For Ridgely Evers, developing new technology has never been an end in itself. In boarding school, technology got him out of waiting tables. In college, understanding technology was a means to world peace. In the 1980s, he preferred to write software than do bookkeeping, so he created an application that evolved into QuickBooks.
These days, Evers has his hand in a number of businesses—strategic consulting for large companies, developing cloud-based software for his own start-up, and running a 28-hectare commercial farm that grows, among other things, olives and grapes for DaVero Sonoma, a company he also owns, which produces gourmet olive oils and Italian wines. Farming is one process, he’s found, that technology cannot affect; plants simply have to grow themselves. And he’s just fine with that. "I really love and am proud of the diversity of my life," says Evers. "I never let myself get pigeonholed."
Evers, an IEEE member, built his first software product at age 17, as a high school student at St. George’s in Newport, R.I. The boarding school, required all students to take a turn at serving in the dining hall. Evers, who hated waiting tables, spotted what he calls a "perfect linear programming problem" in the dining hall. That is, the 250 students and 50 faculty members had to be assigned to new tables every two weeks, following a rule that students couldn’t sit at the same table or with the same professor in a six-week period. Creating the seating charts manually was a nightmare. So Evers wrote a program to do the seating charts, using a Digital Equipment Corp. PDP-8S, with 4 kilobytes of memory, a Teletype, and a high-speed paper tape reader. To make it work, he had to splice paper tape into a loop so the tape reader could cycle through it over and over as it looked for student records. This paper loop was fragile, so Evers had to sit with the computer while it ran his program; if the tape bent or ripped, the program would crash and he’d have to restart it. The senior who managed the dining hall was happy to let him do that instead of wait tables.
Evers bounced around a bit in his early college years, eventually landing at Stanford, majoring in international relations. That doesn’t sound like a very technical program, but in practice, it was. It was a new major, designed to take advantage of Stanford faculty members who had worked on the Strategic Arms Limitation Talks Agreement in the early ’70s, and it included physics and math as well as psychology, economics, and political science. Evers added computer science courses to that mix, simply because they interested him.
Graduating in 1977, he and two other students started a consulting company, initially doing work on nuclear weapons nonproliferation for the U.S. government, then later researching weapons systems and selling their reports. There were always at least two sure-fire customers for each report, he recalls—the CIA, and a company in Wisconsin that he later found out was a front for the KGB. Evers and his partners sold their company in 1980, and he went back to Stanford for an MBA.
Next came a real estate start-up in Phoenix geared to first-time home buyers, based in part on software Evers developed while at Stanford. He modeled a system under which first-time buyers could initially rent a home and then apply the appreciation of the home toward a down payment. To make this a commercial system, he needed to build a user interface; he did that in Basic, on an early IBM PC. The business sailed along until the savings and loan crisis took down his company’s principal lender and the Phoenix real estate market crashed in 1984.
In a second run at the real estate business, Evers and several partners started a brokerage for home buyers, not sellers. He built a system that mined data from Multiple Listing Service computers to make them searchable by a variety of parameters. He also built accounting software for the company, something he had done for his previous ventures.
"Manual bookkeeping is a stupid process," he says. "Spreadsheets are not a good place to do accounting either."
The business never took off, and Evers left, taking the accounting package with him. In 1987, he met Intuit’s founder, Scott Cook. Intuit is now a public company with about US $3 billion in sales annually, but at the time it was a start-up with about 30 employees. Cook looked at Evers’s accounting package and hired him as a consultant to write an invoicing add-on for Quicken personal finance software. Evers started working on the project, but he argued that small businesses needed a dedicated accounting package, not just an add-on to Quicken. Cook finally agreed, and Evers set out to build QuickBooks.
"At that point," he says, "we made a profoundly dumb decision—to build QuickBooks on the Quicken code base."
The Quicken code, written for computers with little memory or storage, was brilliantly tight. For example, the ASCII symbol set, containing the Roman alphabet, numbers, and special characters, is coded in a standard seven bits; computers digest data in 8-bit chunks. The extra bits are typically left blank, but in Quicken, programmers used them for all sorts of critical data—for example, to record whether or not the bank had cleared a transaction. It was a mess to work with, Evers recalls.
"We slogged to get it out," he says, "but we finally got it launched" after two years instead of the anticipated nine months, during which Evers came on board as a full-time employee. "That," he says, "was one of the dumbest financial decisions I ever made, to switch from being a consultant with a percentage of QuickBooks sales to an employee with stock options."
Comments