Google’s tactical admission of middle age has everyone guessing who they are going to buy next. The kookiest of them all is Level 3 Communications, since they have all that fiber and connections within connections etc etc. Wall Street is buzzing with rumors about this, based on some analyst firm’s contention that the deal would make sense. Janco Partners analyst Donna Jaegers tells Denver Business Journal that it is just crazy talk. Sure Google could use all that fiber to build a big-honking GoogleNET but the deal is expensive. $6.3 billion in debt, and $6.18 billion in market capitalization. Yeah, that is too much for Google to take on. But then they bought YouTube!
Like I commented in your Goobed post, “From way out in left field… might this move be confirmation that Google is building a monster network to serve all these bandwidth-hogging goodies? I can think of a few nets, say AT&T and Comcast, that will balk at giving this gorilla a free ride. Battle of the TV 2.0 titans is next.”. The problem with L3 is that it’s a wholesaler, it doesn’t own eyeballs per se and the bulk of its revenue relies on the very people that YouTube/Google pisses off, telcos & cablecos.
Perfect play Om
Imagine the strength of Google voices offering with a carrier of level 3’s strength behind them plus with IPTV estimated figures looking strong this is a marriage made in heaven.
This is simply nutty (more nutty by far than $1.6 billion for YouTube). Why pay billions for access to fiber when you can rent it for millions?
Ok, what is The Next after Web 2.0 ?
If Web 2.0 is syndicating, sharing and taking the internet to its elementary pieces, then the Next is applying activates between the elements of Web 2.0.
If Content in Web 2.0 is an elementary piece, then the Next elementary piece is the relations per se, which are combined in the content, between such contents and which are linked by/via/to it.
Imagine peers having their private “knowledge base” on each of their machines. Imagine each of the peers employs each other’s machines, but contrary to grid computing or distributed computing, they do so in short pulses and in a developed reaction, in which each of the peers forms and not only performs the tasks, while earning and ranking the trust of the others.
If sharing in Web 2.0 is exercised by delivery of contents, then another scale of processing the information is opened up for such communities.
For more about this concept, please visit my site.
google is absolutely building an isp. want proof? go to their job listings and search for “ISP”
and the billions it will cost to do so will be mice nutts when compared to the revenue they’ll drive from each user. how much advertising can they deliver to a user over the course of a users lifetime? not sure what the number is but surely it surpasses whatever the cost is to build.
for clarification to Josh’s comment on why buy for billions when you can rent for millions? the short answer is by owning you have the ability to scale almost infinitely by dropping DWDM gear into the network and lighting up more capacity as you need it. when you “rent” fiber it is typically done by renting a lightwave. owning the network from end to end creates opportunities only dreamed of by @home network.
they can pull it off and create this ‘global village’ but it isn’t as simple as buying a network and being done with it. they need to have the cooperation of alot of parties and that includes vendors of theirs which have been pushed way too hard by google…ask any colo vendor if they will even do business with goog and you will likely get a two letter response. cost of power isn’t the only reason they’re building their own datacenters. remember when you used to get told to ‘treat others as you expect them to treat you’? well somewhere along the way someone forgot to explain this to google.
having said that, i believe a bit of humble pie will go a long way and schmidt has been around long enough to understand that things change and just as there was solaris there will also be linux.
RE: Tom’s comment above, goog may be building a backbone but the real crux is the access networks that they’ll need to connect to to reach end-users. The folks who own those networks have already shown that they don’t want to play nice. Build or buy, that is the question.
hey, Potter. 100% agree with you about the last mile and getting those folks to agree or participate if you will. but…in mt view they have basically gone around the last mile by installing wifi which they offer at no charge. you are right about certain networks not wanting to play nice especially to google. there is also the possibility that goog thinks the networks should be paying them to access the content and svcs that they provide to the network providers end customers…ie, if goog shut down routes to comcast for instance because comcast wasn’t playing nice, the ones to suffer are the comcast subs and ultimately comcast will be forced to play nice or they’ll likely lose their subs to someone who will play nice or that in and of itself will provide even more justification for goog to duplicate the wifi initiatives they’re doing in mt view and sf. earthlink seems like a logical play too in addition to l3 and they(earthlink) have many end users customers and lease fiber from l3 so that expense would go away if everything came under the goog umbrella.
lots of what if’s but think of larry and sergey’s long term vision and all this kinda makes sense. what industries are going to be the victim of capitalism at its finest? IMO, the three traditional media networks stand the most to lose as what google is building has the potential to be a much more efficient, targeted and trackable in realtime system for delivering advertising…
The money would be much better spent lobbying governments to build fibre to the home and blanket coverage of wifi/wimax.
And the single most important google development should be a web browser of their own to deliver their applications. Roughly 75% of google users require a microsoft browser to access Google. And of course 95% use a microsoft operating system to access google.
Buying Level 3 or any other carrier is as far away from a cultural and strategic fit as an oil company buying an IT system integrator.
Simon, an oil co buying an IT systems integrator may be a poor cultural fit, but it would be a great strategic fit. Do you realize how many $B such megacorporations spend on IT and communications (which could be converted to XoIP)?
Far fetched? If GOOGLE had
enough peering traffic (real peering)…would there cost of network of exchanged traffic be free? LVLT (Level 3) acquisition would provide that traffic to exchange. Cost of LVLT? Remember, they have over 5 BILLION in carryover losses that a PROFITABLE company like GOOG could use.
Secondly, LVLT is the “Internet 2” and is in position to gain Government contracts in 2007. They have 12 empty conduits on a 23,000+ mile network with over 5,000 on net building sites.
EVERYTHING will be high demand… Education Research, Video live anywhere, anytime…Vyvx (Level 3) offers real time video worldwide distribution…
HOW much would a network that has the only unlimited capacity
(already a fixed cost, 12 empty conduits)…who needs future bandwidth… $$$$
Late reply to Tom’s post above; the big eyeball nets will never shut down goog, but they will slow it down using packet/traffic shaping. Why kill when you can cripple? Cableco’s are already doing it with P2P traffic…
Hi Jesse, I think Schlumberger were thinking along those exact same lines when they bought Sema for $5 billion in 2001.
In a matter of a few years they divested their IT business to ATOS Origin for around a $4 billion loss.
As someone said to me after the sale, Schlumberger thought they were buying a cow in the end they found out they bought a diary and they had no idea how to run it.
Potter, just saw your reply. Your theory is correct, eyeball nets could shut cripple access to any of their peers and address blocks for that matter. it just isn’t logical to think they would do that because by doing that they would also cripple themselves. don’t forget, those eyeball nets make money through subscriber fees. if a subscriber doesn’t have to pay a fee why would they? goognet will be free. we went down this road when either L3 or PSI shut down peering to each other. within two days the peering was back in place because there was such a backlash from their subscribers and the inet community in general. last point to make is that the eyeball nets also get revenue from advertising and by making it difficult for other internet users to get to content on their network they are limiting the number of ads they serve and trust me, that is where the bread and butter lies. mso’s or cable companies can’t see the shark that they are letting loose on their own plants ever. cable companies are by virtue competitors to one another, can’t cooperate if their lives depended on it and are old school in terms of innovation. what goog is building now was first attempted by cox, comcast and tci via @home network. i worked for @home early on and witnessed how logic and rational reasoning plays no part of their strategy…if it had there wouldn’t be a google because it would have been gobbled up before it became a threat. p2p may work within one network but once one of those bytes has to leave that network it goes onto someone elses network and then it’s gone. this btw could be a big issue. if cable users were all using p2p and their node happened to be the most efficient way to pull content from another network, it would crush the capacity wrt their connectivity…they oversell capacity like everyone else does and if everyone was serving content via their home connection it wouldn’t work as it just isn’t efficient and doesn’t scale. the cable company in this case would be paying for their connectivity and $40 a month or whatever it costs won’t cover their internal costs.
Tom, good points all. However, the example of L3 vs. PSI (Cogent) is an example of what happens when you shut the pipe completely. What I meant by cripple was to turn the valve to trickle, in other words not shut the pipe completely, just partially so that the packets flow more slowly (may be a bad metaphor, but that’s the idea). Large consumer ISPs don’t care about off-net quality of service, they care about on-net and their own services (telephony, video, etc). I agree, it is counterproductive to piss off your customers this way – case in point Om Malik himself is switching ISPs because of poor performance (http://broadband.gigaom.com/2006/10/15/now-is-that-what-you-call-broadband/), but they don’t care, most peeple don’t switch, they just hold a grudge. Message to Om, it’s crappy everywhere, if it isn’t now, it will be shortly, quality ebbs and flows over time based on a number of factors that telcos can control and others that they cannot (peering for example).
Potter,
that sounds like a logical reason for google to provide connectivity services for free to the masses. they already have the backbone in place, so adding a bunch of wifi access points is easy. it actually provides incentive to give away free access because the traffic won’t leave their network and that which does can be handled via their peering and transit relations. look at their mgmt team…they have every piece of the puzzle in place to continue to build their “network” as we’re calling it, which will serve as their delivery platform for services like telephony, hosted business applications(no wonder Ballmer is saying they are too big), video, music, dating, education, what ever.
nobody else will be in a position to do anything electronically as efficient(think bill, alan, urs and jeff) or effective(think individually targeted) as google.