It's a crisp spring Friday in Kansas City and Jim is upset with Google. He's not peeved about a search result or worried about the data Google collects on him. Jim, a slender music instructor in his early 40s, with salt-and-pepper hair and a polka-dot button-up, would like more attention from the company.
He gets internet from Google Fiber, the service that says it can deliver broadband at incredible speeds. But he has issues. It keeps going out.
So here he is at the "Fiber Space," a trendy-looking storefront on a street that straddles the states, splitting the city in two: Kansas on the west, Missouri to the east. The walls are papered with Google Maps printouts. One reads: "Super fast downloads. TV like no other. And endless possibilities."
Jim is calm but irked. He runs a business from home, he explains, and when the connection drops, he loses money.
"This is the same issue as with Time Warner [Cable]," he tells the young customer service rep. "I'm getting this mirror of interacting with this other huge company. And you've always said, 'Our stuff doesn't do that.'"
Google began digging up dirt and laying fiber optic pipes in Kansas City, Kan., five years ago in April. Its first customers were wired the following year.
For the years after, it was unclear — certainly outside of Google — just what Google wanted to accomplish with this first venture outside of its core business. Now it's evident: Google was using Kansas City as a testbed for an audacious project — one to take on broadband providers like Comcast, Time Warner Cable and Verizon, which enjoy long-held duopolies and monopolies across the country, and build out a national service. To provide real competition.
Googlers won't say this out loud, but they despise the cable industry. They find it inert, predatory and, worst, anti-innovation. So Google wants to replace it.
Google has to figure out how to become a nationwide broadband company. Kansas City may offer a road map.
And unlike the other moonshots now housed under Google's umbrella company, Alphabet, Fiber is a project that could succeed in the near future. Google Fiber is now up and running in five markets (it added Nashville last month) and has plans under way for 17 more. Expect that pace to quicken.
Which means Google could end up hearing from many more Jims in the future. That's the downside of telling people you can do better than the monopoly provider they already hate — you have to deliver on your promise.
But Google can deal with that challenge down the road. First, it has to figure out how to become a nationwide broadband company. Kansas City may offer a road map.
I traveled to Kansas City last month and spoke to dozens of Fiber customers at their homes and offices. Nearly everyone I talked to is thrilled with their Fiber subscription — the cost and service are miles ahead, and, boy, is it fast.
In recent months, Google Fiber branched out from the central city to more outlying suburbs. City officials told me Fiber is available to 210,000 households in Kansas City, Mo. — or roughly 80 percent of its population. Former Fiber employees said its pipes easily cover one million households in the entire metro area.
"They pretty much built out an entire city," said Rick Usher, assistant city manager for small business and entrepreneurship in Kansas City, Mo.
Back when Fiber began, a high-ranking Google executive laid out the worst-case scenario for the search giant, former Fiber staffers recalled: Google pours billions into a few test markets, scares the incumbents into speeding up their service, but never gets traction. Worst case is still very good for Google, since the company relies on people going online and searching for things.
But the worst case never happened. People familiar with Fiber say it has hit its initial customer targets in its first three markets — selling broadband to about 30 percent of the homes it has hooked up for service, the industry standard for feasibility. Fiber brought in roughly $100 million in revenue last year, according to sources.
Last month, the business embarked on the next wave of its strategy, starting in Kansas City again. It won approval to place special antennas on city light poles that could potentially beam broadband directly into homes, over the air. Fiber wants to start testing out the equipment in November and plans to have a network up and running by the end of 2017.
Wireless is a big deal for Alphabet. If it works, it means it can deliver broadband without having to build out or buy fiber networks. No dirt to dig up; no last mile to cross. That means its network can swell much more, much faster.
Pipes aren't cheap. It cost Google more than $1 billion to spread across the Kansas City region and will likely cost as much in each new Fiber city, according to sources.
But wireless could be far cheaper — a fifth of the cost of fiber, which is roughly $1,000 per home, according to industry insiders. Other executives, however, say the cost savings would be minimal. Since it's an unproven technology, costs are still being sorted out. Getting a wireless network to work and run cheaply is far from guaranteed.
Even if wireless works, its deployment is real money, even for Google, which made just north of $16 billion in profit last year.
One way to get that money is to tap the debt markets; Wall Street would be less likely to get spooked seeing money borrowed, since Fiber and the other moonshots are now isolated from Google's profitable core business.
In several of its newer markets, Fiber is also tweaking its approach, relying less on expensive digging. In San Francisco, it plans to use the city's dormant fiber pipes; in Huntsville, Ala., it is tapping the city's existing network as its own.
Mapping the Fiber cities
Google is building out its Fiber network city by city (green markers). But the biggest U.S. cable and phone companies are upgrading. Select a city to see its providers.
"We've learned a lot by building networks from scratch. We continue to invest in that model for expansion, but we're also finding new ways to bring our service to cities," said Chris Levendos, Fiber's head of network deployment.
And, maybe, with less cost. Kansas City was a good test of whether Google could set up a viable alternative to the cable giants, but now Fiber knows it has to find new models to really take them head-on.
How to boil a frog
Fiber was born at a nexus of tinkering and worry.
Go back a decade. Google had won search and was minting money. Its founders started tinkering with far-afield tech and began obsessing over wireless networks. Google was also deeply paranoid — it wanted the shortest distance between its services and users.
But Microsoft ran the operating system and the web browser while the big cable and telecom companies like Comcast and Verizon controlled the pipes.
Google was particularly worried about the pipes. Eric Schmidt, the CEO at the time, "pounded the table for many years" about the prospect of ISPs throttling Google's content, an early Googler recalled.
Google then formed Access, the business unit that now houses Fiber, in 2006, to tackle what they saw as a potentially serious threat. After a few half-baked possibilities (founder Sergey Brin once eyed blimps to deliver internet), by 2010 Google grew serious. It brought on Milo Medin, an internet and telecom expert, to run Fiber. Oversight went to then CFO Patrick Pichette, another telecom veteran.
Fiber was designed to operate as an autonomous business unit insulated from the remainder of Google. In theory. Some former employees described frustrations with debilitating ties to the mothership. They had trouble recruiting subscribers, for instance. Google required customers to sign up with Google email and payments accounts.
But Google kept giving Access and Fiber attention and money. It came from the top.
"As the ambition and scale of our projects accelerated, so did our budget and staffing," said Chris Sacca, a founding Access member at Google. "We were lucky because [Google co-founder] Larry Page has always been obsessed with access networks."
The team was motivated, in part, by the stasis of the broadband industry. The numbers back them up. Only a third of Americans have more than one choice for broadband providers, according to a January FCC report. A tenth have no choice.
For inspiration, the early Fiber team passed around a line from Time Warner Cable’s chief Glenn Britt: The CEO was bragging to analysts about the fat margins in his broadband business. To Fiber's team, Time Warner's broadband was a rip-off — slow and overpriced. They wanted to go after executives like Britt.
That may be one reason why Google chose Kansas City, Time Warner territory, from a field of 1,100 cities that applied. Britt was asked at an investor conference how his cable company viewed Google's Kansas City expansion. The late executive replied: "Even if Google builds, we’re not going to wake up and see Google instantly building out the whole country."
But that was always the plan, according to a former Fiber staffer — "to grow to be nationwide at some point."
If wireless works, Fiber can grow much faster and become a national threat.
Wireless was always part of the plan, too. "We talked about wireless deployment," said Joe Reardon, who served as mayor of Kansas City, Kan., when Fiber launched. "It was always a concept in conversations."
Craig Barratt, the CEO of Access, now a separate Alphabet company that runs Fiber, is enamored with wireless technology. Medin, who passed on Fiber's management to Dennis Kish, a Qualcomm veteran, last year, now serves as Google's envoy for wireless discussions in Washington, D.C.
The official line from Fiber is that its wireless tests are just "experiments" and may not necessarily become commercial products, and Barratt won't say that the company intends to compete nationally. Yet those two are very much linked: If wireless works, Fiber can grow much faster and become a national threat.
Why not say that out loud?
One former Fiber executive summed up the strategy, a longtime Google tactic to keep rivals unalert. "If you’re going to boil a frog," the exec said, "you don’t tell the frog."
The frog is finding out anyway.
A pattern has emerged whenever Fiber says it’s coming to town. Pretty soon thereafter, AT&T or Comcast — whichever the incumbent is — lowers prices or unfurls gigabit internet of its own. (Comcast, via its NBCUniversal unit, is a minority investor in Vox Media, which owns this site.)
In Kansas City, residents said Time Warner Cable initially met Fiber’s arrival with scorn. Who needs a gig? Eventually, Time Warner Cable reps started showing up to city meetings. Then the billboards arrived — 300 megabits per second, only $64.99! Followed by the mailers pleading with subscribers who left for Fiber to return.
Usher, the Kansas City, Mo., official, chuckled in recollection. "I call this the seven steps of acceptance of the gigabit revolution," he said.
We met for coffee at YJ’s Snack Bar, a tiny, three-tabled cafe in the city's arts district. Usher greeted the tattooed cashier and two men in flannel sitting at the tables. Then he shook the hand of everyone else who entered. He wore a suit, but wore his silver hair long. He seems to enjoy his job, and clearly likes that Fiber picked his city. Others credit him with carrying the municipal grunt work that made Fiber’s massive civic works project possible.
"He’s the quarterback," said Tim Cowden, president of the Kansas City Area Development Council.
I asked Usher what Google had done well.
"Rather than building it over 30 years, like AT&T and Time Warner have done, they built it in three years," he said. To lay the network, Fiber went from dig to done in a neighborhood in three weeks. Usher compared it to the formation of the railroads. "It’s like hell on wheels. You had these little towns and construction activity that" — his hand shot up — "whoosh plowed across."
Speed has drawbacks. Fiber's quick deployment has sometimes rankled residents. Occasionally, the construction would plow across a sewer line. When residents figured it out — the sewage backed up into their houses — the Fiber crew had moved on.
A Fiber rep declined to comment.
When Google’s wireless plan reached the Missouri City Council, two members voted no. One was Dan Fowler, a lawyer. "From my own point of view, Google has done, in general, a very poor job of communication with people in Kansas City about what they’re doing," he told me on the phone. "And the prime example is me."
Fowler said he was promised a Fiber connection to his home a year and a half ago. It arrived last month. "They have also, through their contractors, a very bad history, from my perspective, of not cleaning up their messes," he added. He said his vote was primarily intended to send Fiber a message.
It was not the only message Kansas City sent.
When Fiber arrived, Google carved out the metro region into 202 "Fiberhoods." These areas could qualify for Fiber installation — but only if a certain percentage of residents agreed to sign up.
This was excellent marketing, guaranteeing demand and even prompting neighbors to go door-to-door on behalf of a then $250 billion company. Usher’s son, an avid gamer, campaigned with a pin reading "Do it for lower ping" — the latency for streaming. (Gaming teens are a frequent rationale cited for gigabit internet; when Reardon, the former mayor, told his family they were adding a Fiber connection, his son, also a gamer, leapt from the table to hug him.)
But the "Fiberhood" approach had a flaw.
Troost Avenue, on Kansas City's Missouri side, cuts the city along class and race. In 2012, Wired ran an article about Fiber's initial rollout that included a map of the neighborhoods registered for the service. It was sliced in half by Troost Avenue — on the west, the city's whiter and richer neighborhoods were registered; on the east, its poorer and black neighborhoods were not.
Kansas City Congressman Emanuel Cleaver wrote that the service could be the start of "potential ‘digital redlining’."
Googlers argue they really care about closing the so-called "digital divide." But they were also not used to grappling with unplugged populations or wrestling with sensitivities of local civic life. When it started, for instance, Fiber was unable to sign up subscribers without bank accounts.
"Sometimes Silicon Valley can be a little bit tone deaf," said a former Fiber staffer.
After attention was drawn to Troost Avenue, the company dispatched several dozen Googlers from headquarters to blanket the street's east side for sign-ups. A Fiber rep wouldn't specify the neighborhood coverage breakdown, but said that "dozens" of the 181 Fiberhoods that qualified for the initial rollout came from that effort and that 94 percent of all central city neighborhoods qualified for Fiber.
Google was sensitive to the issue in other ways. Since arriving, Fiber has selected 100 nonprofits in the Kansas City area to wire free of charge. It recently added a $15-a-month broadband service. And the unit also unfurled a national program to service low-income housing for free.
To Google’s newfound competitors, however, the internet giant was all too competent in its rollout. Comcast and AT&T were miffed at how Google was able to pick and choose its neighborhoods, avoiding "build out" regulations that had required cable companies to offer wider coverage. (The national rules were ditched in 2007 as a barrier to broadband competition.) AT&T has accused Fiber of buddying up to local officials to push ordinances favoring its expansion. In Louisville, AT&T is suing the city over this.
At Fiber's onset, Google saw a clear economic upside in smaller cities. Sources said Medin would demonstrate this with a simple chart: Population density on the X axis; cost on the Y. He then drew a smiley face. Fiber's sweet spot was the bottom of the lip, mid-sized cities where installation was not too expensive and national ISPs did not have major operational focuses.
Fiber is now rolling out in Los Angeles and Chicago, two prime markets for competitors and a test of its economic model. In particular, those big cities will test Fiber’s cable TV service, which has not had considerable traction.
At the onset, Fiber hoped it could sell TV "a la carte" — one channel here, another there — but quickly met resistance from programmers, who have long sold their channels in bundles. Securing pay TV rights was also pricier than Google expected. In its original deal in Kansas City, Google failed to score AMC in its video product. (I met one resident who passed on Fiber because he loved "Breaking Bad.")
The Time Warner Cable Sports Channel also has exclusive rights to several Kansas college basketball games, a big draw in the region.
Carlos Kirjner, managing director of Alliance Bernstein, has tracked Fiber more closely than anyone else on Wall Street. He’s a bull on the business and its plans for a broad, profitable national rollout, estimating that it could connect up to 25 million homes in the next five years. But it’s a bull case built around its internet service, not cable. "Their hope of creating value likely rests on the broadband offer," Kirjner wrote in an email.
That’s the direction consumers are heading. Broadband subscriptions are on the rise — the top cable and telecom providers added 3.1 million connections last year, according to the Leichtman Research Group. But video subscriptions are either flat or falling.
Meanwhile, the floodgates are cracking open on cord cutting or over-the-top cable services (see: Netflix, Hulu, Amazon, YouTube), paired with broadband. This trend suits Fiber nicely.
Rick Usher has a favorite parlor trick. He recently showed it off over dinner with a friend’s family, who doesn’t have Google Fiber. Their teenage daughter was thumbing her iPhone, and Usher asked her to pull up her favorite YouTube clip. He told her to drag the cursor to the clip’s middle. "She literally jumped out of the chair, ‘Oh my God, it didn’t lag!’" he recalled.
Here was the reason for the gigabit revolution, he told me — your internet becomes as reliable as running water. "You’re not worrying about the technology functioning properly," he said.
This idea, that tech fades into the background, is a fixation of Larry Page. Whether his new tech conglomerate can replicate this magic in city after city is the big question for Fiber.
Usher and other Kansas City civic chiefs have another recent parlor trick: Fielding questions from around the country on what it’s like when Google comes through. They say it will probably change your city for the better, though they admit that much of the shine arrives from coming first.
"Charlotte is going to have really awesome internet and they’re going to have some of the same activity we do," Usher said. "But who remembers the twentieth guy who climbed Mount Everest?"
This article originally appeared on Recode.net.