Tag Archives: seattle

Google Fiber goes boringly conventional in Seattle

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

At first it tried to disrupt the broadband industry in the U.S. with full scale fiber to the home deployments, but the financial realities of a capital intensive business with a long term return on investment horizon has forced Google Fiber into a traditional small ISP business model. Its latest move – into a high rent Seattle high rise – is a low risk venture. According to a blog post by its Webpass subsidiary

Today, we announced that Webpass is ready to move into the Emerald City, one Ethernet-wired building at a time.

Webpass provides blazing-fast Internet (up to a gigabit per second!) to residential and business customers, starting with Fifteen Twenty-One Second Avenue, a 40-story luxury tower located above Pike Place Market. We expect to add many more buildings throughout the city, and, starting today, residents of other apartment and condo buildings can reach out to express interest in bringing Webpass to their home.

The only news here is that Google is involved – downtown Seattle already has a thriving market multi-dwelling unit (MDU) gigabit service.

Two questions that the brief announcement doesn’t answer are how is it getting enough bandwidth into the building and what is the business arrangement with the landlord?

Media reports about the Seattle initiative assume that, consistent with Webpass’ market positioning (but not its invariable practice), the building will be fed wirelessly. That seems unlikely: reliably delivering a gigabit (for $60 per month) to dozens of units in a dense urban environment via a radio link is difficult, while leasing dark fiber from the City of Seattle or other providers is easy. I don’t know how they’re doing it, but I’m not making any assumptions either.

Typically, MDU deals between Internet service providers and landlords involve some level of exclusivity, often based on access to the Ethernet or other wiring inside. It’s becoming a controversial practice. The Federal Communications Commission is about to look into it, and Webpass was on the other side of it in San Francisco. It would ironic if Google’s broadband business model goes from being the disrupter to being disrupted.

Comcast offers Seattle the Philly weasel, er, deal

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

After months of fighting, Comcast and the City of Philadelphia reached agreement on a new 15 year cable franchise agreement that included a few spiffs, like expanded eligibility for the low income, $10 a month Internet Essentials program. The announcement came on the eve of a city council vote in Seattle, that would have approved a less generous deal. So, Seattle balked and asked for the same terms as given to Philadelphia. Surprisingly rapidly, Comcast and Seattle negotiators agreed on a few deal sweeteners, including the same IE eligibility upgrade.

All was fine until it came to a council vote and, as a story in GeekWire tells it, a couple of council members read the fine print…

Nick Licata and Tim Burgess expressed worry that the newly-added benefits are not part of the actual franchise agreement, but rather included in a separate letter. “There is some risk that, without consideration, Comcast could change these benefits,” [Councilmember Bruce] Harrell admitted. Given this, the new benefits cannot be enforced by the city, unlike the mandated requirements in the franchise agreement. Even though the city’s law department approved this method, several councilmembers weren’t comfortable approving the new agreement without something more enforceable. The council pushed back the vote to next Monday, Dec. 14.

I’ll take the risk of offering some free advice to Seattle: put it in the contract. In big, bold, capital letters. Otherwise, Comcast will ignore it at its earliest opportunity.

Even if Comcast ran it like an honest promotion, Internet Essentials is not a permanent program. It came out of the purchase of NBC/Universal, but the hard requirement expired a couple of years ago. Comcast extended the offer in an effort to generate good will while it unsuccessfully tried to buy Time Warner and swap markets with Charter, but it could, and likely will, pull it the minute it thinks it doesn’t need the political cover.

Put it in the contract.

To nimby or not to nimby is the dilemma for Seattle and Portland broadband upgrades

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Not every emerald city has a wizard to rely on.

Broadband doesn’t arrive by magic. It needs stuff. Like poles and towers and boxes that don’t necessarily match the neighborhood decor. That simple fact is often lost on nimby homeowners who want to be able to watch four channels of Netflix HD movies at once, but don’t want a small, green box planted anywhere nearby.

Seattle and Portland are two cities where it’s difficult, if not impossible, to install telecoms street furniture. But that might be changing.

Following the midnight exit of Gigabit Seattle, the city’s new mayor wants to make it easier for real companies to upgrade infrastructure. So he’s proposing to end a requirement that broadband providers in general, and CenturyLink in particular, get permission from homeowners before installing boxes in nearby public right of ways. The need to go through hundreds of separate negotiations has effectively blocked broadband improvements. Mayor Ed Murray plans to ask the city council to roll back the individual veto it gave homeowners in 2009, a move CenturyLink claims scuppered 60 upgrade projects affecting 21,000 homes.

Portland is one of the 34 cities blessed by Google and allowed to compete to be on its fiber-to-the-home list. To stay in the running, cities have been told to hand in a fiber-ready checklist by next month. One of the items – easy access to public right of ways for equipment cabinets – is causing consternation in Portland, which currently doesn’t allow them (h/t to Karl Bode at DSL Reports for the pointer). But the city council will consider changing its policy and, perhaps, adopting Google’s standard terms, as San Antonio did.

Public policy debates in San Francisco and Overland Park, Kansas can easily be hijacked by loud, self-absorbed interests. Not coincidentally, those two cities have sunk to the bottom of the Google Fiber priority list, and other broadband upgrade projects have been stalled. Seattle and Portland, though, still have a choice.

Illinois says Gigabit Squared lied repeatedly, wants $2 million back

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Once upon a time, it was strictly formal dress for sunrise.

The company that sold magically cheap fiber and a business case built on fairy dust to Seattle, then left town owing fifty grand is in even bigger trouble in Chicago. The state of Illinois gave Gigabit Squared a $2 million grant to deploy “ultra high speed” Internet access on the city’s south side and, to say the least, isn’t seeing results, according to a story in the Chicago Sun-Times (h/t to the Baller-Herbst list for the pointer)…

Gigabit Squared, a Cincinnati-based company that last May touted the high-speed project in nine South Side communities, “has lied repeatedly” about its intentions and may have spent only $250,000 of the grant money for legitimate purposes, said David Roeder, spokesman for the Illinois Department of Commerce and Economic Opportunity, which issued the grant.

To clear the air the state is inviting Gigabit Squared to an “informal hearing”, which in Chicago-speak is the semantic equivalent of a casual firing squad. The company has until 10 April to RSVP. So far, no direct reply, but the Sun-Times article includes a statement from Gigabit Squared that’s a wonderful mix of disingenuity and bewilderment. Translation: the dog ate the paperwork.

I’m not close enough to either the Seattle or the Chicago project to speculate on whether the failures were the product of ignorant and incompetent management or premeditated fraud or something else. Whatever the reason, the result is a couple of bloody black eyes for legitimate municipal broadband advocates, which is not helpful to our cause. My advice to cities approached by broadband rainmakers beating a drum: if it sounds too good to be true, it is.

Muni broadband wins voters’ hearts in Colorado but not Seattle

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Filling up in Colorado, half empty and still deflating in Seattle.

Seattle voters yesterday turfed out mayor Mike McGinn, their broadband cheerleader-in-chief, giving the job to state senator Ed Murray by a preliminary margin of 56% to 43%. It was a different story in the Rocky Mountain town of Longmont, where residents overwhelmingly approved a fiber-to-the-home bond measure, 68% to 32%.

FTTH was a prominent McGinn campaign promise, both this year and in 2009, when he was elected mayor. The Seattle Times ripped him back in July for not delivering, while a McGinn-backed group – Gigabit Seattle – talked up its plans to build fiber throughout the city. And talk is pretty much all it’s done. Nearly a year since the project was launched, Gigabit Seattle hasn’t raised the money it needs even for its pilot project, which isn’t surprising given its faith-based financing model.

Murray, on the other hand, downplayed broadband during his campaign, which was enough to earn him at least $12,000 in donations from Comcast, an incumbent service provider in Seattle and a bare-knuckle political brawler when its business interests are threatened. There were bigger issues in the Seattle mayoral race, but McGinn’s response to his broadband failures was to turn up the hype rather than turn on the work, making it an easily grasped symbol of his term in office.

By contrast, the muni FTTH campaign in Longmont was poker-faced. The city showed just a couple of cards on the table, giving voters a summary powerpoint presentation and a promotional brochure rather than a bona fide business plan. They’ll get to see it after the fact, though. To sell the bonds, the city is required to publish detailed disclosures about its new broadband business and its existing electric utility. Both are important because if FTTH cash flow falls short, bond payment charges will be tacked on to residents’ electric bills.

The lop-sided result in Longmont shows that whether or not they completely understand the risk, voters enthusiastically accept it. Because they’re putting real money on the table, it’s very likely an FTTH system will be built. That’s not what I expect for Gigabit Seattle, which relies on hope and hype rather than cash.

Slow broadband a drag on Seattle mayor’s re-election campaign

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

I’ll have what she’s having.

Seattle mayor Mike McGinn is running for re-election and the editorial page of the Seattle Times, which has never particularly cared for him, is homing in on his failure to build fiber to every home and business in the city…

With a campaign pledge of broadband Internet for all, Mike McGinn promised big, delivered small, and hopes voters won’t notice the difference.

KUOW-FM, Seattle’s University of Washington-owned NPR powerhouse, reached a similar conclusion, although in a better researched and more nuanced way

When Mike McGinn ran for mayor in 2009, he campaigned on the promise of high-speed internet for all of Seattle. But once elected, he struggled to implement anything close to that. Four years later McGinn still presides over a city of internet haves and have-nots.

The article goes on to talk about what is working and what isn’t. Comcast gets a nod for upgrading at least some of the region to the theoretical 105 Mbps max it offers in other markets. CenturyLink hasn’t done as much, a problem it blames on neighborhood opposition to equipment cabinets on sidewalks.

The real “haves” in the KUOW story are the residents of more than fifty buildings that CondoInternet, a local company, serves. Its business model looks pretty straightforward: plumb a sufficiently large or affluent property with ethernet and hook it up to fiber or wireless backhaul, something its parent company, Spectrum Networks, also does for commercial customers in the Seattle area.

Speeds promised range from 100 Mbps ($60/month) to a gigabit ($120). No performance tests were mentioned, but neither were any complaints. It’s an apparently successful example of the fiber-to-the-basement business model I looked at in a study for the City of Palo Alto a couple of years ago.

McGinn’s latest broadband initiative, Gigabit Seattle, isn’t doing much besides turning up the volume on its marketing machine. All the “have nots” can do is hope it’s designed to outlive Seattle’s campaign season.

Gigabit Seattle raising FTTH attention but not cash

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Adding lift to a trial balloon.

The Gigabit Seattle team is trying to tap into Google Fiber’s buzz by releasing a fiber-to-the-home pricing plan that sounds a lot like what’s on offer in Kansas City, albeit for a few dollars more and with a little less freebie time. Otherwise, there’s been precious little in the way of specific information about the project since it was announced six months ago.

What I wrote then is true today: Gigabit Seattle’s financial vehicle is still a concept car. Zero private sector investors or lenders have been announced, and actual public sector contributions are minuscule.

Service is supposed to begin somewhere in Seattle “in early 2014”. The roadmap outlined in December had the project starting out in a dozen demonstration neighborhoods. No particular construction timetable has been set, even though engineering work was supposed to be well along by now. The latest announcement said that the project team will let residents know next month how they can sign up. Previously, they said that they’ll prioritise neighborhoods on the basis of pre-commitments, again similar to Google Fiber, with a 15% take rate being mentioned as a threshold for moving ahead in a given area.

It’s also unclear exactly who will be building, owning and operating Gigabit Seattle. The company behind it – Gigabit Squared – now describes itself as a “a digital economic development corporation specializing in the planning, implementation and rollout of IT-enabled infrastructure in core markets”. Not a telecoms company, in other words.

Although Gigabit Squared’s CEO says it will own its own projects, it doesn’t have any track record or significant, visible assets yet. Judging by the few financial details discussed so far, it doesn’t have a firm grasp on how much it costs to build an urban FTTH system and the operational telecoms experience of its principals appears slim.

Gigabit Seattle might be able to evoke Google’s business model in a press release, but it’s still a long way from raising the money to pay for it.

Caveat vendor: the customer can say no

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Thank you for your input.

Organizational budgets and goals are set in the C-suite, defining the resources and limiting the options available to IT executives. Then it’s up to them to find solutions that maximize employees’ chances of meeting those goals while minimizing the pain and staying within the budget.

IT executives have to balance the arts of managing up and implementing down. The best outcome occurs when everyone’s needs, wants and dreams are fulfilled. It’s a tough job that requires a diverse set of skills.

But it’s not the same set of skills that brings success on the consumer side of the Internet service, pay television and telecommunications business. There, every customer sits in his or her own C-suite. Consumers set their own goals, work within their own budgets and determine which solutions suit them best.

That’s why I maintain a healthy degree of skepticism whenever I see FTTH initiatives that are led by executives with blue chip credentials in the IT world but lack a commensurate level of consumer market experience. They are well equipped to engineer effective solutions and evangelize the benefits, but they’re accustomed to having the option of making top down decisions. You don’t close sales on a desktop by desktop basis in corporations or institutions.

The difference between customers and employees is that customers are sovereign. Both can say “no”, but it doesn’t mean the same thing. In corporations and institutions, that’s when decisions are made and the debate ends. In the consumer market, that’s the point when selling begins.

Metro broadband: without the political cards, you’re not playing with a full deck

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

Political value: the need for speed at the San Leandro public library.

There’s an argument to the effect that the prices charged for broadband service by telcos and cable companies in urban areas are higher than necessary to provide that service and make a reasonable profit.

It’s not crazy talk. You can make a case that more densely populated areas have lower per household costs – opex and capex – and that more affluent areas have higher profit margins. There are counter arguments too, not least of which is that telecoms network costs should be spread across all users. Personally, I favor the whole system approach – the more people reached, the more valuable the network – but the marginal cost approach has valid uses.

However, it doesn’t follow that an independent competitor in a metropolitan area will be able to charge less for equivalent service or the same for better service. The telecoms business has huge economies of scale: fixed operating costs are high relative to variable costs and large purchases by big companies bring hefty discounts. Particularly for television programming. A local competitor operates at a significant cost disadvantage.

A significant fraction (30%? 40%?) of households passed have to be willing to pay more ($50 per month more is a good placeholder) to either incentivize an incumbent to bring in fiber or support the operating cost and capital requirements of an independent system. The market research I’ve seen says that’s not happening.

People may value significantly better broadband services highly in many senses of the word, but not economically. At least not to the extent that an independent, privately financed metro scale FTTH overbuild in a competitive market is economically sustainable. Not yet.

Something else has to be on the table for an independent FTTH overbuild to work. Construction and operating subsidies, (significantly) below market rate financing, publicly owned assets are examples. In other words, you’re adding political value to whatever economic value is present in order to make a business case.

Whether the political value exists is a legitimate topic for debate, and some communities or state and federal policy makers might conclude that it is. The California Advanced Services Fund (CASF) is one example of policy-driven broadband investment. Leveraging a public owned electric utility, such as in Chattanooga (FTTH) or Palo Alto (dark fiber), is another. So is partnering up public assets and private investment, as in San Leandro. And there are more. And there are counter-examples too.

Claims made by some that ordinary metro FTTH overbuilds are self sustaining investments with no risk to taxpayers are at best distractions. For now, it is as much a political question as an economic one. Debate should be encouraged.

The problem with FTTH is there’s no problem

FacebookTwitterGoogle+PinterestLinkedInRedditEmail

It’s not about finding a mass market solution. It’s about finding a sufficiently acute mass market problem.

The struggle to develop a general fiber-to-the-home (FTTH) or premises (FTTP) business model for city-wide deployments doesn’t result from a market failure. Quite the contrary. It’s evidence that the laws of supply and demand are in full effect.


Demand, meet supply.

People generally get the broadband service someone else – a business or government agency mostly – is willing to give them for the price they’re willing to pay. FTTH market research tracks closely with actual results. If you ask consumers if they’d like faster broadband, they say yes (who wouldn’t?). But when you test price points, they’re generally pleased with what they’re paying now and don’t perceive enough additional value from higher speeds to motivate them to pay more.

From the point of view of a city or other prospective overbuilder, it’s a competitive market. AT&T, Comcast and the rest do a fair job most days meeting most customer expectations. They leverage that complacency to fiercely defend their turf. Successfully, for the most part.

Cities are good at filling broadband infrastructure gaps where immediate economic demand exists, either directly or by bringing a private partner to the table. Lit San Leandro, Palo Alto’s dark fiber and Mountain View’s WiFi system are good examples. But those are specific solutions in largely unique business circumstances that also suit the particular political character of each city.

There won’t be a market-driven case for FTTH until a sizable fraction of the residents and small business owners in a community have a problem that 1. they’re willing to pay an extra, say, $50 a month to fix, and 2. can’t be solved to their satisfaction by existing technology and service providers.

Adding institutional IT budgets to the kitty is not as helpful as some FTTH backers, such as Gigabit Squared, think. An organization with an IT budget hefty enough to make a difference is really looking for wholesale service. Big IT systems need big pipes and budget accordingly. That’s helpful, maybe decisive, for funding a middle mile project, and there are examples where it’s done the trick.

You need a significant fraction of the available homes and businesses ready to spend more now, to tip the balance for an FTTH business case. Until the economic demand (i.e. marginal willingness to pay) develops, the Gigabit Squared model will only work if it leverages political demand: grants, direct tax money, cross-subsidies from other municipal utilities or other public support, in healthy quantities.