A Truck Full of Money (18 page)

Paul's place by the pond had a curious history. The land and the oldest section of his house were once part of the estate of a nineteenth-century magnate named Addison Gage, an entrepreneur of ice. He harvested it from Spy Pond and other locales and shipped it far and wide, to America's warm southern states and also, according to one local chronicle, “to the East and West Indies, to China, to India, to South America, even to Australia.” Paul knew some of the history and would entertain guests with it, drawing a moral: No matter how wacky, an idea can succeed if it aims at a big market (all those hot places without refrigeration) and if there are technologies to execute it (special saws, sawdust in the holds of sailing ships, insulated buildings).

What vision the ice barons of New England must have had, to look at the ice on a Spy Pond and think,
I could send this to India.
It was a fitting heritage for Paul.

For years, ideas for new enterprises had flowed into his mind, usually at night. He would attach names to the relative few that survived the morning. Karl used to register these as Internet domains, and Paul would put them on a master list. In 2013, it contained 158 names. Karl had never known what some of them signified. Paul himself had forgotten some:
APPMOVE, CARDBLOCK, CLOAKLY, GOANDCO, PAYTONES, SEARCHLETS, SOFTWARESURVEY, TOWER
51
. Likewise
DRAGINBOX
. He gazed at it. “I don't remember. But it was something really fucking cool.”

Some names denoted nothing. He just liked their sound:
AVIDFOX, CONNEXAS, CONNEXOS, ENIGENCE, ETELL, HEROPORT, SYNVEO
. Some unattached names came from trips—
ZARZAN
from Norway,
UMANU
and
TUMTUM
from Kenya. Some he'd found by brainstorming through the alphabet. Once, years back, he wrote a short piece of code, a software robot that would find every two-character domain that was still unregistered. The script produced fifty names—such as “B1”—but he bought only a few, and still counted this among his important commercial mistakes. By now, he figured, every two-letter combination had probably been claimed, and each was probably worth at least $100,000. He and Steve Hafner had spent $30,000 on
KAYAK.COM
back in 2004. This had soon looked like a bargain, because they had failed to also buy the German domain name
KAYAK.DE
, and a few years later they ended up paying 100,000 euros for it.

Technically, you can't buy a domain name, but you can in essence rent the exclusive use of one. In the early days of the Web, you almost always did this through an official registry, and you could claim most names for a nominal price. But now, in 2013, about 270 million domains had been registered. Traffic in names had become like a vast international used car business, where the rights to names were bought and sold, some for astonishing prices—typically, brand names or descriptive names likely to turn up early in a Web search ($30.6 million for
INSURANCE.COM
was the current record). Paul made a hobby of playing around in this market, a yard sale haunted by crafty middlemen and shady characters. Just recently, he'd spent $26,000 on two dot-com domains—$10,000 for the initials of his gun control venture,
AGL
, and $16,000 for
ROADWARS
. To recover some of his trading capital, he had recently sold
SNAPCAB
for $10,000. The idea behind that name was a mobile-app-based system of taxis and limousines. It approximated—and predated by about five years—the basic plan of Uber, a current e-commerce phenomenon whose value was placed in the billions. Paul had wanted to build
SNAPCAB
for Kayak, but Hafner had discouraged him—wary, not without cause, of Paul's distractibility.

Prescience is a retrospective virtue. Paul's way of saying this was to tell how he had been introduced to the people who were starting YouTube and had thought,
Yeah. Whatever.
The company on which he had lost two million, Mancala, was a version of the now profitable websites dispensing advice to consumers, a Yelp precursor, a good idea pursued too early, an example of premature prescience.
SNAPCAB
had been prescience unpursued, a might-have-been. Paul's list of names contained a couple more of those:
BOSTONBOUNCE
, which, like Groupon, would have sold discounted coupons to consumers;
ERADIO
, which might have been an early satellite radio company. His list contained a passel of stillborn software talent agencies that would have served both domestic and foreign markets. These ranged from
BOSTONASSIST
to
MUMBAICODERS
. A few names of this species memorialized dreams of stirring up economic development in African countries where he had philanthropic interests:
KENYACODERS
,
NAIROBICODERS
,
KIGALICODERS
.

Of these, only
KIGALICODERS
had come to life, just briefly, in Rwanda. “We found some kids in Kigali. We wanted to teach them to be coders so they could make some money. We built a simple website. We taught them how to bid on jobs on Rentacoder. The company failed, and I don't know why.”

Some names were souvenirs:
XIANGQI.COM, BOSTONLIGHT.COM
, which he had bought back from Intuit.
F
7
was the chorded command that used to summon up his email program and now recalled the decade when he and Karl had emailed each other at least once a day.
PARKINGPATH
recalled the blinding light of a hypomanic hour when he was listening to an urban architect talk on public radio about the importance of trees, and suddenly realized that he must start a campaign—no, a movement—to bring tree-shaded paths to parking lots, one for every two hundred cars.

There was
GOODENOUGHSOFTWARE
, a nod to the way his father had of pounding in a nail and saying, “That's good enough.” The idea sprang from irritation: “Microsoft keeps releasing a new version of Office every year. Who needs all those bullshit features? I would take the ten most successful software apps and strip them down to the twenty percent most people use.” There was
SIMPLELIST
, a design for the simplest way to make a list. And
KWHEEL
, a tool for counting the numbers of colors on websites.
MAILSTATS
came from a program that Karl wrote for him back in Interleaf days; it let your employees analyze their email usage and gave them tips on how to curb the excess. There were
BUZZCARZ
, a now defunct discussion forum about cars;
BOSTONDETAIL
, a mobile service for cleaning cars; and
CARSICK
, an app to resolve arguments between husbands and wives as to who was the better driver, which, along with
PRODRIVEAPP
and
WHIPWARS
(in current teen talk, “whips” meant cars), had led on to Road Wars, its future still in doubt.

SECRETINSIDETHEORANGE
was the name of a movie he planned to make, in order to introduce xiangqi to the Western world. The name comes from an ancient Chinese fable: Two very large oranges are found in an orchard and cut open to reveal, in each, two very small and elderly men intently playing xiangqi. Traditionally, only boys had been taught the game, so he planned to have a young Chinese woman explain the rules on camera. He had long since made an outline of a script and still told himself he'd make the film next year.

Of civic projects that had died or never been tried out, there were
CENTRALAUTHORITY
(to identify and debunk, through parody, lies gone viral on the Web) and
KIDMAIL
(to protect children from online pornography and predators): “I didn't pursue it enough.”

PURPLEWATCH
was a favorite. “I want to be judgmental about people who buy $10,000 watches, but you're talking to a guy who buys $200,000 cars.” He would execute the plan when he found the time: He gets Swatch to make an inexpensive all-purple watch. They sell it for $10,000. The profit goes to a charitable foundation. The people who buy the watches get to say where the foundation sends their ten thousand, and they also get to show off their luridly colored watches to people at their country clubs and to say, as Paul would have it: “I'm rich, too, motherfucker. Except my ten grand went to an orphanage.”

In the realm of ideas, regret was a waste of time. Better to think of the handful of successes.
IVRCHEATSHEET.COM
referred to the interactive voice response systems used by every big corporation. The idea had come to him around 2002, in the days when his father was slipping into dementia and Paul began hearing him try to talk to the robotic voices at Comcast and Verizon. “And I'm like, Okay, I took his car away. Now these machines are taking his phone and TV away. I might as well dig a hole in my backyard and stick him in it. So I was like, The guy should be able to talk to humans. He's paying these assholes at Comcast a hundred dollars a month. Why won't they talk to him?”

Paul raised this question on
PAULENGLISH.COM
. He asked readers to send him the unpublicized numbers and codes that would connect to the phones of actual people at big companies—numbers that would circumvent the offensive, indeed “evil,” IVRs. Paul posted the numbers on his blog. Word spread. Phone numbers poured in. The
Globe
did a story about the local software entrepreneur who was taking on the giants of corporate America.
People
magazine and
The New York Times
ran articles. Paul appeared on NPR and the BBC and national television shows, and his blog went viral—a million visitors in just one month. So he created the website. His assistant at Kayak told him he needed a better name for it. She came up with
GETHUMAN
. Karl, out in Oregon, secured the new name, and Paul created the site. It contained a list of the various secret numbers and codes and procedures that would connect you to a human being at a company—four hundred companies and growing when the
Times
article appeared—and advice on how to use those numbers, and also Paul's thoughts about how companies ought to treat their customers. He told the
Times
reporter he wanted GetHuman to lead a movement that would “change the face of customer service.”

In one of his lectures to students, Paul said: “A lot of successful companies' products are created not with a business plan but instead from how much the person is irritated. And those people just say, I don't care, or This makes me so angry I'm going to dedicate a year to doing this, and then if the product is really good and people start using it, then they figure out how they can make money on it.” GetHuman had accomplished the first part of that sequence. It had gained a large audience, and large audiences can sometimes be monetized. But Kayak was only two years old. Paul didn't have time to figure out how to make money with another website, let alone carry on a movement. Eventually he handed GetHuman to a young coder friend, who made it into a small but profitable company. Traffic to its site was still growing, on its way toward fifty million different visitors a year.

BUZZMED
had turned into a website called Global Health Delivery,
GHDONLINE.ORG
—a flourishing online branch of a system created by Partners In Health and Brigham and Women's Hospital. The idea was to codify and disseminate proven techniques for fixing things like public water supplies in impoverished places and for treating illnesses such as tuberculosis and malaria. For about a year, Paul had stretched his working hours, running Kayak engineering while designing the online forum and assembling a team to build it.

JOINAFRICA
would have extended that work, spreading the Internet all over the continent. An overly enthusiastic public performance ended with a magazine story headlined “Kayak.com Cofounder Paul English Plans to Blanket Africa in Free Wireless Internet.” Paul thought this made him sound racist, a great white bwana who would wire all of Africa: “I had two weeks of near panic attacks.” But he did bring Internet access to medical outposts in Burundi, Uganda, Zambia, and Malawi, buying the satellite dishes and other equipment, hiring others to install them, and sometimes traveling to the sites himself to help.