Blog Post

Broadband Fans, We Have an Innovation Problem

Google (s goog) last week said it plans to build an experimental fiber-to-the-home network that would deliver speeds of up to 1 Gbps. And this week FCC chairman Julius Genachowski outlined a goal of delivering 100 Mbps broadband to 100 million homes as part a “2020 vision” associated with the National Broadband Plan. However, amid what many perceive as good news for the wired broadband industry, the Telecommunications Industry Association and United States Telecom Association said they would not produce Supercomm, an industry trade show, due to “financial projections.” Translation: Wired broadband is in trouble. And it’s the fault of ISPs and Silicon Valley.

Despite a rollout of faster technology from some cable providers, and Verizon’s (s vz) continued fiber-to-the-home buildout, the wired broadband world isn’t looking terribly exciting outside Google’s testbed project. A close inspection of the long-range FCC plan doesn’t have me overly inspired, especially as other areas of the world invest in 1 Gbps networks today.

Meanwhile, in the same time two-week period as all of this wired broadband news, the mobile industry’s largest trade show, Mobile World Congress, took place. It was chock-full of the usual mobile players as well as a who’s who of anyone in the tech scene. And issues associated with mobile broadband, from new networks to spectrum shortages (GigaOM Pro, sub req’d) and how to build applications for mobile handsets (GigaOM Pro), were all anyone could talk about.

Who Needs Wires Anyway?

Wired was tired, and mobile was basking in the glow of the spotlight and investment. But even amid the mobile lovefest, a few discordant notes were sounded. For example, Stephen Bye, VP of wireless services at Cox — a cable company that’s deploying a 3G and later a 4G wireless network — emphasized the limits of wireless broadband.

Sure, Cox has a wired network to sell, but Bye has a point when he notes the shortfalls of wireless when compared to wired broadband. Cox’s wireless  LTE tests offered speeds between 10 and 25 Mbps, which are much slower than Cox’s wired Docsis 3.0 network that can deliver 50 Mbps or more. He also mentioned the increased demands Cox has seen on its wired network and said that sending that kind of traffic over wireless networks wouldn’t work. And wireless broadband traffic is only going to rise. AT&T  (s T) already saw its double from 2008 to 2009 and doesn’t expect that rate of growth to slow, even as it uses more and more of its spectrum. And Cisco (s csco) released information this month expecting mobile traffic to reach 3.6 exabytes per month — 39 times what it was in 2009.

I happen to agree with Bye, and I don’t have a network to sell, but I think the events of the last weeks  paint a pretty depressing picture of broadband in America. And we can only place some of the blame for the lackluster state of broadband on carriers. Some belongs with Silicon Valley and the tech community at large.

Wireless Isn’t The Answer.

For example, the idea that wireless broadband could be a real substitute for wired broadband showcases how crappy our current quality of broadband is. I’ve even weighed whether or not LTE or WiMAX would make a good substitute.

How could I not, when I’m stuck with residential broadband service that delivers 7 Mbps down and 400 kbps up? Wireless services are within striking range of that offering right now. It’s possible I might even get better upload speeds on some wireless networks within the year. My husband is even preparing to dump his pricey T-1 at the office in exchange for WiMAX service from Sprint (s s).

Yes, I have limited choice on the wired side thanks to ISPs failing to invest, but why haven’t tech innovators and entrepreneurs given me something so compelling, and requiring so much bandwidth, that I wouldn’t even consider dumping my wired connection, lest I give up that killer application.

Think Big. Build Big.

So I will blame my willingness to cut the broadband cord on the ISPs’ failure to invest in their networks, but also on a failure of innovation and imagination from technology firms trying to deliver services over fat pipes. Give me something that needs 100 Mbps, so everyone knows why faster broadband is important. Much like Foursquare gets everyone stoked about location, we need an application that requires multiple megabits per second.  I understand that there’s a bit of a chicken-and-an-egg issue here, since delivering a service before too many people have fat pipes will slow adoption, but at least 55 million homes already have the infrastructure to get 100 Mbps. Build something for them.

An emphasis on building products for fat pipes will help make wired broadband exciting again. And despite the investment required by ISPs, many — especially those with mobile networks — will win. After all, as wireless speeds get faster, consumers think they should be able to do just about everything on a mobile network that they can on a wired one.

And for most of today’s applications, that’s actually true. But if we had a bigger performance divide and different applications between wired broadband and mobile broadband then consumers might have an understanding that sometimes the mobile web just can’t compete with the wired one, and that we really need both. For carriers and consumers, that could be a winning proposition.

To learn more about this topic, join GigaOM Pro on Wed., Feb. 24, for the latest Bunker Session event: The New Broadband Buildout.

39 Responses to “Broadband Fans, We Have an Innovation Problem”

  1. I’m late to the show, but I have a point I want to make. You’re putting the burden of creating highly data-intesive programs onto the shoulder of developers to help spur the development of better networks, correct?

    Well, as an ameteur prgrammer myself, I usually strive to make applications as light as possible. And that covers application/data file size as well as RAM and processing power. Your saying that developers need to build bulkier programming just to make downloads slower/more laggy, so that ISP will be incented to build better networks? If I can get the job done and deliver a product efficiently and quickly to a customer, does it really have to be big? If I can create a better compression algorithm (not saying I actually can, mind you) to make my big program transfer faster, then why is that a problem for you?

    • Dude

      At no time did anyone say “build bulkier programming just to make downloads slower/more laggy, so that ISP will be incented to build better networks”. At least not me.

      I’m afraid you’re confusing the size of an app – to this thing called “content”. If you ask my friend Om (or Stacy or Liz) they’ll tell you that video and audio and text (collectively called content) are not only the bulk of what humans create, but how they interact, participate and engage – WITH.

      So NO – the world is NOT focused on the apps – but the content and THAT will benefit from1G connections.

      And then there’s this thing called stability, responsiveness and user experience – in general.


      • (/trying to not come off as defensive!) I understand the difference between the content and the actual software, but thinking specifically of web-based apps (client-side, more than server-side), there is still some bandwidth used, even if it is in only the initial loading. But that aside, my point about compression remains the same…

        While my original point was that applications are written to reduce this overhead, the same is true for content. Hence the usage of the MP3 format (or a plethora of other formats) as opposed to the more raw WAV format. Just because you have content to provide doesn’t mean it has to take up ridiculous amounts of space/bandwidth.

        One more example, then I’ll stop ranting… [sorry! ;)] If I have some graphic intensive content that uses bitmap image data to display all my buttons, I have the option to use true sized images, reduced size (for lower resolution/quality) images, or super sized images (for higher resolution/quality). As the source images get bigger and bigger, the difference in ‘quality’ becomes less and less noticeable, but the difference in data size becomes more and more of an issue. If I have created efficient content, it can look just as good and still be less cumbersome than some alternatives.

        I guess my ultimate point is that bigger is not necessarily better. (some one else hinted that more content – not bigger – would be a more suitable case to make.)

  2. Stacey,

    I disagree with you vehemently here. For one, I don’t see how Silicon Valley could possibly be to blame. The region contributes large amounts of new ideas and technology to getting more throughput from a given physical infrastructure. Things like Metro Ethernet, CWDM and other fiber enhancements. There is no end to the broadband innovation that is taking place here. That’s probably why so many global telcos have outpost offices here just to be aware of the innovation. Sadly, the innovations are deployed abroad more often than domestically.

    Next, you say “Give me something that needs 100 Mbps, so everyone knows why faster broadband is important.” That’s a terrible thing to say. That is the tech equivalent of a scorched-earth “Drill, baby drill.” chant. We don’t want wasteful, resource-hogging applications. We want applications that use shared resources efficiently. For gawd’s sakes, who would ask for more Hummers during an oil crisis?

    Like energy, bandwidth will ALWAYS be a shared, limited resource. I spent four years studying for a degree on the topic of how to most efficiently allocate scarce resources in markets, and I don’t cotton to “drill, baby drill”.

    Do you ask software developers for more bloated programs with more redundant lines of code in order to push PC makers to make faster computers? Do you ask people to eat more so that world food production will be forced to innovate? Do you ask agri-business to use more water so we can be forced to build bigger water infrastructure and de-salination technologies. That advice would be irresponsible.

    The reality is that, as consumers, our ever-increasing bandwidth needs are more often driven by additional applications and services being added to our stream. We don’t want fatter pipes so we can feed some poorly designed adn wastefully bloated app, we want fatter pipes so we can do more, different things concurrently. The demand for more bandwidth will grow quickly enough, we don’t need to Palinize it.

      • Brett Glass

        Marc, at least, seems to have caught the fact that I was paraphrasing Stacey.

        But seriously, what’s up with Stacy and with GigaOm? All they seem to be able to do is trash broadband providers — especially wireless — without knowing how the technology works, what it’s capable of, or how ISPs operate. Perhaps the bashing is just yellow journalism, intended to keep readership up by attempting to inspire outrage. In any event, the above article is uninformed and seriously off base. Wireless can do anything wires and fiber can do — often better. The only constraints are economics and regulation. And as I sit here, in my lab, you can bet that I’m innovating. It’s sad that government bureaucrats will determine whether those innovations will see the light of day.

      • @ Brett
        “Wireless can do anything wires and fiber can do — often better. “

        You must not live in the Rocky’s :-) One might even loose GPS in small Mountain valleys. Let alone wireless.

    • Brett. I’m a wireless guy. But I’m not too stubborn to admit that any signal that can be sent wirelessly could be sent:

      • further
      • with greater capacity
      • more reliably
      • in higher compression
      • with more bundled carriers
      • with more spectrum re-use
      • with lower power
      • more securely

      …over a physical carrier (i.e. cables). Sure, I’m not talking about CapEx or OpEx which could go either way, but I think Stacey’s point is largely supported.

      • Brett Glass

        If you were really a wireless guy (or an engineer), you would know that wired vs. wireless is irrelevant to encryption or security. And the longest distance communications that mankind has ever performed have been wireless, not wired. As for capacity: Give me the spectrum, and I’ll give you all the capacity you want. The only constraints on the capabilities of wireless are political and economic. Period.

      • Brett says: “wired vs. wireless is irrelevant to encryption or security.”

        Um…really. What about the physical access needed to tap a wire, versus the often omnidirectional broadcast of wireless signals? Irrelevant to security?

        “longest distance communications that mankind has ever performed”

        Well, yes, I agree that we have yet to string wires up into space, so duh. I wouldn’t choose a wireless career if there was NEVER a scenario where wireless was appropriate. I just said I’m not so stubborn that I think it’s every situation. You, apparently, are.

        “As for capacity: Give me the spectrum, and I’ll give you all the capacity you want.”

        Well, isn’t that just the point of it, though? Wireless spectrum is a limited resource, and they’re just not making any more. They are, on the other hand, making more fiber and copper.

        “The only constraints on the capabilities of wireless are political and economic. Period.”

        …and physics, and the occasional existence of better alternatives, oh, and Hartley and Shannon’s law. But you knew that, I mean, I do and I’m not even an engineer.

  3. Stacey thanks for taking the time to point out the interconnections, because ironically these niggly details is the nuts and bolts of a broken telecom industry that operates on a broken capitalist business foundation in this country. We are now seeing the devastation to innovation that protecting the status quo can generate. They take their cue from banking and the Auto industry. Keep service low and costs high to maintain the almighty margin. What good are the verizon build outs when Comcast’s monopolistic practices keep me from even getting their service. Like Rick said below, until consumers flex their power in unity to change mindsets, … wont happen in this lifetime.

  4. While a libertarian at heart I feel that fiber to the home is best served by a single not-for-profit public utility who then lease capacity to the ISPs and TV providers who want to target homes in the area.

    Why did up a street three times for Comcast, Verizon/FiOS and AT&T Uverse when all three could piggyback on PublicCo-Fiber?

    I also feel the same is true of cell towers – why pay to hoist three masts on the same building and triplicate infrastructure when a single node could serve multiple operators.

    The benefit of this is that the $ which Comcast would have spent rolling fiber are reduced and they can focus on customer service, content and innovation in their services while their contracted lease payments allow PublicCo-Fiber to build and maintain a robust and reliable network and (as a public utility) their books are open to inspection by the leasees and the public to ensure that any income is focussed on innovation and improvement not lining the pockets of a few investors

  5. I was at a startup working on Ethernet in the First Mile. It closed up shop in 2002. The choice quote from VCs: “The telcos are your only customer, and they will find a way to screw this up.” So yes, in my experience there is little to no VC investment in wired broadband. It seemed to be a general consensus amongst investors we talked to that telecommunications was not a viable market to develop new technology for.

  6. Greg Andrew

    Unfortunately, weak government regulation has led to a duopoly control of wired broadband almost everywhere in this country, and as a result there’s little incentive for those duopolies to invest because the money will keep rolling in no matter what. In most other western countries, governments did a much better job of creating an environment in which isps need to compete and innovate to survive. If it wasn’t for Verizon’s aggressiveness (ie. if they had taken AT&T’s approach, which many advocated, or worse, Qwest’s), our national broadband situation would be much, much worse. Still, the resulting inconsistency of broadband around the country has led to the justified reluctance of companies trying to create businesses using that broadband. Given that and the crap that Google gets for “taxing” the resources of ISPs, who in their right mind is going to invest in anything that is heavily dependent on the kind of bandwidth you are talking about? With the future of network neutrality still very much in doubt, I wouldn’t want to invest my money in anything that depended on the whims of the broadband duopoly; that’s for sure. Other than a select few games, no successful businesses have been built on true broadband in this country.

  7. BuildMoreBroadband

    What you’re describing isn’t an innovation problem. It’s an investment problem.

    Telcos are in the business to make money. They weigh every investment against its perceived returns. It’s like any other business. I might say “I think there should be 10 great restaurants in my neighborhood, within walking distance.” Great. It means nothing. If there is a market for it, supply emerges.

    Telcos (and any other broadband provider) are generally afraid of making investments today, even if they think there may be demand. They see the risk of government regulation (“net neutrality”), higher taxes and all sorts of other kinds of red tape. AT&T CEO Ed Whitacre went to become CEO of Government Motors, where he now gets paid $9 million a year, courtesy of the taxpayer. I guess that was a cushier environment.

    If you want more broadband — and we all do — let’s remove the regulations on operators otherwise seeking to build networks. Tell them there will be no “net neutrality”, no FCC mandates of any kind, and no local politburos asking for “franchise fees” and other extortionist regulations. If you stop penalizing free enterprise, you will get more of it.

    Recipe for a broadband boom:
    1. Take away the threat of “net neutrality”.
    2. Abolish the FCC.
    3. Declare that all telecom/tech is to be 100% unregulated by the Federal government.
    4. Abolish all taxes on telecom/tech.
    5. Repeat the equivalent of 1-4 above on the state/local level.

    • So tell me, where is this Libertarian paradise working on this planet? You think Korea and Japan have less regulations than we have and that is why their BB speeds are higher than ours? I’ve lived in Japan and did business in Japan, Korea, Taiwan and China and I can tell you that that is not the case.

      My take is that the ISPs don’t see themselves in the business of being ISPs. They don’t want to be a utility like the water company. They want to somehow be content providers and sell that content over their privately owned networks. They haven’t figured that out yet and it is that lure of mega-profits that keeps them from simply doing a good job as ISPs.

    • “If you stop penalizing free enterprise, you will get more of it.”

      Haven’t you “Don’t regulate anything, and everything will be great” guys had your shot? That neo-con notion didn’t work in our banks, and it won’t work in telecom either. Markets are good…but only when they are competitive. Oligopolies and monopolies are NOT free markets.

      Do you think telecom in the US is free enterprise? Don’t you suppose that there are significant barriers to entry that prevent a competitive market. Spectrum licensing is a regulatory limit to new wireless carriers. Local franchising rights are limitations to last mile competition. There are natural monopoly characteristics to an industry that needs to provide service to every home in the nation. A good policy would regulate such as to provide a competitive market for service over some fixed, unbundled infrastructure.

      The evidence, also, is against you. Above someone mentioned that Korea and Japan are more heavily regulated, and also get better results. I would cite Britain or France, also much more regulated, where local line unbundling has created phenomenal competition, much better services than we enjoy, at shockingly low prices from a US perspective.

      Your argument that carriers won’t invest in such an environment is proven wrong by the fact that the UK incumbent, BT, is forced to compete as a wholesaler, and as such has made some of the largest R&D and capital investments in global telecom, namely the 21 Century Network.

      I agree with you only on this: forget Net Neutrality. Let’s, instead, copy the worlds proven successful regulatory schemes and get a competitive environment where Net Neutrality is easily dealt with by the market using supply and demand. Net Neutrality isn’t even a subject worthy of discussion in a highly competitive broadband market like the UK. It exists because the market demands it.

      Your 5 point approach is just silly. Why would you propose an experiment in dubious (banking crisis) de-regulatory theology, when we could instead just copy the wildly successful models of so many other nations?

  8. Stacey — ever play World of Warcraft? Or basically any other online offering? The apps are there to take advantage of super-fast connections, but they are hobbled trying to navigate slow internet speeds without a huge loss of perceived performance.


    p>The apps are there already – the ISPs are not.

    • We are still living with low fidelity video and calling it high. Look at what you can buy in a Red video camera for around $10k: 4.5K WS (4480×1920) [about ten times more bandwidth]; 120 fps [four to six times more bandwidth]; 3-D [doubling bandwidth]. That strikes me as 50-100% more bandwidth than streaming an HD video now would take.

      We are suffering through images with easily identified pixels, motion artifacts, screens too small to walk through or sleep on or that force us to turn our heads to see the whole thing, and video glasses where the difference between CGI and reality is obvious.

      • Tee Hee Hee – spoken like a true Silicon Vally guy who can walk to Pixar.

        Out here – in the real world – they wouldn’t know pixel from a pixelated. Mobile bandwidth seems fine and as long as they can get their movies on iTunes, NetFlix or Hulu – everyone is happy.

        And besides – WalMart just bought Vudu – so now we REALLY can get movies on-demand – with shitty bandwidth!

        But Phil is right – the Red Camera ROCKS! Can you send me a free one – please?

  9. Let’s go to utopia. What could we do? How about a personal search engine, which is a partial part of a large engine in the cloud. Which would take ones data and learn parallel to one how to organize ones data. It would actually work for you instead of you clicking like mad to gather some information.

    Problem is this would require some more innovation, massive parallel programing. We will get all these cores,any time soon now. But what are we going to do with them? Run a click happy browser?

    Or we all can just build and distribute our own reality shows for everyone to watch. Question is how sustainable that is in an advertising world. Or crap content is the enemy of advertising. Not that my reality show would be crap, how about an Elk walking slowly through my back yard? Or watching the snow fall, with or without the Elk herd.

    Point is it’s not only pipes which need some innovation. But companies seem to be more oriented to the Elk scenario.

  10. What I think we should embrace is the ‘convergence model’ which takes into account multiple levels of bandwidth, multiple deployment and data endpoints and multiple usages of the same technology.

    Getting deep, rich, scalable connectivity is key to a certain few apps and services, but it also improves infrastructure across the board. Any home or apartment with a fast line – can then attach a wireless ‘repeater’ to it – and share the bandwidth wealth.

    There are still neighborhoods without ANY bandwidth what do ever. So pushing the envelope further, while also raising the bar of our expectations – are not mutually exclusive.

    The mid-mile solutions that OneCommunity has installed in Cleveland can now be leveraged into lots and lots of full fiber 1G connections throughout the city. So instead of wiring up remote rural hamlets, we’re now bringing 1G connectivity to city dwellers.

    But that doesn’t mean they won’t use the wifi at the local Starbucks or browse the web on their iPhones.

    It’s all good.

  11. Part of the main problem as to why the US continues to be behind other countries (Asia, Europe) in both broadband and wireless is the lack of a standard. This is particularly evident in wireless. In the US each carrier has developed their own network, network standards, even devices, so that they can lock people into their network. But in the end this is counter productive because each company is competing against the other in way that slows down development over all. By using a standard for the industry, development for the overall network will be increased rather than disjointed and counter productive.
    Actually it might even get worse, and this is where I definitely blame the ISPs for hampering development. If the big carriers, Comcast, Verizon, Sprint etc. have their way they will be able to filter data. They will be able to slow down any data that is not part of their service. So, what’s the point of faster connections if the ISPs are going to slow down traffic because they think it will give them a competitive advantage. This will only create more and more islands within the internet. If I have Comcast as my ISP and Vonage as my VOIP provider. Comcast would have the ability to destroy the quality of my phone calls. Or if I Google something it takes forever for it to come up because my ISP thinks Google is making money off their network for free. Do they really think Google is not paying to connect to the internet? Am I not already paying them to use their network?
    Bottom line is if the US wants to compete with other leading countries in broadband and wireless, carriers/ISPs need to stop being counter productive and concentrate on their service instead of ways to sabotage each other.

  12. Is this the same Stacey who has written numerous articles that the gov’t should be taking other citizens money and spending it subsidizing faster and faster networks?

    But when it is her own money – she doesn’t see the value in faster?

  13. Well, one application may be cloud computing. That would require a lot of bandwidth if it’s ever going to happen.

    For me, personally it’s just a pipe dream. I’m older and may not make it to 2020. My needs will never be met because those capable of providing the pipe dream, have absolutely no reason to spend a single dime that doesn’t promise a 50-100% return on their investment. So, it’s just a dream. Today’s society is built on promises not actual performance. So, go ahead and drag feet AT&T, Verizon, Hughes, etc. while you pay yourselves multi-million dollar bonuses for imaginary performance.