When Google Fiber Abandons Your City as a Failed Experiment

Illustration for article titled When Google Fiber Abandons Your City as a Failed Experiment
Photo: Elena Scotti (Photos: Shutterstock

There’s a glob of stringy, black rubber spilling out of the street at the corner of Speed and Fernwood in Louisville’s Highlands neighborhood. A block away on Rosedale, the same spongy substance that covers most of Google Fiber’s buried lines in the city snakes in and out of the asphalt.


Google Fiber, Alphabet’s gigabit broadband service, started its relationship with Louisville in 2015. After a two-year delay and negotiations over its rollout, the company adopted a novel but cost-effective plan to bring ultrafast internet speeds to Kentucky’s largest city. This February, only 16 months after it turned on its service, Google Fiber announced plans to turn it off, making Louisville, one of only 19 cities to get Google Fiber since its launch in 2010, the first to lose it. Now, the most visible sign of the tech giant’s screw up is lying all over the roads.

“I literally just got done driving around, looking at streets with Google,” said Metro Councilman Brandon Coan. Strands of errant sealant are “everywhere,” he said outside a coffee shop in his district, one of a few where Google Fiber ran lines before its abrupt departure. “I’m confident that Google and the city are going to negotiate a deal … to restore the roads to as good a condition as they were when they got here. Frankly, I think they owe us more than that.”

The damage goes deeper than the roads. In Louisville, a minor league city with major league ambitions, Google Fiber left behind disappointed consumers, unfulfilled promises of transformative economic development, and a slew of embarrassing headlines for a city that allowed the tech giant to use it as a guinea pig.

But Google Fiber got something out of its time here. It learned that nanotrenching—the cost-saving process of burying fiber optic cables just two inches underground—was a bust. “We currently do not have plans that call for 2 inch trenches, our primary specifications are focused on going deeper,” a Google Fiber spokesperson said in an email.

“It is such a shame to think that we wouldn’t be having any of this conversation if they would have dug their little holes two inches deeper,” Coan said.


In October 2017, when Google Fiber started signing up customers in three Louisville neighborhoods, the ISP was a year removed from a “pause” that saw its CEO resign and more than 100 employees laid off. Turns out, disrupting the telecom industry is wildly expensive, full of intractable regulatory challenges, and opposed by powerful incumbent companies.


The plan in Louisville was to revive the effort with what some were calling Google Fiber 2.0. But first, the city had to agree to one major stipulation: Google Fiber wanted to see if it could scale a method for installing fiber cables that was easier to implement, and easier on the bottom line, than hanging lines from utility poles.

The company had buried fiber lines in other cities, but in Louisville it wanted to go shallower. Nanotrenching called for burying cables just two inches underground, rather than the typical six inches. The company also wanted to cover the trenches with epoxy, rather than the typical asphalt mix. In an email, a Google Fiber spokesperson said the company had “already performed limited trials of two-inch trenches in other markets, and believed it was a promising new method of deployment.” In Louisville, Google Fiber wanted to see if two-inch trenches would work at scale, an attempt that the spokesperson described as “ambitious” and a “lean into innovation.”


Local leaders were concerned. The city sent a team to other cities where a company spokesperson said “limited trials” with two-inch trenches had been conducted, and it sought the advice of outside experts.

“What [Google Fiber] told Louisville was that they didn’t think they could make their traditional construction work, but would we be open to letting them use a new construction technique,” said Grace Simrall, Louisville’s Chief of Civic Innovation and Technology. “There were lots of aspects that we asked a lot of questions about, trying to make them aware that we thought this might have challenges, but they, having done parts of Nashville and San Antonio this way, told us they felt confident they could make this work.”


After years of trying to woo Google—an effort that included spending around $400,000 on legal fees to defend an ultimately unused city ordinance that made it easier for Google Fiber to use utility poles—Louisville was in no position to say no.

Ted Smith, Simrall’s predecessor in City Hall, began working to make the city more “fiber friendly” in 2011 and credited three-term Mayor Greg Fischer, a Democrat, with recognizing that “mid-sized American cities were getting left behind.” (Disclosure: My wife is a Democratic State Representative from a part of Louisville that did not receive Google Fiber.)


“We worked very hard for many years to make the case that Louisville was a city that understands the value of cost-effective high-speed broadband,” he said.

It’s little surprise then that Louisville agreed to let Google futz around with its roads. The town of 600,000 is perpetually trying to catch up with the population and business growth in peer cities such as Austin, Nashville, and Raleigh, which are all Google Fiber cities. Louisville finally had a chance to join the cool kids.


The enthusiasm after Louisville accepted Google Fiber’s offer tells the story. Gigabit internet would help make Louisville “a technologically innovative community,” igniting a “transformation” that would result in a “citizenry that’s more skilled, engaged and prosperous” than ever before, local leaders said.

That didn’t happen. By March of 2018, just five months after Google Fiber launched its nanotrenching trial, the method began to fail. Not only was the sealant spilling out onto roads, leaving fiber lines exposed, but the shallow trenches left cables vulnerable during routine road maintenance.


“If anyone understood road technology or tire technology, [they] would know it’s going to be a problem,” said Jim Hayes, president of the Fiber Optics Association and a decades-long veteran of the industry.

The problem, Hayes said, is that tires easily grabbed the epoxy and ripped it out of the ground. The shorter trench walls also left less asphalt for the sealant to adhere to. Typical microtrenching, a “well-accepted method of installing fiber,” involves burying cables at least six inches deep with a groove that’s no more than an inch wide, Hayes said. “I’ve never heard of any problems with microtrenching.”


Last summer Google Fiber pitched a solution to the problem it created. It would rip out the epoxy and fill in the trenches with asphalt. “It was like, ‘Oh, in retrospect, that seems like an obvious way to fill the hole in the road,’” Coan said.

“They bid out the work to replace the sealant with asphalt,” Simrall said. But before the work started, Google Fiber saw customers lose service when the process of repaving roads damaged fiber lines. “So they shifted their attention to addressing that,” she said.


Then, last month, Google Fiber threw up its hands and quit. The company wouldn’t say how many customers it had at that point, only that the number was “very small” compared to other Fiber cities. Last summer, the local Fox TV affiliate found that the company had pulled permits to wire areas covering about 11,000 households, but it’s not clear how many installations were completed.

In a blog post, Google Fiber explained what happened. “[T]rialing” a “different type of construction method” in Louisville had failed, it said. Fixing it would be too expensive, so instead, it was skipping town, and taking the lessons learned elsewhere.


“This was not an easy decision for us,” a Google Fiber spokesperson said in an emailed statement. “We came to Louisville because Google Fiber is working to change an entrenched industry and we saw a chance to bring great service to Louisville’s residents. Naturally there will be unexpected challenges when you do something ambitious, and at scale, for the first time and we underestimated some of those hurdles. We couldn’t be more appreciative of the partnership we’ve had with the City of Louisville, which embraced the need for competition in this industry as part of its larger initiatives around innovation and emerging technology.”

Simrall insists that Google Fiber was “very sincere in their belief they could make it work,” and had planned to use two-inch trenches in all other Google Fiber cities if it did. That’s little consolation to many Louisvillians, who watched one of the world’s richest companies sweep into town with major fanfare, rip up the roads to test an unproven technique, and peace out when it bombed.


“I was shocked. I was disappointed,” said Candace Jaworski, president of the Louisville Digital Association, a non-profit that pushes for innovation in Louisville and campaigned to bring Google Fiber here. Jaworski was looking forward to one day becoming a Google Fiber customer and seeing the change broadband could deliver to Portland, one of the country’s poorest neighborhoods and one of the first where Google Fiber was available.

“Google Fiber getting in and getting folks into the internet out in the West End was going to be huge with catching up that part of town with the rest,” Jaworski said. “Talk about dangling a carrot and not ever being able to get it.”


Ben Carter was lucky enough to get bite of that carrot. An attorney in the Highlands, Carter said Google Fiber provided wildly fast internet and “rock solid” customer service for the time he had it. When he got the news that Google Fiber was ending its operations in Louisville, he was disappointed, and not just because he would lose the blazing internet speed that he enjoyed primarily for its “Tim Allen, way-more-power-than-you-need” quality.

“It seemed like Louisville was batting above its weight a bit in terms of what cities were getting Google Fiber,” he said of the “coup” that the city pulled off in landing Google Fiber. “Louisville isn’t often on those lists.”


Now that Google Fiber is gone, Councilman Coan is worried that Louisville’s reputation will take a hit. “If you’re not paying attention, you’d say, ‘Oh, well. Louisville is just not a good enough city for Google Fiber.’ That’s not true.”

In the end, it may be that Google Fiber was not good enough for Louisville. Alphabet seems to be scaling back the ambitions of Google Fiber as wireless broadband technology advances and the limitations of fighting entrenched telecom giants become clear, said Matt Wood, the vice president of policy at broadband advocacy group Free Press. Its experimentation in Louisville is evidence enough that Google Fiber was looking for a new way forward. But like so many of its “other bets”—the heading under which Alphabet groups ambitious projects such as its life sciences play Verily, self-driving car company Waymo, and Access, the business unit that operates Google Fiber—the company seems to have lost this one.


That’s not to say Google Fiber isn’t responsible for major changes in broadband access. Some cities, including Nashville, the object of Louisville’s envy, have seen access to gigabit internet soar since Google Fiber arrived. Incumbent ISPs such as AT&T and Comcast have accelerated their fiber operations, including in Louisville, thanks to the competition from Google. That means more people are online, conducting more searches, and looking at more ads.

“Their revenue comes from the eyeball side,” Wood said. “Getting more people connected has always been for them...a means to an end.”


And if its failure in Louisville is the beginning of the end for Google Fiber, Coan said, “It would make me feel better. It would make Louisville feel better.”

Do you have information you’d like to share with Gizmodo? Email tipbox@gizmodo.com or tip us anonymously through Secure Drop.


Adam K. Raymond is a writer in Louisville, Kentucky.


I think that this is symptomatic of a larger trend. Software development has become such a large part of so many business that a lot of software-heads are being put in charge of non-software projects. Those of us who work on the hardware side have learned a lot of valuable lessons (e.g. accessing right-of-way is complex, time-consuming, and expensive) that a meaningful portion of software folks seem to think they can just “disrupt” around or something. But reality doesn’t bend to the will in quite the way that software does.

Another example is in the software defined network space. You think you know networking until you’re faced with the networking equivalent of a blank sheet of paper, which is a “white box switch”. Then it’s like, “Neighbor discovery? Traffic counters? We didn’t know these were requirements!”

And this certainly isn’t to shit on software developers! Where would we be without them?! But experience moving the heavens and the earth in a CPU doesn’t translate very well to moving actual earth.