Tuesday, July 10, 2007

The promise of open platforms in the upcoming spectrum auction

As I’ve written before, Google has become increasingly involved in U.S. spectrum policy issues this year. One of our top public policy objectives is to expand the Internet's reach to more Americans. In part, that means creating new competition to challenge the existing broadband access duopoly (between cable and phone companies), by paving the way for consumers to gain meaningful alternatives via advanced wireless services.

Unfortunately, the wireless airwaves required to develop such a service traditionally have been allocated in a fragmented and inefficient manner. The federal government’s upcoming auction of spectrum in the 700 MHz bands (as part of the digital television transition) offers a tremendous, and probably unique, opportunity to promote competition and web-based innovation.

Earlier this year, Google and other members of the “Coalition for 4G in America” urged the Federal Communications Commission to adopt flexible rules that encourage competitive entry by new and innovative broadband companies. At the time, we stated that our advocacy in the 700 MHz proceeding did not necessarily signal our intention to participate in the auction itself, although no final decision had yet been made.

In comments we filed in late May, we stressed that new entrants face considerable hurdles when competing head-to-head with incumbent wireless carriers. We also noted that a proposal by Frontline Wireless to impose a wholesale/open access mandate on a certain spectrum block would ensure that the owner of that block at least would operate its wireless network in an open manner.

Over the last several weeks, we’ve been taking a closer look at whether and how Google might participate meaningfully in the auction. As part of that look, we've consulted with spectrum auction experts and conducted various game theory scenarios. Our analysis has confirmed that, under the originally proposed rules, the existing national wireless carriers are likely to prevail in the bidding process against a potential new entrant like Google. While we remain interested in the possibility of participating in the auction, it’s clear that the incumbent carriers have built-in advantages that will prove difficult to overcome (particularly the economic and operational barriers to entry for a company like ours, and the relatively greater value and usefulness that spectrum brings to existing carriers).

What would happen if one or some of the existing national wireless carriers win this valuable spectrum at auction? They would probably use it to protect their existing business models and thwart the entry of new competitors -- both understandable actions from a rational business perspective. Beyond the loss of a valuable public resource, however, that outcome would not bring us any closer to fostering much-needed competition in the broadband market, or providing innovative new web applications and service offerings.

Too much is at stake for the federal government to let that happen. Late yesterday, we filed a letter urging the FCC to take concrete steps to make sure that regardless of who wins the spectrum at auction, consumers’ interests are best served. We believe that the winning bidders should be required to adhere to enforceable rules that require the adoption of four types of "open" platforms:

  • Open applications: consumers should be able to download and utilize any software applications, content, or services they desire;
  • Open devices: consumers should be able to utilize a handheld communications device with whatever wireless network they prefer;
  • Open services: third parties (resellers) should be able to acquire wireless services from a 700 MHz licensee on a wholesale basis, based on reasonably nondiscriminatory commercial terms; and
  • Open networks: third parties (like internet service providers) should be able to interconnect at a technically feasible point in a 700 MHz licensee's wireless network.

We believe that adopting these four license conditions collectively will encourage prospective broadband companies to participate in the auction, and be able to bid successfully for the available spectrum. Not only are new entrants more likely to embrace an ethos of openness, but additional forms of competition will emerge from web-based entities, such as software applications providers, content providers, handset makers, and ISPs. And consumers ultimately will come out ahead in that rich and vibrant broadband environment.

In the meantime, there is now potentially positive news coming out of the FCC. Chairman Kevin Martin apparently is about to circulate proposed auction rules to his fellow commissioners, and we're hearing through the proverbial grapevine that his proposal includes several of the open platform conditions we have recommended. If these reports are accurate, we are most encouraged by this favorable development. Obviously we'll need to see the fine print, but such a proposal would represent a step forward for new, innovative entrants to the broadband market.


Tom Coseven said...

Presumably, if Google chose to bid and won, it could implement all four open conditions without FCC mandate. What if the FCC required the first two open conditions (open device & Internet connectivity), but not the second two (wholesale and unbundling)? Would this be enough?

In other words, besides the "big is bad" argument (which is a strange argument for Google to make), why do we need "wholesale" and "unbundling" to create innovation? Wouldn't allowing a user to purchase any device and run any application, by itself, create an open market for innovation?

JoeDuck said...

I'm less interested in "fairness" (a concept that rings a bit hollow to regular folks when coming from market movers like Google) than in the scenario that provides the most robust and scalable broadband infrastructure for the USA. So rather than tell me how you want a level playing field, explain to me how Google's going to make a better broadband footprint for users.

Leechy said...

Tom - the wholesale/unbundling condition (at reasonable commercial terms) is thought necessary as THIS is what brings more competitors into the market. Imagine being forced to rent a competitor space on your network, only to have them compete with you on price. Obviously more consumers are happy ...wireless providers, not so much.

Also, some suggest that if the incumbents own all the airwaves, there's no incentive to come up with innovative new stuff. ...I'd also point out that significant techno innovation has come from the folks at at&t (which was busted up for being non-competitive) via their lucent/bell labs. Big co's like to think of neat service/products to sell us just as much as little ones do.

Ash Rust said...

The 'big is bad' argument is not being presented by Google here, they're simply highlighting the broadband duopoly, which isn't their label. The contrast between the broadband and search markets is eloquently articulated here.

Wholesale access to wireless services and unbundling promotes innovation through direct commercial avenues such budget, luxury or even individually customized pricing models; as just one example. In terms of technology, it allows software systems outside of normal connection scenarios to interact with each other or with hardware across huge areas which previously required wires. Taking another very simple example, it could advance pizza delivery to a stage where you know how far away the delivery guy is from your house, how much change he has left and whether or not he knows the code for your building. You could even drop him a quick message so he knows the code any maybe prioritizes you, because you're in a rush. It's easy to imagine hundreds of other benefits too, admittedly these won't be removed if the duoploy continues but it's important to remember examples such as VOIP which was taken mainstream by Skype and not the large telecoms companies.

There is no doubt telecoms giants do innovate but often their commercial innovations only strengthen their existing business models. Open platforms could spawn innovations which offer all kinds of pricing and quality benefits to consumers and if the telecoms giants really are innovating anyway, what do they have to fear.

Finally, the requirement for a robust and scalable network over an open network, will ultimately reduce reliability and stifle innovation. A reliable network will emerge from a culture of competition and innovation, intially allowing the current 'reliable' giants to control the network removes their incentive to improve; as it won't increase their market share. Furthermore, adopting something easily conceivable (a reliable network) as your goal damages our future potential; 15 years ago few cared about email and the Internet, now most cannot live without it, imagine what we won't be able to live without in another 15 years.

Tom Coseven said...

Ash -- so to net it out. Your argument is that resellers could offer "walled garden" services better than the operators that build the network.

My question was essentially why not offer all services "over the top," since the Martin rules would give you have open device and open aplication connectivity. Your response was that there are walled garden funtions deep in the network that could provide value. So why wouldn't you go to operator and offer them a deal they can't refuse to unlock the walled garden function. The answer is obvious... it would violate net neutrality to offer just one provider access to the walled garden function. However, if the walled garden "fast lane" were offered on a wholesale basis to a a single reseller (say Google), then it would not vioate net neutrality. Problem solved!

Ask yourself the question: Why does Google want the fourth provision?

Open networks: third parties (like internet service providers) should be able to interconnect at a technically feasible point in a 700 MHz licensee's wireless network.

Why does Google want a connection point INSIDE the operator's network, if not to provide an un-neutral fast lane?

Stephen said...

John Deere said... That Google would borrow a page from CLECs (Competitive Local Exchange Carriers) aka new telecom entrants and ask for unbundling and interconnection of a new wholesale wireless network is asking all of us to believe it is in the same place as long distance companies were in the 1980s and 90s. That's quite a stretch. Bottom line, if the FCC ever issued an order requiring unbundling and interconnection and assuming a carrier was crazy enough to ever become that wholesaler (no one would want that obligation), the arguments would come down to price whereas today they are about I guess traffic prioritization and price to som degree. Seems like an awful long way to go to change the argument (from traffic routing to price) or (from one price to another price).

Ash Rust said...

In a free market economy the question should not be 'why not a monopoly?' but 'why should there be a monopoly?'. In this case there is no reason why consumers should not have the right to choose from a wide array of wireless providers.

The requirement for open networks has many uses, for example: critical network traffic which cannot be placed in the trust of a third party. Let's say in 5 years a surgeon can perform an emergency operation via video and robotics but be on the other side of a city, this network traffic is critical and open networks would afford the option of perhaps some guaranteed service which the current telecoms giants might not be able to guarantee.

Finally, if Google was to provide some form of 'un-neutral fast lane', I wonder where you think it might lead. I would suggest it might go to the Internet, as almost all sites on that network appear on Google and Google's profits are almost solely acquired through adverts which appear on its search engine. That being the case, if Google were to favor certain areas of the Internet over others, they would risk reducing profits and a severe consumer backlash.

William said...

why not have it all owned by the federal government and have "free" WiMax for all of us. (paid out of our federal income taxes of course) I understand that this would make it difficult for other wireless providers to compete, but for all other industries in the US that are even remotely connected to communications, it would be like switching from a world where every road is a tollway to how most roads are today. massively liberating for the average person and a true revolution in the media landscape and so great for our democracy as well.

Serge said...

Seems to me that unbundling and wholesale are two conditions that allowed broadband to explode in Europe (UK and France namely), whereas here, where there are no such obligations, we are stuck with a crummy duopoly.
You absolutely need open services (wholesale) and it probably should be "at cost". Then you will truly have competition, based on the ability of a provider to differentiate itself appropriately (see Ash Rust's point: budget access with minimum service - high end service with tailored service)

Mount Camel said...

Ok, we got it that Google wants spectrum wholesale and unbundling between subscriber and service provider. Yes I also agree that Google is a new player in the spectrum space even though it is a market mover in the Internet space.

Not long ago, Google filed the one-pager FCC document talking about auctioning frequencies like they do Ad Sense bidding for advertisers. Since then I have been wondering what's in it for Google to jump in the 700MHz auctions, other than its deep pocket (thus the "because I can" possibility), or that there are the folks who took game theory classes at Stanford and are eager to play out the scenarios with what they learnt.

So what is the answer brought by the four "openness" request this time around? My guess is that it is in response to the big serivce providers' stronger and stronger pressure for Google to pay for consumer traffic visiting its site ($10B annual revenue, not bad to get a scoop of it...)

Presumably, with the 60MHz spectrum that Google would get from the auction with its $4.6B, it would wholesale to those interested, budget or luxuary. Yes we get it the lease will be acquired via bidding like Ad Sense. Terms would include what Ash (or someone else) pointed out, hey, give me a fast lane to the Internet for all subscribers punching in www.google.com.

To make the long story short, Google would now come back to the big service providers and tell them to forget about charging Google for the consumer connectivity. Thus also the Net neutrality initiative...

vRad said...


Spectrum space is limited. And it is open property that belongs to the larger public. If another company thinks of a fantastic new way they can use the spectrum (e.g., they come up with a way humans can communicate via telepathy using the 700MHz band) under current rules, they will have to wait until the next auction to be able to license some spectrum. Or else they will be forced to handover their technology to a band owner. Simply because they are not being provided access to what is basically public property! OTOH, google is asking that the FCC does not completely cut out such upstarts (why should a major player now, have such a huge advantage over a company that may be started next year?) but rather provides them with an avenue to implement their ideas, while still providing the licensees a fair share of the pie.

Obviously Google will benefit from it, but there is nothing wrong in doing what is best for your company. What the FCC should be doing is what is best for the public at large. Such open rules are indeed best for the public at large (prevents carriers from monopolizing) and theoretically, the FCC should be completely in favor of them. Unfortunately, lobbying speaks louder. Which is why I am glad that at least Google is on our side, this time.

Epitiro said...

Richard hi,

I read with interest your panel discussion piece at Santa Clara
University today. I'd like to draw your attention to our isposure
desktop agent - a freely downloadable application that allows users to understand DNS, PING, HTTP and most importantly raw TCP and UDP throughput speeds from their home. The TCP & UDP speed tests work by
terminating on hosted servers located on the major telco backbones.

See http://www.isposure.com

A more interesting module we've just completed is executing those same tests but using traffic indicative of Peer-to-peer sessions (packet payload signatures, ports and protocols). Testing has shown that this
clearly indicates those ISPs who are implementing traffic shaping
mechanisms in their networks.

We've not deployed this solution or the TCP/UDP speedtests in the US to date as we're worried about overloading the endpoint servers with requests given the sheer sale of the US.

Should it be of interest to you it might be worth exploring siting those servers within google's hosting centers and re-purposing our solution in support of your goals.

I'd appreciate your thoughts ...

best regards,
JP Curley
Chief Technology Officer
epitiro - global broadband benchmarking