Move over Bill Gates and Michael Dell, computerdom’s newest magnate is just 27. Joe Liemandt is out to change the way we buy and sell things.
By Josh McHugh
This story originally appeared in the 6-9-1996 issue of Forbes. This republished web version accompanies a 2018 profile of Joe Liemandt, who has returned to the ranks of the world’s richest people with a new tech empire. ___________________________________________________________
IN JANUARY 1990, 21-year-old Joseph Liemandt telephoned his parents from Palo Alto, Calif. Joe, an economics major, was dropping out of Stanford University just a few months before graduation. He wanted to start his own software company. “You’re a moron,” growled his father, Gregory Liemandt, a highly successful businessman. Gregory pleaded with his son to finish his degree, but Joe wouldn’t yield. He had a hot idea and the idea couldn’t wait until graduation.
“My market window’s closing!” he told his father, and the next day he threw himself completely into running the software business he had started six months earlier with a handful of schoolmates.
Six years have elapsed and Joe’s now 27. Dad died in the meantime, but not before becoming both proud and supportive of his rebellious son, whose Trilogy Development Group of Austin, Tex. has 200 employees, and is in the process of hiring another 100. Still private, Trilogy Development — FORBES estimates — had revenues of $ 70 million last year, and earned about $ 20 million before taxes. This year revenues will probably top $ 120 million. Based on Wall Street’s valuations of publicly owned software companies with similar profiles, Trilogy could be worth $ 1 billion, given or take a couple hundred million, if Liemandt wanted to cash out or raise capital — which he doesn’t.
What was the “market window” that led Joe Liemandt to drop out of college?
“I knew when I got to college I was going to start a software company,” Liemandt says. “I started doing tons of research. I’d sit in the library going through lists of the top 50 software companies.” He did part-time consulting jobs for Palo Alto computer companies during the academic year and the intervening summers. On these jobs he began to notice that when he ordered computer equipment, the shipments often arrived late, or with parts that were missing or incompatible with his computers. Smart technology companies were dumb when it came to selling and delivering their products. So he began to study what is called sales configuration — how orders are processed. He was surprised at how primitive the process was. Much of it was handwritten, using cumbersome inventory manuuals and often requiring lengthy consultations between a company’s salespeople and its engineers.
As he tells the story, Liemandt’s initially calm demeanor changes. He begins rapidly clicking the pen he is holding, then suddenly pushes his chair back from a conference table and springs to a whiteboard. “I focused my research,” he says. Grabbing a red marker, he begins to sketch a rectangular diagram representing the cost structure of the typical U.S. manufacturing company. About half of the average manufacturer’s noncapital spending budget, Liemandt says, is made up of general and administrative costs, R&D expense and manufacturing costs. All of those areas, he adds, have been automated and computerized for years.
Now the punch line: “The rest of their budgets — more than 40% — the manufacturers spend on sales and marketing,” he exclaims, circling half of the diagram. “I thought, ‘Holy cow, no one’s done this [computerizing the sales process]!’ And this,” he says, circling the “sales” part of his cost-structure rectangle with the red marker while still rapidly clicking the pen in his left hand, “is [potentially] a $ 10 billion market.”
By his junior year at Stanford, Liemandt had done enough research to know that others were sniffing around the same problem. In his programming classes he learned that Digital Equipment Corp. was using expert systems, an offshoot of artificial intelligence written in something called rule-based programming, in an effort to computerize its own sales configuration. Rule-based programming consists of a series of if/then constructions — if the Hewlett-Packard salesperson changes the printer in the package she’s selling, then the system needs a different kind of cable. If the package uses a different kind of cable, then the price of the entire system changes.
Liemandt could see that rule-based programming was a promising approach. But building a useful mathematical model of a complex product line required so many thousands of if/then equations that the rule-based programming solution, by itself, was unwieldy. In his classes Liemandt learned there was another approach to the configuration problem. A Xerox programmer, Sanjay Mittal, was trying to build inventory and product-line models using what’s called constraint-based equation systems. This kind of math is used in everything from optimizing airlines’ passenger yields to interpreting military satellite photographs. But as it turned out, constraint-based programming wasn’t capable of efficiently handling the large volumes of data required for a workable sales configuration model. The problem remained unsolved.
Joe Liemandt knew that others were working on the problem, and he began to worry that his window was closing. Hewlett-Packard and IBM were pursuing internal projects. These were aimed at creating programs powerful enough to catalog their huge inventories and smart enough to constantly reconfigure product offerings as product lines and prices changed. Joe decided to risk his dad’s ire, abandon his diploma and work full time on sales configuration software. Liemandt convinced four fellow Stanford students to throw in with him. John Lynch dropped out with Liemandt. Christina Jones, Chris Porch and Seth Stratton thought about it, but juggled schoolwork with Trilogy until they graduated. “Their parents hated me,” chuckles Liemandt. “You spend a hundred grand to send your kid to school, and he wants to drop out to work for another kid.” Diploma in hand, Stratton left Trilogy and was replaced in the lineup by Thomas Carter in 1990.
So young Joe Liemandt and his team got to work applying algebraic algorithms similar to ones used by mathematical economists to build general equilibrium models of the economy to create patterns that enable manufacturers to build synthetic models of their product lines. He and his crew spent thousands of hours “writing code” — programming that is — late into the night, sucking down Jolt cola for its caffeine content. Corporate strategy meetings were held across the street at the Old Pro, a Stanford bar, and were fueled by “a lot more beer than Jolt,” Liemandt remembers.
Much of this programming was trial-and-error, but by the time he quit school Liemandt had figured out how to handle thousands of if/them equations in a system that was easy enough to be used by the average salesperson. He would combine the two programming approaches then being experimented with. He would use constraint-based equations to impose some order on the rule-based programming approach DEC was pursuing. Later Liemandt and his team would blend in a third approach, object-oriented programming. Very simply put, this approach enables programmers to organize pieces of software into independent “objects,” or modules. This allows programmers to easily add new information, such as when a product line changes, or to fix errors in the software without shutting down the entire program.
In early 1991 Trilogy’s programming was beginning to jell, but Liemandt received bad news. His father was diagnosed with terminal cancer of the esophagus, and given a year or two to live. Partly to be closer to his parents, Liemandt moved Trilogy, now a 13-person enterprise, to Austin, Tex. Trilogy had by this time created an early version of the software is now sells, and had applied for patents covering its algorithms (two have been granted; three more are on application). But the company still had not sold its product to a single customer.
By moving to Austin, however, Liemandt was able to hire Dave Franke away from research consortium MCC. A heavyweight in computer science, with years of programming experience, Franke, then 38, gave Trilogy some much-needed credibility among the corporate customers Liemandt hoped to reach. One month after Franke came aboard, Hewlett-Packard signed its first contract with Trilogy; a $ 3.5 million deal for sales-configuration software and support services. HP was in effect abandoning its own pursuit and giving the business to Trilogy. “It was the greatest feeling in the world,” Liemandt says of getting HP’s business.
With the nod from HP, other big customers began ordering from Trilogy — Boeing, Silicon Graphics, Alcatel and scores of other high-technology and telecommunications firms. In 1994 erstwhile competitor DEC paid Trilogy’s programmers a high compliment when it began to buy their software.
Lacking this kind of software, the salesperson must serve as a kind of middleman between the client and the company’s engineers, with constant consultations between the sales call and the final order. Further complicating things, different pieces of a manufacturer’s equipment might not work well when hooked up to a customer’s existing equipment, an unpleasantness often discovered only after the order had been delivered. Trilogy software, tying together all the relevant information, greatly shortens and simplifies the process. According to Trilogy, the typical U.S. manufacturing company spends 2% of its gross revenuers to rectify human errors and miscalculations in the configuration process. Saving even half of that would be worth billions of dollars to businesses.
Using Trilogy’s Selling Chain software, a manufacturer loads into the program all the options and product specifications he offers. In effect, he’s turning all his product catalogs into electronic files. Once all this information has been turned into bits and bytes, the salesman can huddle with his customer and ask the company what/if questions. What if my customer links his new Sun workstations to an HP OfficeJet printer-fax-copier, rather than to an HP LaserJet printer? Would that work with the customer’s IBM server? What would it do to the price I can quote him? Does the customer need new connecting cables? Special plugs?
In 1994 Beaverton, Ore.-based Sequent Computer Systems was meeting its sales targets, but was regularly exceeding its cost-of-sales budgets. Robert DeBakker, Sequent’s director of order fulfillment, found the company ($ 540 million revenues) was spending lots of money correcting mistakes that originated in sales configuration. “Come quote time, we’d have a conference room full of catalogs and price books,” says DeBakker, explaining that Sequent customers sometimes received servers without room for the required disk drives, or without connecting cables, or with the wrong kind of preinstalled software. Fixing these mistakes cost Sequent money, time and customer loyalty.
In 1994 DeBakker shelled out about $ 700,000 for Trilogy’s software and services. He says the investment saved Sequent $ 3 million in its first year, and would have saved more had it been implemented sooner. “It made me a hero,” says DeBakker of Trilogy’s software.
Boeing is counting on Trilogy’s software to help it cut costs. A 747 is made up of over 6 million parts, and a customer can choose among hundreds of options. How many seats? What kind of avionics? How many bathrooms? Do you want carbon composite landing gear or steel? Every option the customer chooses affects the availability of other options, and changes the plane’s price. It takes the sales agent days or weeks working with company engineers to make sure all the chosen pieces fit together, renegotiating the price at every step. “It can take 22 trips,” says Doug Frederick, head of information systems for Boeing’s redesign effort. Trilogy’s software sharply cuts that waste of time and resources. The software knows which parts go together and how much they cost. So the salesperson can sit down next to the customer, turn on his laptop and configure the 747, complete with a price quote, in one trip.
Who’s the kid who beat the technology giants to this giant market? Says Wade Monroe, Trilogy’s chief financial officer, and at 54, one of Trilogy’s elder statesmen: “You know how some parents clone their kids to become sports stars? That’s sort of what happened with Joe in business.”
From earliest youth, Joe Liemandt was steeped in business lore. When Joe was a boy in the 1970s, his father, Gregory Liemandt, was the head of planning at General Electric’s components and materials group in Pittsfield, Mass. His boss was an up-and-comer named Jack Welch. Young Joe played hockey with Welch’s son, Mark, spent weekends at the Welches’ and often went on ski trips to Lake Placid with the Welches. Over the years the Welches and Liemandts have remained friends. Joe Liemandt didn’t lack for suitable role models.
In 1983 Greg Liemandt quit GE to run UCCEL in Dallas, a maker of software for mainframe computers, and Joe’s course was set. His mother, Diane, remembers her adolescent son rifling through his father’s briefcase in ssearch of business plans to read. When he was 16, Liemandt spent the summer of 1984 as a programmer in UCCEL’s research and development division, where he learned a lesson he would remember when he started his own company. “They had huge teams of programmers — 50 to 100 people — working on a single project,” Liemandt recalls. “It was a fiasco.” Coordinating the efforts of so many programmers left little time for the actual programming, and projects soon fell apart.
Quick to learn, Joe applied that lesson in his own business. “You need one or two programmers on a project at a time. Any more, and it becomes a nightmare,” he says. Liemandt entered Stanford (chosen, he says, for its proximity to Silicon Valley) in the fall of 1986. He later signed up as an economic major, but spent much of his time taking computer programming calsses and hacking around the computer lab. He knew already that software would be his career, but it wouldn’t be entertainment software. “The product,” he says, “had to be something my dad or Jack Welch would buy.” The something, as it turned out, was sales-configuration software.
An obvious question: Did the example of Harvard’s most famous dropout, Bill Gates, inspire Liemandt’sdecision to leave Stanford early? No, he says, dropping out was “a side effect of, of . . .” he searches for the right word “. . . of a feeling that you’re in a race, and school is something that’s in the way. Nine months is an eternity in software [a reference to the senior year in college he blew off]. Ask Netscape if they’d give up nine months. Nine months out of their browser work, and Netscape isn’t in business today. Somebody else would have done it.”
Starting Trilogy was the easy part. Keeping it going in the early years wasn’t. Liemandt’s father had made about $ 30 million when Computer Associates bought UCCEL in 1987, but Gregory made it clear that he would not finance the vision that lured his son from school. “Outside of me slipping him $ 100 bills in the airport so he would eat,” recalls Diane Liemandt, Joe’s mother, “there was no [family] seed money.” Joe Liemandttried to get Kleiner Perkins and other venture capital firms interested in Trilogy, but even in Silicon Valley the idea of backing a crew of 21-year-olds is a bit hard to swallow. No go. So Liemandt took consulting jobs and began leveraging one credit card against another. “At one point, I had 22 credit cards out there,” he says. “Luckily, I never missed a payment.”
Liemandt and his sidekicks realize that the configuration software business is too good to be ignored by competitors. Trilogy does not have the field to itself. A few smaller competitors have sprung up, but the real threat now comes from larger outfits. European software powers — SAP of Germany ($ 1.9 billion sales) and Baan of the Netherlands ($ 216 million) — have established themselves in automating back-office operations, as had Oracle Corp. Now all three companies have started to make forays into sales automation.
But the astounding successes of Microsoft and Intel have proven that in the computer world the spoils go to the company that grabs the market share early on and keeps it, leaving competitors in the position of always trying to catch up. So it’s ful speed ahead for Trilogy to keep its early high market share as the market grows. “It’s a land grab,” Liemandt says. “Whoever gets the market share and partners first, wins. We’ve got two years until it [the land grab] is over.”
Liemandt says old family friend Jack Welch encouraged him to pull out the growth stops, when Liemandt and his mother recently visited Welch for a weekend in Connecticut. “After talking to Jack, I feel like I’m going half speed,” Liemandt reports. “He tells me, ‘Don’t be such a wuss, Joe.'”
Trilogy draws most of its new hires from the computer science departments of Stanford, MIT, Harvard, Berkeley, Carnegie-Mellon and Rice. While Liemandt can’t match Bill Gates in salary, stock options and name recognition for getting top recruits, he makes up for it by giving the youngsters big projects to run right away. “Here’s more responsibility than you think you can possibly handle — go!” is how Samer Dholakia, 22, a recent Stanford graduate, describes the attractions of working for Trilogy.
Liemandt knows most of the kids he’s hiring are too smart to work for him — or anyone else — for long. “There are a lot of people here to learn how to start a software company,” he says. “That’s great.” He says he’d rather get a few years of boundless entrepreneurial energy from his employees than get many years of mediocre performance.
An inevitable question: When are you going public, Joe? He answers that he’s had 20 calls in the last year from investment bankers drooling to take Trilogy public. “The market’s way overpriced, absolutely,” he says, agreeing that an IPO would probably value Trilogy somewhere around the $ 1 billion mark. But he didn’t spend all those youthful years dreaming about being a businessman for nothing. The earlier he goes public, the more equity he will have to give up. “You’ve got to starve before you give up 51%,” he says. Bill Gates waited 11 years to take Microsoft public, and staying private allowed Gates to weather radical changes in the early software industry without having to worry about quarterly earnings dips.
Anyhow, Liemandt says, “Making 8 zillion dollars is not the win. If that was it, we’d go public tomorrow and cash out. But if we don’t change the way people buy and sell things, then we blew it.” Besides Liemandt, who owns about 60% of Trilogy, otherprincipal shareholders are the Stanford schoolmates remaining from the early days — John Lynch, Chris Porch and Tom Carter. Christina Jones traded her Trilogy holding for a chunk of a Trilogy subisidary that she hopes soon to take public.
Joe Liemandt and his partners are, in a way, the prototype of the new entrepreneur. He is now around the age Bill Gates, Steve Jobs and Michael Dell were when they acquired early success. Liemandt cites a few factors that give youth such an advantage in today’s business. “Young people are in a much better position to take risks, because they don’t have families, mortgages and house payments to take care of. You’ve got to be willing to jump off a cliff for your idea.
Risk is almost an obsession with Joe Liemandt. He likes to visit Las Vegas and the crap tables with associates. He thinks it’s good for morale. “Nothing,” he says, “brings a group of people together like risk.” In 1994 Liemandt bet a Trilogy programmer his blue Lexus that the programmer couldn’t finish a projectg in time for a client proposal. The programmer upped the ante: If Liemandt lost, he would have to fork over his Lexus and drive an Aspire, Ford’s smallest subcompact car. Lexus-less, Liemandt crammed into the Aspire for over a year.
Along with an appetite for risk, Liemandt cites a generational change that makes the computer industry so hospitable to the very young. Computers aren’t something they learned. Computers are what they were brought up with. “Many people graduating from college today have been using computers since their early teens. Kids who have been using the Internet since they were 15 — they’re 25 now, and they’re ten-year Internet experts. [Technology] is changing so fast, there is nobody who has ‘expertise’ in the area. What matters is the ability to learn, adapt and figure out what the answer is. You’ve got to be willing to get in over your head and struggle to make things happen.
“Six months from now, whatever you’re doing, you will have to do something different.” that short sentence could well serve as a credit for today’s aspiring entrepreneur.
Let’s block ads! (Why?)