I Go Delphic, Snort Oak Leaves, And Give Four Reasons Why Google Will Bid To Win in the 700 MHz Auction (despite what the smart money says)

Analysts who watch Google and watch the wireless world really, really, really don’t want to think of Google as getting into the wireless biz. This spring, I heard an awful lot of “Google won’t bid” or “Google can’t win” or, my personal favorite, “you think Google is going to bid? Are you on crack?”

As regular readers know, while I have occasionally been a shade grumpy about how Google worshipers have attributed all things in the auction to our Great Google Overlords, I have been surprised at the reluctance of most analysts to accept that Google really does want to win licenses. For example, when Verizon announced it would open its network to third party devices, analysts suggested this would take the pressure off Google to win licenses itself. When Google announced it definitely would participate in the auction, a number of analysts again questioned whether Google was really serious about winning or whether it just wanted to insure the $4.6 billion C Block reserve got met. Although as shown in this article here, some analysts expect Google to press hard to win, the conventional wisdom among analysts has jelled into “Google is only bidding to make sure the C Block conditions stay in place.”

These analysts have sound reasons for thinking Google would be mad to bid. Google never wanted to be a network provider. Sure, they’ve dabbled a bit by investing in broadband over power line (BPL) and dipping a toe in muniwireless (neither of which has amounted to very much). But Google never took the potentially ruinous plunge from being an applications provider (its realm of dominance) to becoming a network provider. Worse, the estimated $5-$6 billion price tag for the C Block licenses is only the beginning of the cost to actually build a network. According to one widely reported estimate by Google itself, it would cost another $12 billion to build the network once Google has the licenses.

Nor is the wireless industry considered ripe for expansion. If anything, analysts expect further consolidation as smaller carriers find it tough sledding against the vertically integrated giants AT&T and Verizon (which jointly control the bulk of residential subscribers, can offer a nice set of wireless and wireline bundles, and enjoy other advantages that make them tough to beat). Even with Google’s genius for creating new capital opportunities, the conventional wisdom goes, how on Earth can Google ever recoup this mammoth investment as yet-another-wireless carrier in the highly-commoditized world of wireless telephony. And the one thing that might have worked, creating its own compelling “walled garden” that encourages users to go with Google wireless to enjoy access to features they routinely access in the wireline world, is the one thing Google has sworn up and down it won’t do. To put icing on the cake, the formation of Android and the inclusion of national carriers T-Mobile and Sprint make it impossible for Google to create its own walled garden if it changes its mind after winning.

With all this to consider, small wonder analysts by and large don’t see much chance of Google making a serious run to win. They believe that Google wants someone else to win, but offer an open network Google can ride on. So while bidding to make sure the spectrum gets bought makes sense, actually wanting to win the licenses doesn’t. Hence the convergence of the conventional wisdom that Google will leave it to Verizon or someone else rather than tie a multi-billion dollar albatros around its neck and potentially crash its stock valuation (especially if you hold Google stock).

For the reasons given below, I will play the contrarian. I think Google will bid and fight hard to win licenses. Indeed, while I expect Google to target C Block, it may well go after D Block or some of the other licenses as well, if that’s what it takes to build a national footprint. Google might still get outb id by Verizon and other carriers, but I don’t think that’s Google’s plan. I think they are in to win.

Why? See below . . . . .

First, we will assume that I am the Great Internet Delphic Oracle and, by the process of burning oak leaves and sniffing the fumes, have peered deep into the heart of Google and discerned their true desires. (Yes, I’m aware the Delphic Oracle was dedicated to Apollo and probably relied on hallucinogenic gases released from rock fissures and that it was the Oracle of Dodoni that listened to oak leaves to hear the prophecies of Zeus. But I’m short on volcanic gases and up to my knees in oak leaves at this time of year. If you think that invalidates my prophecy, you don’t need to keep reading.)

[*OAK LEAF SNORK* Man, that is gooooood oak leaf]

So here goes. Let us begin by observing that CTIA, Verizon, AT&T and the rest of its members wrote the epitaph of the traditional wireless industry as it has existed from 1994 until today when it told Google “Well, if you want to get on a wireless network, then why don’t you bid for licenses and build one yourself.” At the time, this seemed like a pretty witty response, since everyone knew that a combination of manipulation of the bidding rules as they existed until now and other advantages enjoyed by incumbents against possible new entrants made it virtually impossible for Google to acquire licenses. But, as we shall see, the bidding rules have changed and Google does not plan to play by the traditional wireless rules.

1. Google Has A Different Vision For the Wireless World It Can Only Achieve By Owning Licenses.

OK, the biggest problem is that analysts are looking at this as if Google wanted to break into the wireless phone business and introduce the fabled “G-Phone” similar to the Apple iPhone. But that’s not what Google (and the rest of Silicon Valley) want. What Google really wants is far more audacious. Google wants to eliminate the entire wireless “phone” industry and replace it with the “mobile broadband” industry. In this world, people do not buy “mobile phone service” with the option to load all manner of various features for additional prices onto their phones. People buy a wireless service contract for a “dumb pipe” similar to what they buy (now) from cable and dsl companies.

Needless to say, the existing wireless carriers — to the extent they can even contemplate such a thing — regard Google’s vision with unbridled abhorrence. But Google has one big advantage — everybody else wants the same world Google does, they just don’t know it yet. That may sound absurd, elitist, patronizing, etc. But the fact is that most people don’t realize what they want until someone with entrepreneurial vision thinks it up and sells it to them. In 1990, when I tried to explain to people why I needed email now that I had graduated college and would gladly pay for a flat rate service rather than the 25 cents an email GENIE or COMPUSERVE wanted to charge, they did not get it. “Why on Earth would anyone ever want that?” Was the most common response. In 1993, when I started at the Department of Energy, all new hires wanted to know if we could get email access. “Why on Earth would anyone want that?” Was the initial response. Until people got it, then discovered they wanted it at home as well. So I believe that, given the option, people will discover they really want an open wireless network. But they have to have the option available to them first, and the only way Google can do that is by winning licenses.

In the U.S. until now, the mobile phone has been first and foremost a phone. Over the years, phone companies working with equipment manufacturers have attached various services and gadgets, like cameras. But it is still primarily a phone and the contract I always buy is a phone contract. I get internet or text messaging or anything else as an addition to my basic phone contract. And we accept this, because it is the way things always have been and no one is offering any different.

Compare this to the dsl and cable modem experience. You get a basic contract and you can do anything you want. Want to plug in voice over IP? OK, you can do that. Want to download a song or video from iTunes? It works just the same way. And, most importantly, if want to use Google search, or Google Blogs, or YouTube or Google Calendar or anything else Google and have Google serve up GoogleAds and track your Google activity all over the GoogleWeb, you can (at least today) do this with cable or dsl service without paying any additional fees and without Google needing to get permission from the cable or telco that provides the broadband access service.

Google would very much like to replicate this wireline experience in the wireless world. But this would require the equivalent of a slash-and-burn on the existing wireless business model. Unsurprisingly, existing wireless carriers will not become party to this radical reshaping of the industry unless they have no choice. But if consumers can chose a wireless network that gives them flat rate dumb pipe (as the basic contract, with the option to add on more customer service for additional fees), then over time they will, and the closed wireless carriers will respond to the competitive pressure by changing their business models.

I believe this because that is exactly what happened in the dial-up ISP world in 1996. At the time, America Online gave you 10 hours a month as part of your flat fee and then gave you metered pricing for the rest. But a lot of stripped down plain vanilla ISPs offered a flat per month fee. So people signed up to AOL, learned how to use the internet, noticed that they wanted a lot more than ten hours of access, and switched to one of the rival flat rate services. Eventually, even AOL had to switch to complete flat rate in order to reduce churn.

So Google figures that if it can create open alternatives, it can get the dominant players to switch over time out of fear that they will lose customers who want the flexibility and all around usefulness of an open dumb pipe where you can pay extra for hand holding, rather than be forced to pay for all the bells and whistles as a bundle. It’s absolutely true lots of people like bundles — AOL had 20 million more subscribers than its nearest rival at its peak in 2000. Many people like hand holding and flashy bells and whistles, and will happily pay extra for them. But many people do not, and even the people who like hand holding also like choice and like to tailor the bundle for themselves. An open Verizon will still make squindoodles of money selling bundles and hand holding to willing customers. But they will also sell a lot of plain vanilla connections to people who want to “roll their own.”

Already, Google has enticed the two national carriers not affiliated with landline phone companies to join their open system as a way of differentiating themselves from AT&T and Verizon and attracting customers. Verizon has responded with its own “openness option”. But Google understands that the telcos do not want to go nearly as far as Google wants. To truly transform the wireless world, Google needs to have its own wireless licenses so it can cause a truly open wireless network to come into being.

And for Google, more than wireless is at stake. Google can read the writing on the wall for wireline, even if dumbass regulators in DC can’t. Given enough time, the cable/DSL duopoly will gradually morph away from the existing open internet model to become more and more like wireless is today: you buy a basic contract for the core service and everything else costs extra. The provider bundles everything, controls the nature of outside attachments, applications, etc., etc., always taking its chunk off the top and driving up the price to everyone else. But if Google is successful in transforming the wireless world, it will also stop the transformation of the wireline world. By contrast, if Google can’t stop the transformation of the wireless world, it is probably screwed on the wireline side as well.

2. Google Has No Desire To Be A Network Provider. But It Wants To Be A Network Architect.

Analysts are right that Google absolutely does not want to be a network provider. It’s a business that requires a lot of investment and a lot of marketing, customer service, and other things Google doesn’t provide now and doesn’t want to provide. But Analysts draw the wrong conclusion from this and decide that Google doesn’t want to win licenses because it doesn’t want to build a network.

But as Google spokescritter observed in this New York Times article:

A Google spokesman said that if the company prevailed, it could lease the spectrum, offer wholesale Internet service through other companies or operate a nationwide wireless Internet network that would compete with DSL and cable Internet access.

Consider why open source software competes against Microsoft, despite the best efforts of MS to kill it. The answer is because open source plays by a very different set of rules. Open source can’t be “Netscaped” because it doesn’t play by the same rules as Microsoft. So the tricks that put MS at the top of the proprietary tree and help it cut down any serious competition don’t work on open source products.

Google is pretty much in the same position vis-a-vis the wireless world as open source is to MS. If Google actaully wanted to become a competitor to Verizon or AT&T, it would face all the obstacles and the limited rewards the analysts are talking about. But it doesn’t. It doesn’t even want to be in this stupid business. But it desperately wants the wireless world to change. And everybody, from Congress, to the FCC to the wireless industry made it abundantly clear to Google that the only way to change the wireless industry was for it to win licenses and build a different kind of network. Otherwise it was just a miserable freeloader trying to ride on other people’s networks.

So OK, Google gets it. But that doesn’t mean it will build the network itself. And if it does, it will not build a network that competes with Verizon and AT&T. It will build a dumb pipe network, which means it will operate under an entirely different set of constraints and with a radically different cost structure. But, more importantly, they probably won’t build it at all.

Because, to repeat, the fact that Google doesn’t want to run the network is a strength not a weakness. Google wants the consequences of an open network, and wants to absolutely ensure that the network will be totally and completely open. Google could never trust another carrier to abide by its vision of the “C Block” conditions. Verizon or anyone else who won would always want to interpret the “openness” requirement as consistently with its existing “walled garden” architecture as possible. But if Google controls the spectrum itself, Google can let other companies build their own networks provided they abide by Google’s openness rules. Given this recent article in Businessweek describing how not just Frontline, but other companies want to offer spectrum wholesale if they win the auction, it seems likely that Google will find willing operators around the country willing to build and operate the network under rules set by Google. Since Google will make its money on the applications, advertising etc., not on the wholesale profits or even selling the wireless access to the customer, it can work quite nicely with these potential network operators without the problems that have afflicted the existing MVNO/resale market — much the way open source works with clients on an entirely different model than the models developed by vendors of proprietary software.

When you combine Point 1 + Point 2, the idea of Google being serious about winning makes more sense. They cannot get what they really want any other way. And they will avoid the doomsday scenario feared by the investment analysts. Even if they bid $6 billion for licenses, they will not spend another $12-$20 billion on a network or get trapped in a high cost/low margin business against more expert opponents. Google is not stupid enough to try to become the next Verizon or AT&T. Rather, like a Toxoplasma Gondii protozoa driving mice to seek out cats, Google will drive Verizon, AT&T and the rest of the wireless industry to behaviors that accommodate the Google life cycle — even if the confused mouse gets eaten by a cat in the process.

Unlike mice, AT&T and Verizon are doing their best to resist by either bulking up on other “uninfected” spectrum or trying to ensure that they control the definition of “open access” for network devices. But a third factor gives Google the chance of grabbing the spectrum it needs and thus wedging itself firmly into the wireless community as the irritant that will create the pearl of open wireless networks.

3. Anonymous Bidding Changes Everything.

For the first time, the FCC will operate under a system of anonymous bidding. Analysts really have not digested just what that means to bidding strategies and behaviors. In fact, just about nobody has really thought about this with the possible exception of my friend and fellow Wetmachiner Gregory Rose, and people pay him to explain this stuff. While I cannot say precisely how anonymous bidding will change the outcome, I am utterly confident it will; and I am equally confident analysts haven’t really thought about how much this changes likely bidding behavior and outcomes.

For a start, the notion that Verizon’s willingness to open its networks takes the pressure off Google to bid relies on the idea that Google would know that the high bidder is Verizon. But the suppose the high bidder is really someone unknown, or someone hostile to the C Block conditions and who therefore plans to make implementation of the “open access” requirement as meaningless as possible. If you’re Google, how much do you want to chance that? Wouldn’t you rather have the spectrum yourself, and make sure that network device openness gets done right?

Similarly, if you are Verizon, you feel very differently about bidding against Google than about AT&T. Sure, you want the spectrum and you don’t want rival AT&T to get it. But at least AT&T is the devil you know. Would you pay more to keep it out of the hands of Google than out of the hands of AT&T? Anonymous bidding prevents this kind of strategic thinking. So what will Verizon do when it sees a higher bid? How will it asses the value of the spectrum? Will it make pessimistic assessments about the identity of the high bidder and assume that the high bidder is the worst possible winner? Or will it try to value the spectrum for its potential return without consideration of the possible “blocking” premium. And what about D Block? Suppose Google fakes for C Block but really bids for D Block? What if it tries to build a footprint via the smaller licenses? Alternatively, will a company like MetroPCS, that has previously avoided head-to-head bidding against larger companies, decide to go for C or D Block on a theory that it will be mistaken for Google or one of the larger telcos and that other bidders will therefore shear off? (MetroPCS as the Pufferfish of spectrum auctions, pretending to be much larger than it really is and thus frightening away opponents.) If the gambit fails, MetroPCS or similarly situated company is unlikely to suffer retaliation, as it might have in previous auctions.

Google appears to understand that under this new set of rules, a well funded new entrant has a host of new opportunities to overcome the advanatges of incumbents. And at the same time, Google cannot count on a more “open friendly” bidder winning the licenses. To meet its goals with certainty, Google must acquire a national footprint. At the same time, parties trying to block Google (or other new entrants) from acquiring a national footprint will have a very difficult time doing so. It is impossible to be strong everywhere, and parties that have relied on the mutual interests of similarly situated bidders to guarantee that a new entrant will meet with stiff resistance on every front find themselves forced to rely on their own devices without assistance.

When these uncertainties are combined with the aggressive geographic build out rules for the non-C Block licenses (if the reserve prices are met) and the conditions on C Block and D Block, this promises to be the single most unpredictable auction in FCC history. Such uncertainty favors bold well funded newcomers like Google, to the disadvantage of traditional winners like Verizon and AT&T.

4. When Google Commits, It Does So All The Way.

Finally, there is Google’s past history as a guide. Google may spend a lot of time dabbling in different businesses or investments. But when the time comes to act, it acts boldly and commits itself. Consider Google’s acquisition of YouTube for what some analysts judged as the excessive sum of $1.65 billion, especially in light of the copyright headaches the acquisition would bring. But Google committed and made the purchase, accepting that it would become a magnet for copyright lawsuits. And while Google has taken some steps to appease the IP mafia by setting up a video filter, it remains locked in high-profile multi-billion dollar lawsuits.

Google’s decision to jump into the wireless auction and take on the likes of AT&T and Verizon mirrors its decision to acquire YouTube and take on media giants such as Viacom. It’s extremely high-risk, and it remains unclear to conventionally-minded analysts focused on the bottom line how Google will come out ahead. But Google did not reach its current position by playing safe or undertaking half-measures. I don’t expect it to do so in wireless anymore than it has in such daring strategies as challenging the dominance of Microsoft Office with web-based applications or buying Youtube. Having decided last July to put itself on the line in wireless, we can expect it to play to win.

Mind you, I don’t say that Google definitively will win. Verizon and AT&T are not slouches when it comes to getting what they want, even when they have to fight fair (they may not prefer it, but they can do it if they must). And we will need to see after December 3 who submitted short forms to know how many hungry newcomers or crafty incumbents will show up for the ultimate Steel-Cage Spectrum Smackdown. But I think Google absolutely will play to win, rather than back away quietly if the C Block reserve price is met.

O.K., the oak leaf buzz is seriously wearing off now. I am done prophesying and must now go in search of some Ouzo and feta cheese — as soon as I get these stupid toga-wearing spiders off my face.

Stay tuned . . . .


  1. I really hope you’re right. And,anecdotally, “they just don’t know it yet,” takes 30 seconds to overcome. That’s how long it took me to explain this all to my non-geek girlfriend, who immediately said, “That’s what I want to happen.”

    Regardless of whether the wireless phone business is low margin or not (we could argue about accounting, maybe) … people still think that it is not, that the cell phone companies are ripping them off.

    I’m liking my t-mobile @ home … an extra $10/mo. for all the minutes you can eat over the Internet, anywhere.

    But ultimately, I expect to have the choice of paying for one wireless dumb pipe, not cell+cable+cable for my office (from a different provider in the next town with a special, higher commercial rate.) Bleagh.

  2. The only thing I don’t like about this is that Google plans on being the gatekeeper. Google’s going to have to overcome a lot of “They only want to build the network to monitor everyone” FUD. Not that AT&T, Verizon, and ??? aren’t all doing the same thing.

    For everyone who thinks the mobile phone market is a low-margin market: the only mobile provider with low margins is Metro PCS. Everyone else is making money. How is it that telcos can afford all the ads, lobbyists, and lawyers which are making our lives so miserable?

    It’ll be nice to go back to the wholesale/retail model. It’ll be like the wild west of the ISP era, when there were 30 isp’s in a region, all trying to sell you dial-up at $10 a month. The national carrier provided backhaul, but they didn’t see a dime.

    If Google collapses the telco infrastructure, I’ll gladly pay them the $10 a month, until the next broadband revolution.

  3. One small correction.

    There were some tests on volcanic gasses around the location of the Temple at Delphi…They just make you sick, no hallucinations.

    It turns out that the selection procedure for the priestess got them rural women, who also form the bulk of fortune tellers in the world today….They have a knack for self induced trances.

    It’s the human mind, not volcanic chemistry.

  4. Matt:

    Check the article I linked to above. It argues that the previous research discrediting the “volcanic fumes” theory was inadequate and the new tests suggest that hallucinogenic fumes were involved.

  5. This is the greatest blog post I’ve ever seen. And no, I’m not a splogger; this really is the best post I’ve ever read.

    Please, Google, bid high. Even higher than Harold on all those smoked leaves.

  6. Harold,

    Your analysis makes sense, including the potentially major shift in bidding strategies driven by anonymous bidding.

    What I’d add is the relative compatibility of Google’s likely wireless access model (relatively low-cost, dumb-pipe open-IP; revenue models weighted more toward ad-support vs. end-user fees) with use of white space spectrum to supplement 700 MHz network capacity and coverage (assuming the FCC approves such use).

    As you know, there’s potentially a lot of white space spectrum, and it seems possible that some of its buildout costs could be supported by end-user-funded FON- or Miraki-like network extension models that employ low-cost standards based mesh technologies.

    It also seems reasonable that Google might encourage such models and even subsidize them in ways that could get the network built faster and with a win-win cost-sharing model that would strengthen rather than threaten Google’s wireless business model (in contrast to their perceived impact on incumbent wireless business models).

    Networks and especially devices would need to support this 700 MHz/white space compatibility, but the relative timing and spectrum-bands involved seem pretty well suited for this to happen.

  7. cheating – how do you keep AT&T and Verizon working together to keep Google out? AT&T all our bids will end in 97 and Verizon all bids end in 93 – you get the idea – but “codes” could be slightly more complex.

  8. Mitch: Excellent point re: white spaces.

    Brandon: Actually, after a spectacularly obvious case of signalling with trailing digits in a spectrum auction back in the late 1990s, the FCC started taking signifcant measures to counter such signalling. Bids must be in specific increments, and the FCc can stop the bidding at any time if FCC staff monitoring the bidding detect suspicious patterns.

    It is always possible for clever folks to try to find new ways to cheat. But most bidders avoid outright violations of the rules. Usually, they try “rules lawyering” to find ways to legally flout the rules.

  9. While I believe there is merit to your argument, I find it very sad that we are auctioning off spectrum that belongs to the people and praying that the winner is someone who will do the right thing. Honestly, why can’t FCC keep the spectrum and lease it to network operators, just like Google (may) be planning to do, and enforce openness? There is no reason, except it’s against the religion (Privatize Everything!) of the Beltway idiots.

  10. Jaded:

    Blame Congress, not the FCC. Congress mandated auctioning the spectrum in 2005 to raise money. I agree that auctions suck (and have posted so numerous times). But unless we get Congress to change the law, we are screwed.

  11. AT&T will bid and win a Lower Band Nationwide License on the cheap, that does not need to be OPEN as the Upper Band C Block does. VW needs the Uppper Band C to remain competitive to AT&T in Data Services.
    Google needs access and control of “the Last Mile” if it is to become the Nationwide provider of Content/Hosted Apps delivered over serious Broadband Low Latency networks. Replacing much of the Best Effort Internet services.
    Most Local Incumbents (Independant Telco ILEC and CLEC) look to the 700Mhz as a must have spectrum to allow them to address the increased demands for Broadband Wireless Services to compliment their Wired services-They will bid and win CMA markets and use new 700Mhz RAdios based on WiMAX feature sets and release products in late 2009-2010.

Comments are closed