNow. Somewhere in the crowd, working the scrum for selfies like a lounge act between sets, was a 63-year-old programmer from New Castle, Indiana.
Forbes, which had sent two reporters, found him "sparkly-eyed." He had booked extra time between appearances, the magazine noted, purely to grip and grin. Neither Wayne Newton nor Celine Dion, it observed with a straight face, had anything on Fred Luddy in Sin City that day, at least not among this crowd. What Luddy said about the experience was strange, and worth pausing on, because almost nobody in his position says it and fewer mean it.
"When all of these people are happy to see you, honestly you feel like a rock star," he said. "It's kind of an undeserved feeling, because they were the inspiration. You folks had all of the ideas. I just wrote them down and thought about them."
I just wrote them down. Founders say all sorts of humble things on conference floors; humility is cheap on a stage and expensive in a boardroom. But Luddy's version has a specific texture. He does not claim to have been lucky or surrounded by geniuses. He claims to have been a stenographer, the man at the back of the room with a pencil, taking dictation from the people who actually do the work. It is the self-description of a clerk. He was by then a billionaire, and the company he had started in his house fourteen years earlier carried a $30 billion market capitalization, more than 6,000 employees, and 4,000 customers, 850 of them in the Forbes Global 2000. Forbes had just named it the most innovative company in America.
Fourteen years before the Venetian, he had been broke. Thirteen years before, he had been the company's only employee. Six years before, even after the vindication of an IPO, the whole enterprise was worth about $2 billion, and it came very close to being sold for a good deal less than it would be worth twenty-four months later. The man in the selfie line had lost everything once, had nearly given the company away once, and had once fired himself from the top job. That he was standing in a fake Venice being adored by help-desk managers was not inevitable. Several times it was barely likely.
Everything that follows circles that sentence. He just wrote them down.
By the Numbers
Fred Luddy and the Company He Wrote Down
$35MPersonal fortune lost in the collapse of Peregrine Systems
1ServiceNow's headcount until mid-2005
$880KRevenue in fiscal 2005, the first full year of sales
$2.5BVMware's 2012 acquisition offer, blocked by Sequoia
$13.28BServiceNow revenue in 2025
29,187ServiceNow employees in 2025
$60MGift to Indiana University, the second-largest in its history
Thirty-Five Million to Zero
The number was $35 million. That was what Fred Luddy's position in Peregrine Systems had been worth, and it was not theoretical wealth. He had earned it across more than a decade as the San Diego software company's chief technology officer, the person who built the product the company sold. Peregrine made IT service management software. These are the systems a large corporation uses to log a broken laptop, route the ticket, track the fix, and document why the email server is down. It was unglamorous, essential, and lucrative. Luddy's products there and at earlier employers would eventually generate revenue, by one accounting, well in excess of a billion dollars.
Then Peregrine's executives were found to have been inventing a meaningful share of that revenue. Restated financials, an SEC investigation, criminal cases, and in 2002 a bankruptcy filing. The company that had been one of San Diego's software anchors became one of its most notorious accounting frauds. Luddy was not implicated. He had built the thing. He had not counted it. But a fortune held in a fraudulent company's stock does not distinguish between the people who cooked the books and the people who wrote the code. It went to zero for all of them.
He was nearly fifty. He had no job, the savings were gone, and the industry he had spent his career in now carried his former employer's stink. What happened next is the shortest interval in the story, and the one on which everything depends.
In November 2003 (he has been precise about the month, if not always about the day) he began writing a new piece of software. The entity that would hold it was called Glidesoft. It was, in Wikipedia's dry formulation, intended "to provide the same services previously available from the then defunct Peregrine Systems." He was going back into the same burning building.
The timing was psychological rather than strategic. Luddy started the company, by his own telling, two weeks before his fiftieth birthday, because he believed some threshold was about to close. "I couldn't wait," he said later, "because there was something psychologically that said I couldn't start a company at 50." The public record cannot quite agree on what year he was born, and so cannot quite agree on which side of that line he landed. Wikipedia currently gives November 24, 1954; an older revision gave 1953; the company was registered on November 5, 2003, and incorporated in California in 2004. I mention the discrepancy because the man himself cared enough about the deadline to sprint for it, and because the sprint matters more than whether he technically made it. He beat a clock that existed mostly in his head.
Years later, accepting the Horatio Alger Award in 2023, an honor reserved for people who have overcome adversity, he described the wipeout in terms that would sound like a greeting card if he did not have the receipts.
— Fred Luddy, on receiving the Horatio Alger Award, 2023I believe that adversity is what makes you who you are. Losing my job and my life's earnings was one of the best things to ever happen to me. That situation forced me to figure out exactly what I wanted.
What he wanted, it turned out, was to write code. Not to run a company, though he would run one for seven years. Not to be rich, though he would be rich again, more than thirty times over. He wanted to sit with a laptop and solve a problem that annoyed him. The money had always been incidental to that, which is perhaps why losing it clarified so much. Remove the $35 million and what remains is a man who would rather be programming than doing almost anything else.
He said as much, with a kind of wonder, in 2013: "when people are able to find a passion, a deep passion, that they are reasonably good at and they can make money, it's almost unfair."
Almost unfair. It is the remark of someone who never fully believed the world would pay him for this.
The Boy Who Fetched the Balls
New Castle, Indiana, sits on the Big Blue River about forty-five miles east of Indianapolis. It is a manufacturing town, and its high school, where Fred Luddy was a member of the class of 1972, carried the name of the Chrysler Corporation. His father was an accountant. His mother, Mariol, taught at a Catholic school. The family would go on to send something like a dozen of its members through Indiana University: both parents, a sister, two brothers, and assorted others. Education was the family business, more or less.
As a boy, Luddy worked as a ballboy, retrieving tennis balls for players whose names now read like a hall-of-fame plaque: Arthur Ashe, Stan Smith, Rod Laver. It is a job of pure service. You crouch at the net, you sprint when the point ends, you hand the ball back to the person whose name is on the scoreboard, and you vanish before the next serve. Nobody watches the ballboy unless he screws up. Nobody remembers him if he doesn't. It is tempting, and probably too neat, to see in this the template for a career spent building the invisible machinery that lets other people do their jobs. I'll note the rhyme and move on, the way a good ballboy would.
He wanted to write code from the age of seventeen. At eighteen, according to a biography he has used on the speaking circuit, he wrote his first software program: an automated supply chain and maintenance utility for a business unit of American Standard, the plumbing-fixtures company. Think about that for a second. A teenager in Indiana wrote enterprise software, the boring kind that tracks parts and schedules repairs. That was the same species of problem, maintenance and the orderly fixing of broken things, that he would spend the rest of his life on.
He arrived at Indiana University in Bloomington in 1973 and did not stay. The standard version is that he dropped out because he was spending too much time programming. The version he prefers, in that same speaking biography, is more interesting. He "tried to attend Indiana University and Sir George Williams University," the latter a Montreal school that later became part of Concordia, "but found part-time programming for these institutions to be far more compelling than the classroom experience." He did not drop out of college to code. He dropped out of college to code for college. The universities kept him; they just kept him on the payroll rather than the roster.
The person he credits for this is Edward Whalen, a professor of economics at IU who also served as the university's director of budgeting. Whalen was a man who lived in two worlds at once, the seminar room and the ledger, and he offered the dropout his first job in programming sometime in the 1970s. Luddy has named him in nearly every account of his life that Indiana University has published. Whalen is a figure of a recognizable type: the mid-level institutional man who sees something in a kid that the institution's formal machinery cannot process, and quietly routes around it. The ticket gets escalated. The kid gets a desk.
From there the résumé moves through the mainframe era like a geological core sample. Amdahl Corporation, the company Gene Amdahl founded to build IBM-compatible mainframes. Boole and Babbage, a performance-software house named for two nineteenth-century mathematicians. A company of his own, Enterprise Software Associates. Along the way he wrote, by his own account, a competitive operating system to IBM's mainframe offering, a set of performance improvements for the IBM mainframe environment, a PC video-conferencing and collaboration system, and a Unix-based application development environment. These are not small projects. Writing an operating system to compete with IBM's is the kind of thing people put on their tombstones. Luddy lists it in a subordinate clause.
In 1990 he joined Peregrine. In 1993, on a project at BellSouth, he saw what was then one of the largest IT help desks in the country. "When I went to Bell South in 1993," he recalled in 2011, "they had one of the largest IT help desks with 40 people." Forty people answering phones about broken computers was, in 1993, enormous. By 2011 his company would be selling six or seven thousand seats at a time to Citi and Johnson & Johnson.
The arc of his career, in other words, ran exactly parallel to the rise of a function that barely existed when he started: the corporate department responsible for keeping everyone else's technology working. It is a department that is noticed only when it fails, a ballboy department. Luddy spent thirty years writing tools for it, and by 2003 he had a fairly specific grievance about how those tools treated the people who used them.
Smart People, Dumb Software
"I love and hate technology," Luddy told an interviewer at the 2013 Knowledge conference, and then explained the conditions under which each emotion applies. He loves it when it is enabling and helpful. He hates it, specifically and personally, when it makes him, or someone else, feel stupid. "That is why I focus on simplicity."
Seven years later he gave a long interview to Forbes that ran under a headline in his own words: "I Don't Like How Technology Makes Smart People Feel Dumb." In it, according to the magazine's description, he told the story of a data-entry clerk named Phyllis who inspired his best idea. I have not been able to reconstruct the full anecdote from the record available to me, and I'm not going to embroider it. The point is that the founder of a company now worth well over $150 billion locates the origin of his best idea in a data-entry clerk. Not a CIO, not a customer advisory board, not a McKinsey slide. Phyllis.
The grievance had a precise shape. In the early 2000s, consumer software was undergoing a quiet revolution. Luddy liked to cite My Yahoo, the customizable portal page, which some 80,000 people configured to their own taste with absolutely no training. Nobody sent them to a class. Nobody certified them. They just did it, because the software was built to be done. Enterprise software, by contrast, required training courses measured in days. It forced people, as he put it in 2014, "into a form metaphor": a world of rigid fields and submit buttons, designed by the vendor's engineers for the vendor's convenience. The consumer internet was getting fun. The office was still boring.
When Sequoia Capital later explained why it had invested, it pinned the decision on exactly this asymmetry. Luddy, the firm wrote, "wasn't motivated by the big market for software or because he wanted to make a bunch of money."
— Sequoia Capital, on its first meeting with Fred LuddyHe did it because he found it completely ridiculous that the software used at home was so useful and fun, yet the software used at work was so clumsy and frustrating. It speaks to how clearly Fred thinks.
Completely ridiculous is the right register. This is not the language of market sizing. It is the language of offense. Luddy looked at the gap between My Yahoo and the help-desk software he himself had built at Peregrine, and he was annoyed in the way only a craftsman can be annoyed by his own trade's inferior work.
His vision, which he dates with a programmer's terseness, "2003, November," was "to be able to enable business people to create meaningful workflow applications." Not programmers. Business people. The clerks and coordinators and help-desk managers who know exactly what the process is because they run it every day, and who are told, every time they want to change it, to file a request with IT and wait. He wanted to hand them the pencil.
The legend of how he started is almost too good. According to Sequoia, Luddy wrote the initial code for ServiceNow (a working version of the product, the firm emphasizes) on a single flight from San Francisco to London. Sequoia's partners were frankly impressed: "This wasn't a consumer app but fairly complex enterprise technology." The more prosaic version, which Luddy has given himself, is that it was just him, his laptop, and a lot of coffee. Both are probably true. The flight is the moment of crystallization; the coffee is the months on either side of it.
He named the company Glidesoft. The name lasted until 2006, when the company became Service-now.com, the hyphen lingering for a few more years before it dropped away. The original word never left. Today, deep in the platform, ServiceNow's scripts are written in JavaScript, and its database queries run through a proprietary object-oriented interface called Glide. Every customer who has ever written a query against a ServiceNow instance has typed the old name of the company, usually without knowing it. It is a fitting fossil: the brand was replaced, but the verb describing what software ought to feel like remained, load-bearing, inside the machine.
A Platform That Could Do Anything, Which Is to Say Nothing
The first problem with Luddy's software was that it worked too well in the abstract.
He had built something general, what the industry would come to call a platform-as-a-service, before most of the industry had a name for it. It was a set of modular components, database tables that could be wired together, and workflows anyone could reconfigure. He showed it to people. They nodded politely. The trouble, as he recalled in 2013, was that the pitch was "it can do anything." In enterprise sales, anything is the most dangerous word in the language. A buyer cannot get budget approval for anything. A buyer gets budget approval for a specific line item that replaces a specific thing that is failing.
So he narrowed it. He and the small group around him built a set of applications on top of the platform that would create immediate business value. They started where Luddy knew the ground better than almost anyone alive: IT service management, the help desk. The general platform became, at least on the sales sheet, a specific replacement for the clunky ticketing systems that large IT departments already hated. In 2011, a reporter asked the obvious question, which was whether this was just Peregrine again. Luddy's answer is the most combative thing on the record from a famously mild man.
"It is the next generation of what Peregrine was trying to do," he said. "But we threw out every single rule that Peregrine played by, and came up with a whole new notion and a whole new way of doing things. We're stunningly different."
The rules he threw out are visible in the business model. Peregrine sold software the old way, installed on the customer's servers, licensed in large upfront chunks, upgraded in painful cycles. ServiceNow would be delivered over the internet as a subscription, priced per user per month, with that cost, as one early account put it, "ranging down from US$100." Early on, the product was offered free in exchange for feedback. Luddy needed the truth more than he needed revenue. Most decisively, according to the account Sequoia later assembled, the company sold seat-based licenses with no seat limits, a strategy the insiders called "going wide." Once a customer had the platform, anyone in the organization could use it. The pricing was designed not to extract maximum value from the first sale but to let the thing spread.
And spread it did, through a channel nobody had designed. When a department head in HR or facilities wanted software for some new workflow, they did what department heads do: they went to IT. IT, which already had ServiceNow running the help desk and knew it could be bent to almost any shape, said, essentially, use this. The IT manager, the most put-upon person in the corporate org chart, became the product's evangelist. The platform that "could do anything" got to do anything after all. It just had to enter through the service entrance.
The most extravagant example came from CERN, the European Organization for Nuclear Research, outside Geneva. It is the home of the Large Hadron Collider, and also, it turns out, of a great deal of garbage collection, fire safety, and lawn care. CERN's administrators took what Luddy's team liked to describe as the "IT" out of "ITSM" and began running service management for waste, fire services, people, transport, and green space through ServiceNow. The physicists searching for the fundamental particles of matter were filing their landscaping requests through software designed, at its root, so that a help-desk manager in San Diego wouldn't feel dumb.
For a long time, there was barely anyone to do the selling. Luddy was the company's only employee until mid-2005. That is worth sitting with: roughly a year and a half after he began, with beta versions already shipping to customers in 2004, the headcount was one. Then JMI Equity, a growth investor, put in $2.5 million, and Glidesoft hired five people. Among the co-founders the record lists are David Loo, Don Goodliffe, Bow Ruggeri, and Patrick Casey, engineers of the type who sign on with a fifty-year-old ex-CTO because they've seen his code. Mark O'Donnell, one of the original employees, would later summarize the mood of those years in six words: "Nothing's impossible, we can build anything." The first office, as Luddy later showed a company video crew, was a total fixer-upper.
The numbers began to compound with the indifference of arithmetic. Fiscal 2005, the first full year of sales: $880,000. Every year after that, revenue more than doubled. In 2007 it reached $13 million, and the company went cash-flow positive and opened its first Silicon Valley office, in San Jose. By January 2011 it had 275 employees in San Diego, Chicago, New York, Atlanta, London, and Frankfurt, plus a partnership with Accenture, which had more than 100 consultants devoted to ServiceNow implementations. For the fiscal year ending June 30, 2011, revenue hit $125 million, with 400 employees and 800 customers: Qualcomm, UBS, Deutsche Bank, Morgan Stanley, Flextronics, Home Depot, Coca-Cola, McDonald's, Intel, SanDisk. For much of this period, the company was based in Solana Beach, a surf town north of San Diego, far from the region's established tech corridors. ServiceNow, a local paper noted in 2011, "isn't well-known beyond San Diego's small circle of software companies." It was, in a phrase Luddy used himself, "kind of the brains behind IT operations," and like most brains it was invisible from the outside.
In 2009, Sequoia invested. The partners were Doug Leone and Pat Grady. Grady would later make a career-long point of Luddy's motivation, telling an interviewer that founders who focus on solving a customer problem, he named Fred Luddy and Zoom's Eric Yuan, tend to outperform those driven by money or status. It was a bet on temperament as much as market. It was also, in retrospect, a bet that would require Sequoia to save the company from its own exhausted leadership, including Luddy, twice.
The Founder Who Fired Himself
By 2010 the doubling had become a problem. The company had been built lean, bootstrapped in spirit even after it took money, run by a programmer who preferred programming. Now customers were arriving faster than the infrastructure could hold them. Sequoia's later retelling uses the phrase "buried under its own success." Running lean, the firm concluded with hindsight, is good. Running too lean is a dangerous liability.
Luddy, by the account he gave on Sequoia's Crucible Moments podcast in 2024, had what he described as a crisis of confidence about his own ability to scale the business. He enjoyed product. He did not enjoy the operational grind of running a company with hundreds of employees across six offices and two continents. Sequoia, which had been watching the same thing, suggested a new chief executive. Luddy agreed, and in the spring of 2011 he stepped down to become chief product officer, a role, he told the San Diego Union-Tribune that September, that "better fits his programming roots."
It is easy to praise this decision in retrospect and hard to overstate how rarely founders actually make it. The literature of Silicon Valley is full of founders who were pushed. It is thin on founders who walked to the door and held it open. Luddy's version was not entirely voluntary, since boards and investors have their ways, but every account, including the most skeptical, describes him as a willing participant in his own demotion.
The man he held the door for was Frank Slootman. Born in the Netherlands, trained as an economist, he had most recently run Data Domain, a storage company he took public in 2007 and then sold to EMC. Slootman was in nearly every respect Luddy's opposite. Luddy was a craftsman; Slootman was a field general. Luddy listened and wrote things down; Slootman issued orders and expected them followed. Years later, after he had gone on to run Snowflake, Pat Grady would single out Slootman's defining habit as "dealing with reality" and his "intellectual honesty." It was a polite way of describing a man who did not soften bad news for anyone.
The stories from Slootman's arrival have the quality of parables. Early on, by the account given in Crucible Moments, he dismissed an employee for failing to remove unused sports tickets after being told to. It was a small, almost absurd infraction, and the point was precisely its smallness: Slootman's directives were orders, not suggestions, and the company needed to learn that before the larger orders arrived. Slootman himself has described being terrified to open his email every morning.
He had reason. The cloud infrastructure that hosted ServiceNow's customers was fragile in ways that made seasoned executives pale. Deutsche Bank, among the company's marquee accounts, suffered critical outages that were eventually traced, in one of the most quoted details of the whole saga, to a used Dell server in a closet. A global bank's IT operations depended on secondhand hardware in a broom cupboard. When Slootman arrived, Doug Leone later said, the company was "90 days from going out of business."
The fix was brutal and expensive. ServiceNow went on a hiring spree that roughly doubled the workforce in months, deliberately driving the company into short-term losses to buy the talent and technical capacity it needed. Annual reports from the period show headcount rising from 729 to 1,269 in a single year. An engineering team led by Dan McGee standardized the entire technology stack, homogenizing the infrastructure so that every customer instance ran on the same predictable foundation. Chronic stability problems receded. Gross margins improved. As Sequoia put it, there was no playbook for this; ServiceNow was a cloud company before the cloud had manuals. The team started down the wrong road more than once, ripped it up, replaced people, and iterated until they got it right.
None of this was Luddy's work, exactly. But none of it would have been possible without the specific architectural decision Luddy had made at the beginning: a single platform, a single data model, extensible in every direction. Slootman took that extensibility, in Sequoia's phrase, and "merchandised" it, turning what Luddy had built as a general toolkit into a widening catalog of formal product offerings. Luddy built the instrument. Slootman wrote the setlist and booked the arenas.
There is a version of this story in which the founder resents the operator who scaled his creation. There is no evidence of that here. Luddy sat on the board and kept writing code. And in 2011 he received the Ernst & Young Entrepreneur of the Year Award, an honor for a company he had just stopped running.
The $2.5 Billion Mercy
In 2012, with the infrastructure stabilizing and an IPO on the calendar, VMware came calling. The offer, unsolicited, was $2.5 billion.
Everyone in the room wanted to take it. This is the detail that makes the episode worth studying rather than merely celebrating. It was not only the employees with options who wanted the money, and not only the executives who had just spent a year in crisis mode. The founder wanted it too. Luddy, by the Sequoia account, personally favored the deal. He was exhausted. He had lost a fortune once, and $2.5 billion is a great deal of stability for a man who remembers what zero feels like. The company's leadership leaned the same way. Sequoia's own retelling puts it plainly: Fred, Frank, and the board "were all ready" to sell.
Two investors were not. Doug Leone and Pat Grady argued that the company was badly undervalued and that the right comparison was not the next two years of revenue multiples but the next decade of market expansion. They predicted, in venture shorthand, a "10-bagger," something like a $10 billion company. To make the argument stick, Grady brought in Bill Chandler, a lawyer who had written the relevant law on the subject. Chandler established that the board of a private company receiving an offer had a fiduciary duty to shop the deal for the maximum value, which made a quick sale to VMware legally untenable. The deal died on procedure, which is how many of the most consequential deals in business die.
ServiceNow went public in June 2012, raising $210 million. Morgan Stanley ran the offering one month after it had taken Facebook public, a debut that had gone famously sideways, and ServiceNow's went up. By the end of 2012 its market capitalization was $3.7 billion, already more than a billion dollars above the VMware offer. Within two years it reached $10 billion, just as Leone and Grady had predicted. Grady, looking back, called the sale that wasn't a "train wreck" narrowly avoided. Sequoia titled its podcast episode From Starting Over at 50 to Dodging a $150B Mistake.
It is an investor's story, told by investors, and it flatters investors. Take it with appropriate salt. But the core facts are not in dispute, and they cut against the founder-mythology instinct in an instructive way. The founder, the man whose clarity and motivation Sequoia had bet on, was wrong about the most important financial decision in the company's history. He was wrong for a perfectly human reason. He was tired, he had been broke, and the number was large. The people who saved him from his mistake were the ones he had invited in.
There is a symmetry here worth noting. In 2002, a company Luddy built was destroyed by executives who lied about the numbers to make them look bigger. In 2012, a company Luddy built was saved by investors who insisted the real numbers would get bigger, and who used the machinery of corporate law to force the board to slow down long enough to find out. He has told the first story as the best thing that ever happened to him. He has told the second, by all accounts, with gratitude and a certain bemusement.
Thirty Seconds of Contentment
After 2011, Luddy's title was chief product officer, and for the next several years his public appearances took on a distinctive form. At ServiceNow events in Europe in 2013, during a talk called "The Future of Now," he did what company accounts describe as his signature move. He built a bespoke application on the ServiceNow platform, live, on stage, in a matter of minutes. No slides could do what that did. The founder of the company would sit at a keyboard in front of an audience and prove, in real time, that the software did what he said it did, and that a business process could be turned into an application in less time than it takes to describe it.
The demonstration was the argument. ServiceNow's entire premise was that people who are not programmers should be able to build things. "We want people to do programming without realizing they are doing it," Luddy said in 2014. Watching the founder do it without apparent effort was a kind of permission.
He was, by his own description, never satisfied with any of it. At Knowledge14, asked whether he was paranoid about competitors, he gave the most revealing answer in his public record.
— Fred Luddy, Knowledge14, April 2014I'm always frustrated. When we complete something, I am pretty content with it for about 30 seconds. I've never been happy with anything I've produced for more than a few minutes. I see flaws, I see opportunities.
Thirty seconds. It is the attention span of a craftsman rather than an owner. Owners admire what they have. Craftsmen see the joint that didn't quite close.
The interviews from these years have a consistent theme, which is that the enemy is not a competitor but complacency. Luddy liked to cite BlackBerry, which had once held, in his telling, something like 85 percent saturation of its market and spent its energy making its existing product incrementally better. "They were looking at the current strengths," he said, "not what the market expectations were going to be." His diagnosis was general: "most organizations become very myopic on their own product." His benchmark was explicitly consumer-grade. "We would like to be thought of as more of an Apple and Google, compared to Oracle," he said in 2014, drawing a line between companies that grow by innovating and companies that grow by financial engineering and acquisition. A year earlier he had been blunter: "The way that large software companies try to solve stuff is to buy a lot and claim it is integrated."
He understood the force driving all of this better than most enterprise vendors, because it was the force that had produced his company in the first place. Employees were bringing consumer expectations to work. They had seen Dropbox and Evernote, and they started using them whether IT approved or not. Shadow IT, the industry called it. Luddy had a better phrase for the mindset. People have a job to do, he said, and "their attitude is, 'I will proceed till I'm apprehended.'" It is a line that could double as an autobiography. A teenager writes enterprise software for a plumbing company. A college student skips class to code for the college. A broke fifty-year-old rebuilds the product of the company that just bankrupted him. Proceed till apprehended.
At Knowledge15, he described what came next as a "platform popularity contest." ServiceNow had to become the place people went to get work done, instead of sticky notes or email. The only way to win was to make using the platform easier than not using it. And the only way to do that was to be willing to throw out old code and redesign from the ground up. "I feel ready for the next 10 years," he said. "We spent 10 years building something that I think has established a really great foundation. And now, people just want to work differently."
The company outgrew him, as companies built well do. Slootman ran it until 2017. John Donahoe followed, then in late 2019 Bill McDermott, the former chief executive of SAP. McDermott was an operator of almost theatrical intensity, the son of a working-class Long Island family, who had spent his career at precisely the kind of giant enterprise vendor Luddy had defined himself against. ServiceNow under McDermott and his CFO, later president, Gina Mastantuono, crossed $10 billion in revenue in 2024 and reported $13.28 billion in 2025, with 29,187 employees. It began making the large acquisitions Luddy had once said it would avoid. In March 2025 it bought Moveworks for nearly $3 billion, and later Armis for $7.75 billion. It described itself, as Slootman had begun to a decade earlier, as the "ERP for IT," and then as something larger, a system of record for the whole enterprise.
Luddy moved to an advisory role around 2016. He had been a director since June 2004 and chaired the board from 2018 to 2022. In 2018 he crossed into the billionaire rankings. Forbes photographed him at his beachfront home in Del Mar and noted that success had meant more time to spend with his ten-year-old son, and to work on his next project, making American healthcare more affordable and efficient. By October 2023 his net worth was estimated at $1.2 billion.
The company still runs its queries through Glide.
Coopetition
There is a strange, cheerful coda to the ballboy, and it happened in San Diego.
A few years into ServiceNow's public life, Luddy was invited to meet Billie Jean King, who was in town exploring whether to expand World TeamTennis, the league she had co-founded in the 1970s, to include a San Diego franchise. He went. He emerged, as one local magazine put it, a proud owner of the San Diego Aviators. He had a reasonable sense of what pro sports cost: his longtime friend John Moores was the former owner of the San Diego Padres. So he experienced what he described as reverse sticker shock.
"I was buying the whole team," Luddy said, laughing, "for probably less than the agent's commission on a third-string shortstop!"
He took majority ownership in 2015. In 2016, in only its third season, the team won the championship. In March 2017, he and Mark Ein, a Washington entrepreneur who owned the league's Washington Kastles, bought King's majority stake in the league itself. Ein is a man who had built his own fortune in venture and private equity and then spent it on an unfashionable sport for the love of it. The arrangement, with two team owners also co-owning the league they competed in, required what Luddy called "coopetition." "We have to cooperate with each other, we have to build the best product, but at the same time we have to compete at the team level."
What is revealing is what he loves about the format. World TeamTennis is fast. It uses a service clock to keep games moving, scores by simple points rather than tennis's arcane love-fifteen-thirty-forty, and plays with an on-court DJ. A full night delivers five matches in a couple of hours: men's and women's singles, men's and women's doubles, mixed doubles. "I really love it," he said, "because all the tennis decorum [is] kind of thrown out the window. Our format is just so straightforward."
Listen to what he wished people knew: that they "can come with almost no planning, they can plunk down less money than it costs to go to the movies, they can see world class tennis, and in two hours and 15 minutes they're going to see five different matches."
No planning. No training. No decorum. A scoring system that doesn't make smart people feel dumb. It is My Yahoo with rackets. The boy who once crouched by the net fetching balls for Arthur Ashe had grown up and bought the court, and the first thing he did was throw out the rules that made the game hard to follow. The service clock ran.
Software That Compounds
In January 2025, the Salk Institute in La Jolla appointed Luddy to its board of trustees. The institute was founded in 1960 by Jonas Salk, who, famously, declined to patent the polio vaccine. Its president, Gerald Joyce, praised Luddy's fit with Salk's ambitions in biocomputation. The announcement noted that Luddy was working on projects to "revolutionize spreadsheets," surely the most consequential and least beloved piece of software in corporate life, and to improve the diagnosis and treatment of rare genetic diseases in children. He was seventy, or close to it, depending on which birth year you believe. He was still programming.
By the summer of 2026 he had started another company. It is called StarLifter, and on June 30, 2026, he sat with CNBC's Jon Fortt for nearly an hour to explain it. StarLifter, he said, was building the "decision layer" missing from enterprise AI: software that reasons over a company's data in real time and recommends or triggers the actions that humans now hash out in slow meetings, ingesting a company's policies and handbooks to understand its business context. It was early, with three customer engagements, and well-capitalized enough that raising more wasn't a priority. Tola Capital, led by Sheila Gulati, had led the seed round.
He explained the thesis on LinkedIn in terms that, for anyone who has followed his career, land with a jolt of recognition.
"When I started ServiceNow, the problem was clear: work was fragmented across systems, and people spent too much time managing processes instead of getting work done," he wrote. "Today, the next bottleneck is decision-making." Organizations had spent heavily on ERP, analytics, workflow automation, and now AI, and yet "many important business decisions still require people to collect information from different systems, figure out what's changed, determine what's important, gather in a room to make a decision, and coordinate action across teams. In many ways, people have become the middleware between data and execution."
Human middleware. It is the same grievance he had in November 2003, transposed up an octave. Then, the insult was that software forced intelligent people into rigid forms and then made them feel stupid for not mastering them. Now the insult is that software leaves intelligent people to do the dull work of stitching meaning together across systems that should already understand each other. He gave examples with the specificity of someone who has watched it happen: a renewal at risk, with three departments each holding a different piece of why; a supplier disruption that the team who solved it eighteen months earlier has rotated away from, so the playbook starts over.
"The goal isn't to replace people," he wrote. "It's to eliminate the manual work required to understand what's happening."
In a separate post he made a claim that reads, from the founder of one of the most successful enterprise software companies in history, almost like a confession: "Most enterprise software depreciates from the day it's installed. The opportunity in front of us is to build software that compounds — software whose value grows the longer an organization uses it, because every decision, every override, every operational event becomes durable institutional memory the system can reason over."
Durable institutional memory. A system that remembers what the enterprise has learned. It is, if you squint, the job description of the person who writes everything down.
The other thing he has spent his second fortune on is a building in Bloomington.
In 2015 ground was broken on a new home for Indiana University's School of Informatics, Computing, and Engineering, which had been founded in 2000 as the first school of its kind in the country. Luddy gave $8 million toward the $39.8 million project. The building is 124,000 square feet, designed by Indianapolis-based RATIO Architects with Pelli Clarke Pelli, and finished in twenty-seven months from conception. It contains fifteen classrooms and labs, nineteen conference rooms, thirty-one interview rooms, an innovation center, and a public café. It opened in January 2018 and was named Luddy Hall, "in honor of the many IU alumni in the Luddy family, including Fred's mother, father, sister and two brothers."
At the dedication on April 13, 2018, the capstone of a weeklong celebration the university called LuddyFest, Luddy stood up and said: "The structure has exceeded all expectations. The drawings were magnificent, but to see it in person, I just might finish my undergraduate work."
He also said something less charming and more pointed. "The shine has come off Silicon Valley. It's too expensive. It's not a good place to live, and I think that we have such an opportunity here to become such a hub of innovation. In Indiana, it should be our turn now."
His mother, Mariol, then living in a retirement community in Carmel, was in the audience. "The dedication was totally awesome," she told a local reporter. "(Fred) did a wonderful job speaking. He's been extremely successful, extremely, and he's very generous." It is hard to improve on a mother's double extremely.
On October 17, 2019, a Thursday, IU President Michael McRobbie announced in Luddy Hall that Luddy had given $60 million more, the second-largest private gift in the university's two-hundred-year history. It would fund the Luddy Center for Artificial Intelligence with an initial focus on digital health; six endowed chairs, six endowed professorships, and six endowed faculty fellowships; and scholarships, including some reserved for high-achieving Hoosier students. The trustees renamed the school the Luddy School of Informatics, Computing, and Engineering, effective immediately. In January 2023 the name extended to the Indianapolis campus. In June 2025, at the school's twenty-fifth anniversary, McRobbie paid tribute to Luddy's generosity, welcomed Tom and Heather Luddy in the audience, and remembered "their vivacious late mother Mariol, whose beaming pride in the naming of the school will remain an indelible memory for me."
In March 2023, sixty students from the Luddy School's living-learning center flew to California, departing Bloomington at 3 a.m., and met the man whose name was on their building. A freshman computer-science major named Jennifer Johnson came away with a sentence she repeated to the university's news service. "Fred Luddy assured me that no matter the failure, I can soar again." He had talked, she said, "about knowing your strengths, and also the strengths of those around you." It is the speech of a man who once fired himself and once tried to sell too early, and who was rescued from both mistakes by the strengths of those around him.
But the image I keep returning to comes earlier, from the winter before Luddy Hall opened. On December 16, 2017, at 10 a.m., in Simon Skjodt Assembly Hall, the arena where Indiana basketball is played, before some 1,971 graduating students, Indiana University conferred on Fred Luddy an honorary Doctor of Humane Letters. Forty-four years after he arrived in Bloomington and left the classroom for a programming job in the budget office, he gave the winter commencement address.
Afterward he posted the video to LinkedIn. He called it the greatest honor of his life. He thanked the employees and customers of ServiceNow for enabling it. And then, in the tone of a man who has never entirely believed the room was cheering for him, he added a note for anyone who clicked through:
"If nothing else, you will love my hat — starts about 19 minutes in."
Part IIThe Playbook
What follows is an attempt to extract the operating logic from Luddy's career. Most of it is not the logic of a born executive. It is the logic of a craftsman who happened to build a company, made two catastrophic-seeming errors of judgment, and was saved from both by recognizing, or being made to recognize, the limits of his own temperament. The principles below are drawn from the evidence above. Some are about product. More of them, surprisingly, are about knowing when to stop being in charge.
The principles
- 1.Build for the person who feels stupid.
- 2.Re-enter the burning building with a new set of rules.
- 3.Architect general, sell specific.
- 4.Price for spread, not for extraction.
- 5.Stay a company of one until the product is undeniable.
- 6.Recruit through the service entrance.
- 7.Fire yourself before the outages do.
- 8.Hire your temperamental opposite, then obey him.
- 9.Distrust your own exhaustion at the moment of sale.
- 10.Treat contentment as a thirty-second state.
- 11.Make the demo the argument.
- 12.Follow the grievance to its next octave.
L
The Decisions That Made ServiceNow
Luddy's career, reduced to the moments where a different choice would have produced a different company.
1973
Arrives at Indiana University; soon finds programming for the university more compelling than class. Edward Whalen offers him his first programming job.
1990
Joins Peregrine Systems; rises to chief technology officer.
2002
Peregrine collapses into bankruptcy amid accounting fraud. Luddy's $35 million goes to zero.
2003
November: begins writing the platform that becomes Glidesoft, roughly two weeks before his fiftieth birthday.
2005
Sole employee until mid-year; JMI Equity invests $2.5 million. First full fiscal year revenue: $880,000.
2007
Revenue reaches $13 million; first year cash-flow positive.
2009
Sequoia Capital invests, led by Doug Leone and Pat Grady.
2011
Steps down as CEO; Frank Slootman takes over. Luddy becomes chief product officer.
2012
VMware's $2.5 billion offer is blocked; IPO raises $210 million in June.
2019
Gives $60 million to Indiana University; school renamed in his honor.
2026
Launches StarLifter to build the "decision layer" for enterprise AI.
Principle 1
Build for the person who feels stupid
The most durable product insight in Luddy's career is emotional, not technical. He hated technology that made smart people feel dumb, and he built a company around refusing to make that software. The reference point he chose was not an enterprise competitor. It was My Yahoo, a consumer product that 80,000 people customized without any training. The comparison was deliberately unfair to the enterprise vendors, and that unfairness was the point: once you hold work software to the standard of home software, the incumbents look ridiculous, and "completely ridiculous" was exactly the phrase Sequoia used for his motivation.
The useful move here is choosing whose experience to optimize. Enterprise software is bought by executives and used by clerks, and most vendors design for the buyer. Luddy traced his best idea to a data-entry clerk named Phyllis. When the end user is the design target, the product gets simpler. Simplicity turns out to be the thing that lets the product spread beyond the department that bought it.
Tactic: Identify the least powerful person who will use your product daily, watch them use the incumbent, and design against every moment where they look embarrassed.
Principle 2
Re-enter the burning building with a new set of rules
After Peregrine, Luddy could have fled IT service management entirely. He did the opposite and rebuilt the same category within roughly a year of watching it destroy his savings. The decision looks reckless and was actually the opposite. He knew the domain more deeply than almost anyone, he knew precisely what the incumbent products got wrong because he had built some of them, and he knew the buyers. What he changed was not the market but the rules: subscription instead of license, cloud instead of on-premises, configuration by business users instead of consultants. "We threw out every single rule that Peregrine played by," he said.
Sequoia's summary of the lesson is right: you don't have to be first. A category you know intimately, attacked with a different architecture and business model, beats a novel category you are guessing about. Domain expertise compounds; it is the founder's least appreciated asset after a failure.
Tactic: List every structural assumption your previous company or industry treated as fixed (pricing unit, delivery model, who configures it, who buys it), and build the new company by inverting at least three.
Principle 3
Architect general, sell specific
Luddy's platform could do anything, and that was precisely why it initially failed to sell. "It can do anything" gives a buyer nothing to approve. The fix was not to narrow the architecture but to narrow the pitch: build a specific application, IT service management, on top of the general platform, and sell that. The generality stayed underneath, waiting.
This two-layer structure is the core of ServiceNow's long-run economics. The specific wedge got the company through the door. The general platform let it expand, first to HR, then to facilities, and eventually to CERN's lawn care. Slootman later "merchandised" the extensibility into a widening product line, but he could only do so because Luddy had refused to build a single-purpose tool in the first place.
Tactic: Keep the underlying data model general enough to support uses you haven't imagined, but go to market with one named problem, one named buyer, and one named incumbent you replace.
Principle 4
Price for spread, not for extraction
ServiceNow's early licensing had no seat limits. Customers could let anyone in the organization use the platform, a strategy insiders called "going wide." This sacrificed revenue in the short term in exchange for something more valuable: organic proliferation inside the customer. Once the help desk was running on ServiceNow, every new workflow request that reached IT was a potential expansion, and IT had every reason to recommend the tool it already owned.
The broader lesson is that pricing is a distribution decision. A per-seat price with hard caps optimizes the first invoice. A pricing model that lets usage spread optimizes the tenth. Luddy also gave the product away early in exchange for feedback, which is pricing as research: the cheapest way to learn what customers actually need is to remove the friction that keeps them from telling you.
Tactic: Audit your pricing for any mechanism that punishes a customer for letting more of their colleagues use your product, and remove it if expansion revenue is your growth engine.
Principle 5
Stay a company of one until the product is undeniable
Luddy was ServiceNow's only employee for roughly eighteen months. Beta versions shipped to customers in 2004; the first five hires came only in mid-2005, after JMI Equity's $2.5 million. The working version of the product, by Sequoia's account, was coded on a single transatlantic flight. This was not penny-pinching. It was a refusal to build an organization around a product that did not yet prove itself.
A founder writing code alone has no coordination costs, no meetings, and no one to persuade. The tradeoff is speed of scaling later, and ServiceNow paid that price in 2010 when the infrastructure buckled. But the sequencing was right. The product was undeniable before the company was large, and the company never had to pretend.
Tactic: Delay your first hire until a customer is using the product for real work, and define in writing what that hire unblocks that you cannot.
Principle 6
Recruit through the service entrance
ServiceNow's most effective salespeople were not on its payroll. They were IT managers, the overworked, under-thanked people at the bottom of the corporate prestige ladder, who discovered they could solve other departments' workflow requests by pointing them to the platform they already ran. Sequoia called IT managers the company's "beachhead to evangelize the product for wider adoption."
The insight is to find the user whose job gets easier when your product spreads. That person becomes your channel. Luddy had spent thirty years building for IT departments and understood their incentives intimately: they get blamed when things break and ignored when things work. Give that person a tool that makes them the department that says yes, and they will sell it for you.
Tactic: Map who inside your customer receives requests from other teams, and make your product the fastest way for that person to say yes.
Principle 7
Fire yourself before the outages do
In 2011, Luddy stepped down as CEO of a company growing more than 100 percent a year. He was not failing at product; he was failing at scale, and he knew it. The infrastructure was hanging by a thread. Deutsche Bank's outages traced to a used Dell server in a closet, and Doug Leone believed the company was ninety days from collapse. A founder who had stayed in the chair for reasons of pride would likely have presided over the end.
The key detail is that Luddy did not leave. He moved to chief product officer, the role that "better fits his programming roots." Founders often frame stepping aside as a binary between running everything and running nothing. Luddy's version kept him working on the thing he was best at, while handing the thing he was worst at to someone better.
⇄
Founder and Operator
How the ServiceNow leadership handoff divided the work.
| Dimension | Fred Luddy | Frank Slootman |
|---|---|---|
| Core instinct | Build Write the code, listen to users | Execute Issue orders, enforce discipline |
| Relationship to infrastructure | Lean, scrappy, bootstrapped in spirit | Overhaul the stack, standardize everything |
| Hiring posture | Solo until mid-2005 | Double headcount in months, accept losses |
| Product contribution | General, extensible platform | Merchandise extensibility into new product lines |
| Risk he addressed | Product-market fit | Infrastructure collapse |
| Post-handoff role | Chief product officer, director, later chair | CEO through 2017, later CEO of Snowflake |
The founder-to-IPO literature, from Sequoia's podcast to books like Frederic Kerrest's Zero to IPO, keeps returning to some version of this handoff because it is so rarely executed cleanly. Luddy's case is unusual mainly in how little drama it generated.
Tactic: Write down the three operational problems that will kill your company in the next year, honestly rate your own competence at each, and if you score yourself lowest on the most urgent one, start the search.
Principle 8
Hire your temperamental opposite, then obey him
Slootman was everything Luddy was not: blunt, operational, unafraid of firing someone over unremoved sports tickets to establish that directives were orders. The pairing worked precisely because of the contrast. A founder who hires a successor in his own image gets more of what the company already has. A founder who hires his opposite gets what the company is missing.
The harder part is what comes after the hire. A founder who remains in the building, on the board, with the most institutional authority of anyone, can undermine a new CEO simply by being consulted. There is no evidence Luddy did. He stayed in his lane, built product, and let the operator operate. The handoff's success depended as much on the founder's restraint as on the CEO's talent.
Tactic: When you bring in an operator, publicly define the decisions that are now theirs alone, and refuse to be the court of appeal for anyone who disagrees with them.
Principle 9
Distrust your own exhaustion at the moment of sale
When VMware offered $2.5 billion in 2012, Luddy wanted to take it. So did the board and the CEO. The company had just survived a near-death infrastructure crisis, and a man who had once lost $35 million had reasons to value certainty that his investors did not share. Leone and Grady blocked the sale by insisting the company was undervalued, and by using a lawyer, Bill Chandler, to establish that the board had a fiduciary duty to shop the deal. ServiceNow was worth $3.7 billion by year-end and $10 billion within two years. It later passed $150 billion.
The lesson is not "never sell." It is that the moment an offer arrives is often the moment a founding team is most depleted, and depletion distorts judgment toward safety. Sequoia's framing is useful: evaluate the offer against a five- to ten-year view of the market, not a two-year forward multiple. The founder's emotional history, especially a history of loss, is a legitimate input into personal financial decisions and a dangerous input into enterprise valuation.
Tactic: When an acquisition offer arrives, write the company's ten-year bull case before reading the offer letter, and require anyone who supports selling to argue against that case in writing.
Principle 10
Treat contentment as a thirty-second state
"When we complete something, I am pretty content with it for about 30 seconds." Luddy's restlessness was not a personality quirk to be managed. It was the engine of the product. He cited BlackBerry's 85 percent market saturation as the cautionary tale: a company that kept improving its current strengths while the market's expectations moved elsewhere. His prescription was a willingness to throw out old code and rebuild from the ground up, even on a successful platform.
The operational version of this is to measure the product against users' expectations, which are set by the best software they use anywhere, rather than against its own previous version. Incremental improvement against yourself is how dominant products become myopic.
Tactic: Once a quarter, have your team rank your product against the best consumer software they personally use, not against your direct competitors, and fund the largest gap.
Principle 11
Make the demo the argument
Luddy's signature move at company events was to build a working application live on stage in a few minutes. For a product whose central promise was that non-programmers could build things, there was no more persuasive evidence. The demonstration didn't describe the value proposition; it performed it. "We want people to do programming without realizing they are doing it" is a claim. Watching the founder do it in real time is proof.
This works only when the product is genuinely that easy, which is the hidden discipline. A founder who commits to live demos commits the engineering organization to a product that can survive them. The stage becomes a quality-control mechanism.
Tactic: Commit to a live, unscripted product demonstration at your next major customer event, and let the preparation expose every place the product isn't ready.
Principle 12
Follow the grievance to its next octave
At roughly seventy, Luddy started StarLifter around the same complaint that produced ServiceNow, transposed upward. In 2003 the insult was that software forced smart people into rigid forms. In 2026 it is that software leaves smart people acting as "middleware between data and execution," manually stitching meaning across systems in slow meetings. The grievance is identical in kind: capable humans are doing work that technology should be doing for them.
Founders often look for a second act in a new domain. Luddy's pattern suggests a different approach. Find the original grievance, ask what it looks like after your first solution succeeded, and attack the new bottleneck your own success revealed. ServiceNow solved fragmented workflows; StarLifter targets the decisions those workflows now feed. He also named a standard that applies to any software builder: most enterprise software "depreciates from the day it's installed," and the goal is software that compounds.
◎
Conventional Wisdom vs. the Luddy Approach
Where his career departs from standard founder advice.
| Conventional wisdom | Luddy's approach | Phase |
|---|---|---|
| Start young; startups are a young person's game. | Start two weeks before fifty, with thirty years of domain knowledge. | Origin |
| After a failure, change industries. | Rebuild the same category with inverted rules. | Origin |
| Hire early to move fast. | Stay a company of one until customers are using it. | Early |
| Cap seats to maximize revenue per account. | No seat limits; let the product spread. | Early |
| Founders should stay CEO as long as possible. | Hand the CEO role to an operator at the first sign of scale failure. | Scale |
| Take the big offer when you're tired. | Let investors force a ten-year valuation lens. | Risk |
| Grow by acquisition and claim integration. | Grow by platform extensibility and product innovation. | Scale |
Tactic: Restate the problem your current company solves as a grievance about wasted human effort, then ask what wasted effort remains for the people who already use your solution.
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 IIIQuotes and Maxims
In their words
— Fred Luddy, on founding ServiceNow just before his fiftieth birthdayI couldn't wait, because there was something psychologically that said I couldn't start a company at 50.
— Fred Luddy, San Diego Union-Tribune, September 2011It is the next generation of what Peregrine was trying to do. But we threw out every single rule that Peregrine played by, and came up with a whole new notion and a whole new way of doing things. We're stunningly different.
— Fred Luddy, ServiceNow Knowledge13, May 2013When people are able to find a passion, a deep passion, that they are reasonably good at and they can make money, it's almost unfair.
— Fred Luddy, at the dedication of Luddy Hall, Indiana University, April 2018The shine has come off Silicon Valley. It's too expensive. It's not a good place to live, and I think that we have such an opportunity here to become such a hub of innovation. In Indiana, it should be our turn now.
— Fred Luddy, on LinkedIn, 2026Most enterprise software depreciates from the day it's installed. The opportunity in front of us is to build software that compounds — software whose value grows the longer an organization uses it, because every decision, every override, every operational event becomes durable institutional memory the system can reason over.
Maxims
- Write it down. The best product ideas belong to the people doing the work; the founder's job is to listen carefully and record them faithfully.
- Software should never make a smart person feel dumb. If your users need training to do what a consumer app teaches in seconds, the product has failed, not the user.
- Know the burning building. Deep domain expertise from a failed company is the most underpriced asset a founder can carry into a new one.
- Anything sells nothing. Build a platform that can do anything, then sell it as the answer to one specific problem.
- Price for the tenth sale. Licensing that lets usage spread inside a customer is worth more than licensing that maximizes the first invoice.
- One employee is a strategy. Stay alone until the product is undeniable; an organization built around an unproven product institutionalizes its uncertainty.
- Fire yourself in time. When the company's most urgent problem is the one you are worst at, hand it to someone else and go back to what you do best.
- Exhaustion is not a valuation. The moment you most want to sell is often the moment your judgment about the company's future is least reliable.
- Thirty seconds is enough contentment. See the flaws and the opportunities in everything you finish, because your users already do.
- Proceed till apprehended. People with work to do will route around rules that slow them down; build the tool they would choose anyway.
