The Slow Death of Google Fiber? CEO out, expansion halted, employees fired

Google_fiber_logoI’ve long maintained that Google isn’t in the ISP business for the long haul. I said over four years ago that the odds of your city seeing it were astronomically low. Well, now Google is basically saying the same thing. The CEO announced that they have halted expansion of the network, let go employees in towns where they haven’t build, oh yeah, and he’s leaving for “other opportunities”, maybe they should have counted with synel professionals to reorganize. I’ve had multiple first-hand reports of users in Provo who have been unable to get signed up for unspecified reasons, even after Google said they were coming. So what happened?

Google simply bit off more than they could chew. Investors do not like Google pouring money into something this capital-intensive with an ROI so far out. In every market they have attempted to deploy in, they are hit with constant roadblocks from incumbents, something any sane person with industry knowledge could have foreseen from miles away. The lack of voice and 100Mbps products lead to lower than expected adoption and even the loss of customers in Provo, something that I (among others) warned about immediately. Google eventually added these, but it seems to be too little, too late.

Google’s original promise was to form public-private partnerships with cities. Once they launched in Kansas city, it became clear that the “partnership” was reduced to operating like a standard duopolist while using brand power to extract all kinds of benefits from the city. The same thing happened in Provo when the city took a multi-million dollar bath on a network that was around break even on operating expenses and debt service. The model was “give us everything we ask for because we’re famous and fabulous”. The obvious cherry-picking and red-lining was swept under the rug with promises of “eventual” universal rollout, something that now looks increasingly unlikely.

I don’t mean this to just be a smug “I told you so” post (though I would be lying like Donald Trump if I said I wasn’t taking at least a little glee in having been right for so long as the haters yelled at me). It’s to point out that real broadband improvement starts at home. It means you, your community, your city all working together to improve the outcomes. Google was about as close as I’ve ever seen to a large scale broadband Santa Claus and it appears poised to pratfall on the stage.

My take is that we’re seeing the slow decline of Google Fiber. Cities who have it now should be working on their contingency plans for if (or, more likely, when) Google decides to pull the plug. Cities who were hoping for it (including both those who were and were not in talks with Google) need to move on to a new plan. It could be a true public-private partnership, a full-on municipal network (UTOPIA will still be happy to have you), or a privately-funded user-owner cooperative. What won’t work, be it Google or Comcast or CenturyLink, is hoping that you can just wait your way into better broadband.

Google is botching iProvo, but will anyone investigate?

Google_fiber_logoA lot of people tend to turn off the critical thinking the minute Google comes to town with their Magic Fiber Elixir. I’ve already spilled a lot of digital ink on why I think it’s a bum deal, so I don’t need to rehash that here. What does need to be asked is if their slick online marketing campaigns and brand power are being used to avoid any level of accountability from the cities they make deals with. Right now, I think the answer is a resounding yes.

I’ve heard from multiple sources, some of them very close to Google Fiber, that take rates in Provo are not only well below Google’s expectations but below what Veracity had achieved while they still operated the network. The estimates I hear put take rates in the low-20s whereas Veracity had peaked around 30%. That giant plunge would be almost the entirely of the MDUs taking service from the last time those numbers were available. Part of this is to be expected. Google offered up a “free” tier of service for seven years to anyone who paid a pittance of a connection fee. In student-heavy Provo, it shouldn’t be much of a surprise that this ended up being a very, very popular route to getting online.

I know some people are saying “so what? The network is theirs to make flourish or fail.” That’s not quite true, though. If you look very carefully at the contracts, the city has the right of first refusal to get the network back for the $1 that Google paid to use it. The odds of the city not exercising that right are extremely low. Citizens are still paying for the bonds via a utility fee. There is an immense public interest at stake here should Google decide to pull out of the business any time after the end of their seven-year obligation. When the city handed the network over to Google, it was covering operating expenses and the bond. A 25% hit on subscribers means going right back to propping it up from city coffers.

When the city owned and managed the network, there were monthly reports on subscribers and revenues broken down into segments. Once Broadweave came in (and was later acquired by Veracity), most of the numbers were sealed up as company trade secrets and the only public data was if the payments were being made or not. Veracity was more open that Broadweave about take rates and network challenges, but there was still a lot of data left up to speculation.

Once Google came in, the meager data dried up. I have little confidence that either Google or Provo’s elected officials (looking at you, Mayor Curtis) would give anything resembling a direct answer if asked, assuming they gave any response at all. Even worse, it seems that journalists who proudly proclaim to be the public watchdog aren’t going to even ask those kinds of questions, uncritically reprinting each press release as gospel truth.

If you live, work, or have any personal interest in Provo, you should pushing for answers before the city inherits another financial mess.

Learning the Wrong Lessons from iProvo and Google Fiber: A Rebuttal

I suppose this kind of response to the Google Fiber announcement was inevitable, just as it’s completely false assumptions are. The author gets a number of predictable things wrong.

[A]nd finally Provo will be out of the Telcom business.

No, it won’t. It’s still paying for the bond, it has to shell out another $1.7M to make it happen, and there’s a provision that they get the buy the network back for $1 should Google decide to pull up stakes. Granted, that may never happen, but to say that Provo has washed its hands of the matter is patently false.

[I]t’s not like there was some huge cache of potential customers waiting for join the Net.

That’s a really funny thing to say about a network with a 35% take in spite of having multiple failed private providers and mountains of negative press. Bear in mind that Verizon was thrilled to have a 18% take rate on FIOS after two years. If anything, many people were holding out while they were waiting for a less-tainted provider to be an option.

[F]iber is clearly not “future-proof” as claimed at the time.

And here’s where an inch of knowledge on telecom gets you into trouble when you try wading into the ocean. The fiber itself is fine. Most of the work is in digging trenches, attaching to poles, putting in conduit, and running the lines. The electronics, while vital, are a relatively small portion of the overall network. It’s also worth noting that the 100Mbps electronics put in place almost a decade ago are still providing a service that neither Comcast nor CenturyLink can match or beat. That’s some pretty good longevity on any network equipment.

It’s worth noting that Google is deploying 1Gbps electronics when 10Gbps, 40Gbps, even 100Gbps electronics exist. Would the author slam Google for being behind the curve over that, making the same dubious claims over “future-proof” networks? Of course not. Any network is designed to take advantage of the best you can get for the money now and plan for upgrades in the future. It’s become painfully obvious that neither incumbent has done a particularly good job of doing so.

[W]ireless networking has greatly increased its speed and range, and cellular data has moved from a novelty to a mainstay of most cell phone plans.

Again, Mr. Platt gets in trouble by talking about technical things without any technical knowledge. Wireless almost always depends on fiber backhaul. When you use microwave backhaul like Clearwire and Sprint, you end up introducing a lot of latency into the connection which renders it unsuitable for any real-time application. Wireless also hasn’t come anywhere near catching up to fiber in terms of speed. It’s just barely starting to get to a point where a wireless ISP can offer up 100Mbps speeds.

It’s also comical to cite cellular as an alternative when, again, the speeds can’t match wireline. Even the best LTE connection can barely muster a real-world speed on par with CenturyLink’s oh-so-hard-to-find top-tier ADSL2+ product. That’s only going to drop as more LTE devices get into the hands of consumers. Most of those cellular plans come replete with very low caps, high overage charges, or some kind of throttling or filtering which makes them completely unsuitable for business use and hardly an alternative for residential users.

In the long run, it’s still unclear whether and how long wired internet connections will be relevant.

I think the preceding two paragraphs lay this one to rest. If wireline was dying, why would Verizon have poured billions into it, to the chagrin of investors, especially when they own America’s largest cellular company? Why would Google be pursuing deals in various cities to promote fiber-to-the-home? Why is FTTH so explosively popular in Hong Kong, Seoul, and Tokyo? This is yet another point on which Mr. Platt falls on rhetoric as a substitute for knowledge and comes up lacking. While companies like CenturyLink who lack the will and/or ability to upgrade their wireline networks are dying a slow and painful death, that has everything to do with being a terribly run business, not the relevance of their industry.

Most importantly, Internet service is far outside the essential role of government.

This is a common refrain, and it often comes with a big dose of selective outrage. The telecom sector has been rife with government intervention and cronyism almost since its inception. AT&T was a legally-protected monopoly right up until they were broken up in 1984. The major players in the industry got huge tax breaks in the Telco Act of 1996, the price of which has surpassed $300B. Google, on whom Mr. Platt lavishes praise, has received massive tax and financial benefits from the local governments where they plan to do business. Where is the outrage here? Or is the outrage reserved for when public money isn’t being spent on private enterprise? Sir, your principles ring hollow.

[T]hank you Google for buying out network at the appropriate price of $1.

The over-simplification of what the deal actual is shows yet more layers of gross ignorance. The network had an assessed value of $25M, and much of that had to do with the negative perception created by a string of grossly incompetent private providers (HomeNet, Mstar, Broadweave). For an economics professor, he’s not doing such a great job at following the money.

But this is where Mr. Platt’s true motivation sneaks on out in all of its ugly glory:

Please, let your monthly utility bill stir thoughts about the proper role of government. If this reflection somehow prevents citizens and politicians alike from future misadventures into private enterprise, it just might be worth it.

Translation: I’m glad that you can suffer and the taxpayers can be thrown under the bus in order to prove my ideological points. I’d do it again in a moment.

This kind of attitude is far too prevalent in the discourse about both iProvo and UTOPIA. The idea that making the projects fail in order to make taxpayers suffer so that you can be vindicated on your prediction is abhorrent at best. If you see someone with this attitude trying to get into any kind of government position, you’d do best to run in the other direction. Fast.

It’s sad that there’s no shortage of people who are confident in their lack of knowledge, nor that they spend so much time trying to get their ill-informed opinions into print. Let’s just hope that they become footnotes in the debate rather than carrying any real gravitas.

How likely is it that Google Fiber takes over UTOPIA?

Even before Provo announced that Google Fiber would be taking over iProvo, there’s been a lot of speculation that Google Fiber could potentially take over UTOPIA. UTOPIA made their pitch with 1100 other cities, and I produced my own video explaining why a partnership would be a good deal for both parties. Since then, Google has drastically altered the original terms of the arrangement, throwing both open access and municipal involvement under the bus (unless you consider providing tons on concessions “involvement”). Despite Google and UTOPIA being in talks shortly after the RFI closed, I don’t think we’re likely to see any kind of takeover or partnership between the two unless there’s concessions from one or both sides.

I think the largest sticking point is going to be open access. UTOPIA has repeatedly stated that their goal is to offer a world-class infrastructure that any private company can use to provide services. It’s not just enough to provide a third pipe; the network must allow you to pick between companies that differentiate themselves on what they can do with it. This kind of competitive environment has been proven to drive innovation and lower consumer prices across the board. While I have no doubt that Google’s network will do both, it runs the risk of being so superior to existing options as to become a new monopoly.

Google’s focus has been on residential customers. While this is a critical segment for service, there’s almost no mention of business needs. Yes, small businesses and startups can probably do just fine using the same tier as residences, but many need more than that. UTOPIA’s biggest customer uses 20Gbps of bandwidth. If all of their users on the network were on gigabit, they’d need even more to keep up with all of them. The lack of focus on the business end of the pipe degrades the value of the gigabit connections for residential customers. UTOPIA has a complete end-to-end vision; Google does not.

Another problem is the financial terms of any arrangement between the two. Google got a very sweet deal on iProvo, effectively a perpetual free lease with a $1 security deposit. The city is still going to have to pay off the debt on the asset themselves. In exchange, they’re hoping that improvements done to the network will improve it enough that it will be an economic net positive. I think the city could have negotiated a better deal and Google would still be doing well on it financially. I have no doubt that the UTOPIA cities, who are much more eager to pay off the bond, would hold fast for better terms.

If the numbers from iProvo translate to UTOPIA, Google would have to spend somewhere in the range of $40-50M to connect houses currently passed by the fiber and upgrade them to gigabit. This doesn’t include building fiber rings to areas not yet covered. That could easily add another $150-200M to the tab. Should they manage decent take rates (35% of customers paying for service split evenly between Internet-only and double play, and another 35% taking the freebie service), they’d earn $78.1M per year on expenses (including the bond) of about $36M per year. At the high end of finishing the network, it would take them almost six years to break even.

Given that Google seems to be aiming for seven-year commitments, that might be a price they’re unwilling to pay. The cities would have to make some kind of concession to sweeten the pot, and it would likely include tossing existing providers off the network and covering at least some portion of the bond debt. These actions would cause a decent amount of backlash both from residents as a whole and the power user subscribers who have been evangelizing the network for years. If Google’s goal truly is to increase broadband penetration, I’d like to think they would accept any offer that doesn’t make them lose a small fortune.

Google Fiber: The best deal Provo deserves, but not the best deal it can get

Yep, that’s my final take on Google Fiber taking over iProvo: even though you can, you’re not willing to do any better, so go ahead and take the deal.

Google is effectively getting a free lease on the network for a $1 security deposit. Yes, legally, it’s getting “sold” for $1, but Provo has the right to buy it back for the purchase price if Google either doesn’t meet its service and upgrade obligations or decides to stop providing service. That may sound like a decent deal with Google pumping $18M in upgrades into it, but Provo doesn’t have the best track record with getting ownership back, do they?

Meanwhile, Provo is left holding the debt and paying $39.6M over the next twelve years. The city seems to value the network asset at just $25M, and other offers to buy the network were low-balled at $10M. Why has so much value disappeared? It’s because of a string of poor choices with service providers. First there was HomeNet. Then Mstar. Then Broadweave. By the time Veracity came to the table, they didn’t have quite enough oomph to overcome the immeasurable brand damage done by their predecessors. A network that used a $40M bond and who knows how much in federal grants  (which is what built the initial fiber rings) has managed to lose value simply based on perception.

I think Provo can do better. My back-of-the-napkin math is that Google picks up 70% of the total subscribers in town. Half of those are likely to use the free service and pay in one-time revenues of $367K, barely anything in the big picture. Of the remaining half, I’d bet they’ll be evenly split between Internet-only and double play for an average monthly ARPU of $95 each. That works out to about $14M per year in revenues.

On the expenses side, the upgrades are costing Google about $2.57M per year over the seven-year commitment. If Google were to assume the bond payments (and let’s assume they end as soon as they leave), they would still have $8.13M annually to cover expenses related to network operations. This doesn’t even include the revenues that Google is likely to get from TV ad revenue. In short, Google could both assume the bond and do well financially.

That said, I don’t think the council is going to try pushing for those terms, even if they are a win-win. Just as with most things in Provo politics, the die has been cast by the executive and council approval is merely a formality. The short public review period of under a week makes it perfectly clear that questioning the deal is not welcome. I think this deal is a neutral as it currently stands, but I wouldn’t be surprised if there’s a bit of buyer’s remorse down the road.

I was wrong about Google Fiber, but this isn’t a golden deal for the city either

After more than a few false rumors concerning Google Fiber in Utah, I dismissed the latest one as being as equally untrue. Boy was I wrong. Like so wrong that I probably can’t express it myself. (That’s okay; plenty of you are doing it for me, some more kindly than others.) I seriously overestimated Provo’s aversion to risk, and it appears they are putting a lot on the line to make this deal work with Google. I’m not convinced that the city is getting a very good deal. I’m also worried that these important details are getting ignored in your excitement to have a big name like Google running what has been a heavily abused asset. (Seriously, you folks who put up with Mstar? My condolences.)

The basic terms of the deal is that Google gets to take over iProvo for $1, effectively giving the $50M network away. The city maintains bond payments of $3.3M per year for the next twelve years as Google will not be assuming the debt. The city is effectively giving them $90M to take over the network in a no-bid process. Google can also walk away when the seven year deal is finished, leaving the city with none of the assets and five more years of bond payments. Google has to spend a scant $18M to provide gigabit services to 35K households. If they get a take rate of just 20% for Internet-only (a ridiculously low-ball estimate), they’ve made their money back in only three years. You can bet that more than a few other companies and providers would have liked to get that deal. For all of the accusations I made of giving the network away in prior arrangements, this one makes them look fair by comparison.

In order for this to break even on the rather immense subsidy being granted, the city will need to earn an average of $12.9M per year in new tax revenue for each of the seven years. Based on the city’s 2012 filing with the State Auditor’s Office, this would equate to an almost 35% increase in total tax revenues. Given that the amount of land left to be developed in the city is relatively small, I can’t see a way in which the city actually accomplishes this. Google is making easy money at the city’s expense.

So what are you getting for your money? For a one-time fee of $30 (assuming you already have the connection in your home), you can get 5Mbps service for seven years, the length of Google’s commitment. (The standard packages of $70/mo for gigabit Internet or $120/mo with TV will be available.) City facilities get gigabit for free for the seven-year period. Each household will effectively be paying $368.57 per year in bond payments and the loss of the asset, so jack those prices up about $30.71 per month.

Is it a good deal? That remains to be seen. A lot of you are rightly excited to be getting the kinds of speeds that UTOPIA customers have been enjoying for almost a year now. But this isn’t a bailout. This isn’t a free lunch. You’re paying a pretty goodly amount for this arrangement, and you should be asking if it’s worth it. I know I am.

No, Google Fiber Is Not Coming to Provo

Update: Yep, I totally missed this one, the one time out of dozens that it actually proves to be true. Glad the troll commenters are having a good time with it.

Provo (or, more specifically, Mayor John Curtis’ office) has been hyping up an “epic announcement” for weeks now. Somewhere in the speculation came a number of rumors that Google Fiber would be swooping in to either take over or supplant iProvo. Quite frankly, that’s an absolute load of malarkey. Nothing in the history of either Provo’s handling of its fiber optic assets or Google’s launch of Google Fiber cities would suggest anything like it.

For starters, we have to consider the way in which Google announces new fiber cities. Its history has been to send out a press release at least a few days out to generate buzz and send invitations to the tech and national press. This is not a company that announced these kinds of projects without a serious amount of fanfare.

Google also has a history of liking to roll its own solutions. One of the ways it got so big so fast was to create servers based on their own custom specs, not buying expensive off-the-self servers. Google Fiber is no different. They seem to relish the idea of starting from scratch and making something uniquely theirs. iProvo would require extensive upgrades to support 1Gbps connections and an entire head-end replacement to support modern MPEG-4 video.

There’s also the point of Google’s tacit support for municipal networks. Swooping in to take over a network would undermine their unspoken support for poking incumbent operators in the eye. Their goal has always been to shame ISPs into providing better service, even if that means throwing them under the bus in favor of municipal options. Taking over a network doesn’t match up with that at all.

The kicker, though, is Google’s terms for setting up shop. So far, Google has set it up so that they can walk away from their network if they decide it’s not for them. That probably sits just fine in Kansas City and Austin where they aren’t really on the hook for much. But Provo? They’ve spent half a decade trying to run away from iProvo. Any deal that doesn’t involve them washing their hands of the network would be anathema to every single action the city has undertaken under two different mayors and a lot of different council members.

Make no bones about it: Provo is not getting the bailout it so desperately desires. Hyping up these rumors does nothing to actually fix the situation, but it does distract from real looming problems like Veracity’s lease of the network expiring in two months. Instead of trying to find a white knight savior, it’s time for Provoans to demand that their elected officials address the elephant in the room instead of continuing to punt.

PS I’ve already told you to stop daydreaming and build the same kind of network that Google is. What are you waiting for?

Legislation Alert: HB273

Some years, muni telecom gets a break. This year isn’t one of them. Rep. Keith Grover has introduced HB273 which would effectively ban Provo from using utility surcharges as a way to cover iProvo debt payment shortfalls. It’s unclear as to if this would have any effect on UTOPIA, but the bill does include some vague language concerning charging “just and reasonable” rates. Depending on how the legislation is implemented, this could prevent UTOPIA from differentiating wholesale prices depending on volume or require that they match rates with other networks.

Any way you slice it, this is a bill very explicitly targeted at removing operational flexibility from municipal networks. I highly recommend you contact Rep. Grover to urge him to reconsider.

iProvo to go back to the city, Veracity will lease for 14 months

In an unsurprising move, Veracity Networks will transfer ownership of the iProvo network back to the city of Provo as it has been unable to generate enough revenue to make the purchase feasible. Provo has already instituted a monthly utility charge to all residents and businesses in the city to pay for the current debt load on the network. Veracity will continue to operate the network under a 14-month lease agreement, paying Provo $95,000 per month plus a cut of revenue from future subscribers. Interestingly, all customer revenues are going into an escrow account that will pay the city before Veracity gets its share and the city will be responsible for all network infrastructure costs including installations.

The agreement has been framed as a chance for Provo to figure out what it wants to do, and there doesn’t appear to be anything ruled out. I wouldn’t be surprised to see new providers added to the network, or for the city to seek out new partnerships, either with another private company or a public entity like UTOPIA. The final vote will be on March 20.

UTOPIA and iProvo as Campaign Issues

The Deseret News brought up that UTOPIA and iProvo are proving to be campaign issues in the upcoming municipal elections. (h/t: Brian Merrell) The article itself provides a good background on the financial details of the networks (and an added bonus that the Utah Taxpayers Association is little more than a lackey for Comcast and CenturyLink), but not much in the way of where candidates stand. Personally, I’ve found them to fall into one of three categories.

The first is the obvious municipal network supporter. They’re in favor of the network and are willing to do what it takes to make it successful. Long-time boosters like Dan Snarr and JoAnn Seghini fall into this category. You’ll find them to be few and far between because of the amount of flack so many of them catch.

The second is the opponent who wants to acknowledge reality. They don’t think joining the network was a good idea, but they know the reality is that the decision has been made and they have to make the best of it. In some ways, they are the best option since they won’t pull punches when something isn’t working the way they think it should. John Curtis is a good example.

The last is the opponent who has plenty of complaints, but nothing in the way of solutions. They’ll rant all day long about how the network was a terrible decision, it should have never been done, and no good can come of it. They completely fail to propose any real solutions, and the “just sell it” attitude pays no heed to making the city take a financial bath in order to prove their ideological point. These people are a reckless danger to any city they govern since they are willing to make you, the taxpayer, pay dearly in order to kill off their non-favored programs. George McEwan is a prime example of this kind of low-information irresponsibility (and he was thankfully eliminated from the election for failure to file financial disclosures).

So as you head to the polls in your city, ask yourself which of these candidates is going to make good decisions for the city. Ask yourself who each one of the candidates fits from your list. You’ll know what to do.