This startup could replace your cable internet, but faces hurdles

Rob Pegoraro · Contributing Editor

Mon, Dec 18, 2017, 11:47 AM

Starry wants to take on the likes of Comcast and Verizon with its own DIY network transmitter.

BOSTON — For most Americans, getting broadband internet at home requires signing up with one of two corporations they may not like much: the phone company or the cable company.

This lack of choice has persisted through years of failed launches of third options. Powerline broadband looked promising in the mid 2000s before flopping in field trials, and “fixed wireless” systems like Clearwire’s WiMax 4G service was either too slow or couldn’t support unlimited use.

A new wireless ISP called Starry aims to end that losing streak with a $50/month service that offers downloads and uploads at 200 megabits per second without data caps.

But while it seems to be off to a sound start here, the company will need to deal with obstacles ranging from simple foliage, to marketing to historic-review approvals before more people can connect via its airwaves.

Distance, not density

Starry delivers broadband over millimeter-wave frequencies. Other companies have done that, but Starry’s network relies on fewer but more precise transmitters covering longer distances.

“You need point to multipoint,” said CEO Chet Kanojia in a meeting at Starry’s downtown offices in early December. “That’s the chief differentiator.”

The roughly 80-pound transmitters that Starry places on cellular-network towers target those beams with electronically-steered antennas, as explained in a February report provided to the Federal Communications Commission. This allows a range of up to a mile and a half in cities — without needing the usual direct line of sight between antennas.

It also frees the company from having to hang transmitters on utility poles or street lights, which often represents a massive obstacle to broadband deployment.

Starry has eased its job further by focusing on multiple-dwelling buildings: It puts a receiver on the roof, then the owner connects that to existing network wiring to give residents a new amenity. Kanojia said the firm now has 240,000 residences eligible for service.

One early subscriber praised Starry in an email conversation.

One of the 80-pound transmitters Starry places on cell towers.

“The service has worked great for us!,” said Terry Xu, a software developer in Cambridge. He sent screenshots of speed tests showing downloads and uploads consistently exceeding 200 Mbps, sometimes 300 Mbps. Ping times were at worst 16 milliseconds, easily responsive enough for everyday use.

Aside from “a few hours a month max” of planned maintenance, always with advance notice, Xu said he hasn’t noticed any downtime.

Regulations and other risks

Beyond expanding outside Boston — expect news about other markets early next year — Starry also plans to reach single-family homes. During my visit, Kanojia showed off the antenna it’s designed to fit in the bottom of a window; the antenna sits outside in a horizontal tube, with the rest of the electronics in a pod on the inside.

Reaching individual houses will subject Starry to an adversary most ISPs don’t face: plants.

“Minimizing exposure to foliage matters,” Kanojia said. The signal can get through one or two trees, but not a stand of them.

Two telecom analysts called out risks to Starry’s continued buildout.

“I think it’s great in terms of the product idea,” said Alex Besen, CEO of the Besen Group and author of a study of Starry based on publicly-available information. But he worried that beyond-line-of-sight reception won’t meet the company’s plans.

Benoît Felten, CEO of Diffraction Analysis, cast doubt on Starry’s hope to move to customer-installed antennas: “All of the existing examples of fixed wireless (the commercially viable models) rely on precision installs.”

Starry’s existing installations have already seen delays: Xu had to wait three months for his building to get online. Spokeswoman Virginia Lam said issues like historic reviews can prolong building deployments, but most take about eight weeks.

Kanojia said streamlining those municipal regulations would ease Starry’s work. What about the set of net-neutrality rules that the FCC aims to delete Thursday?

“That hasn’t bothered us,” he said, pointing to Starry’s absence of data caps.

Few tech founders know the risks of government power better than Kanojia: The Supreme Court killed his TV-via-internet startup Aereo when it held that it violated the spirit, if not the letter, of copyright law.

Competing with incumbents

Starry has benefited from patchy competition. Verizon (VZ) only began bringing its Fios fiber-optic service to Boston last year, leaving Xu (already a cord cutter) looking at a Comcast (CMCSA) connection with slower uploads or a poky Verizon digital subscriber line.

(Verizon’s media division, Oath, includes Yahoo Finance.)

Roger Entner, lead at Recon Analytics, defined Starry’s ideal rival as phone-based DSL. “That’s your only option?” he said. “You’re going to do this in a heartbeat.”

Entner also warned that exclusive marketing deals between telcos and building owners could hold back Starry. In a July filing with the FCC, the firm asked the commission to prohibit those marketing lock-ins.

But those incumbents won’t stay pat forever. Comcast is moving to upgrade its upload speeds, while Verizon’s Boston Fios buildout should finish by 2021.

“A third player there is going to struggle to carve itself a niche from zero brand recognition,” said Felten.

Kanojia said its infrastructure-light business model will allow it to break even with very few sign-ups among eligible customers — “probably 3 and a half, 4 percent.”

Eventually, 5G wireless will let Big Telecom compete not just over wires but over the air for your broadband business. But how many people who are tired of the cable-phone duopoly will want to wait until 2019 or later for 5G to reach their home?

That’s the opportunity for Starry — and other companies that can get fixed wireless in gear sooner.

More from Rob: