Google, Ads, and Kill Switches

There’s a mixed bag of news for Google to confront their new CEO, and it’s mixed in multiple dimensions.  Android and the basic business of search are showing a combination of positive and negative signs, and confusion is never a good thing.

On the Android side, there’s excitement over the new Honeycomb version for tablets, and also an indication that Android is gaining at least numerical shipment ascendency.  It’s passed Symbian as the most popular phone OS, and there are projections that Android tablets will pass iPads soon too.  So it seems all good for Android, except that Google hasn’t been able to monetize Android like Apple has monetized its smartphone and tablet platform iOS.  A big part of that is the ruthless discipline Apple has enforced on apps.  Apple kicked some Sony apps off because they let you buy ebooks around iTunes, and that suggests that Amazon and Barnes & Noble may also be in for problems.  Android has a more fragmented store framework to begin with, and Google has less control and gets less direct money.  Thus, success in a market sense doesn’t necessarily translate to something Google can bank.

In the search area, the problem is the competition from Facebook and Twitter.  Advertisers have run the numbers a bunch of ways, and what they’re finding is a mixed bag.  On the one hand, it’s clear that search ads get more conversions, meaning that people click on them and act on the results more often.  That’s likely because people search for products they want to buy, creating a direct link.  But social-network ads get more “impressions” meaning that they’re presented to a lot more people.  That means that even with very low conversion rates, they might still outperform search.  At the very minimum, this kind of mixed news could mean that some ad dollars will flee search for social networking, and if the return is indeed better then it might never come back.

The tide of operators looking at monetizing Internet traffic is a more unambiguous negative for Google.  All week, EU operators have been saying that it’s not tolerable for them to invest in additional capacity to support the business goals of OTT players when they don’t get any of the money.  In Europe, the focus seems to be on creating an OTT-pays-for-peering model that would provide the access ISPs some revenue.  Here in the US, that option seems foreclosed by the FCC’s neutrality order, but just in case there’s proposed legislation that would absolutely close that choice of payment off.  If that happens here, it’s likely we’d see what’s already happening in Canada, which is a sharp reduction in the per-month bandwidth allotment and a corresponding increase in what at least some users would then pay for Internet access.

For years, I’ve pointed out that the Internet usage model and business model had a fatal parasitic flaw.  OTT giants could develop because they exploited bandwidth that was incrementally free, and thus present the market with a kind of no-cost (or lower-cost) business model while network operators faced a no-revenue model.  If that sort of thing is carried far enough, it’s a fatal disease.  I think Google under Schmidt realized that they needed a constructive accommodation with the access ISPs, and that’s what prompted the Verizon/Google discussions on “neutrality”.  Whether Page sees the same need isn’t known.

The situation in Egypt is another policy thorn.  Can the government cut off Internet service?  In the US, as a practical matter, the answer is that it could providing that the big access ISPs honored a government request.  There’d be no legal compulsion at the moment for them to do that, but I think that if a good case was made they’d likely comply.  But the Administration is looking for enhanced cybersecurity, and some of the measures that could be taken might codify an obligation to shut down the Internet or a part of it on request.  The Internet community, of course, is inalterably opposed to this sort of thing, but I suspect that at least a slight majority of the population would accept or even welcome the idea.  Where we’ll strike the balance here remains to be seen.

Bits Don’t Rule and We Don’t Rule Them

FCC has filed a response to two provider lawsuits on net neutrality (one by Verizon), saying that because the order has not yet been published in the Federal Registry it’s not technically in effect and cannot yet be challenged in court.  That seems a rather lame move, but as I noted last week the current filings are in the DC Court of Appeals, which the operators believe would be likely to find in their favor given their ruling in the Comcast case.  That’s why they filed early to start with, and why the FCC is now hoping to slow things down.  If cases can be filed in multiple appeals courts there could be a lottery to decide which will hear the matter, and that might favor the FCC.

Internationally, we’re seeing some hardening of provider attitudes on neutrality and related issues.  In Canada they’re heading for usage pricing and much lower thresholds for incremental costs.  Telenor in Norway has indicated that the practice of having OTT content like YouTube simply pull capacity out at no charge must “stop”.  I think all of this is related to the same factor that caused Wall Street to downgrade Ciena today, and to Huawei’s record $28 billion earnings (it’s closing in on Ericsson’s top spot).  Return on bandwidth is plummeting still, and that means that operators are at risk for not being profitable at the transport/connection level.  If you’re a price leader, this is a good situation to be in, and Huawei is while Ciena is not.  Operators can’t rely on cost reduction alone, though.  They’ve got to somehow stem that process of commoditization, and that’s possible only if they can improve revenue per bit with some mechanism of per-bit pricing.

Neutrality rules, at least as some would write them, might foreclose the option that would actually be best for users; let the OTT players pay for delivery.  The other option, which is where Canada is tentatively headed, is to establish incremental pricing and usage tiers.  The “let’s go on as usual” choice doesn’t seem likely, or at least it isn’t likely to create a good outcome.  Network investment in infrastructure would likely start to decline sharply as early as 2012 without any relief, and no regulator can order operators to run an unprofitable business if those operators are public corporations.  Already, Eurozone carriers are looking to developing markets to invest because the ROI is higher than it is at home.

It’s higher outside connection/transport too.  Verizon bought Terramark, a leading enterprise-targeted cloud computing provider, and this shows that network operators in general and Verizon in particular are aware of the cloud computing opportunity.  It’s not that the cloud is the lowest apple or the biggest in financial terms, but that cloud infrastructure has so many possible revenue missions that it makes sense to get involved quickly, and effectively.  Terramark has a good customer base and reputation, and its VMware-virtualization slant on the cloud is quite harmonious with the enterprise interest in hybrid cloud computing rather than pure public services.  Network operators are also much more credible providers of backup and overflow-targeted cloud services than anyone except prime IT vendors, which makes this particular space a hot opportunity.

Lessons from the (Earnings) Season

LinkedIn, not consumer-directed sort-of-rival Facebook, is filing for an IPO.  The move may be a sign of confidence in the markets for early 2011, a sign of lessening confidence beyond the first half, or simply another indication the financial markets are eager to make a quick buck.  One interesting thing about the move is that one of our Wall Street friends calculated that the market capitalization (total stock value) of Google and Yahoo, when combined with the hypothetical value of LinkedIn and Facebook, would exceed the world’s ad revenues from all sources.  Go figure.

Staying with financials (it’s earnings season after all) Amazon shares dropped after the company reported 36% higher sales and 8% higher earnings but lower margins.  One must wonder how the Street believes that online retailers would be gaining sales versus other retail models without discounting, and how discounting could be equated with anything other than lower margins.

Bringing these two themes together we have the results of the commission impaneled to investigate the 2008 financial crisis; the report has lots of bad things to say about regulations and greed and other factors, but in my view it’s light on the real problem.  Wall Street wants to create wealth faster than the economy creates value.  Many of the stocks we’d see as being untainted by financial scams are still objectively priced above traditional market justifications of P/E multiple (which would be about 14 in the real world).  That says they’re being bid up, and being bid up simply means that somebody is promoting the notion that they’re going to keep going up despite fundamentals.  To me, that’s a bubble no matter how you try to disguise it.

Microsoft turned in an interesting quarter, with good numbers and good stock response coming not out of blazing success in its traditional spaces but from good Kinect reception by the market.  This clearly illustrates the dilemma even in fundamentals—the consumer is now the engine of technology rather than business, and that means that fads and not plans drive the markets.  Microsoft is being praised (through its stock) for coping with the changes, much as IBM has benefitted from its ability to reinvent itself periodically to cope with the transition from mainframes through minis to PCs, and from hardware differentiation to middleware.

In the broad economy, we seem to be seeing a continuation of the division between countries doing better and those doing not so well.  Debt issues cloud the Eurozone, Japan’s debt has now been downgraded, but China is coping with runaway growth and the US is showing signs of greater economic strength and improving employment.  GDP growth in the fourth quarter came in a bit under expectations (3.2% versus 3.5%) but consumer spending was up 4.4%, which bodes well for the coming year.  Interestingly, the US’s position is likely due to aggressive stimulus and rescue actions, and these are now under political pressure.  Presuming that the anti-debt posture of Washington these days is more than just theater, you’d have to wonder (as some economists are already wondering) whether we might not risk slipping back into stagnation by trying to cut debt too quickly and cutting programs too much.  Loss of confidence and jobs created by debt-reduction-induced economic declines would hurt the deficit more by increasing pressure on social programs and reducing tax collections, something the states are already finding out.  Hopefully the national planners will do so too.

Hulu’s New Business Model is Bad Industry Juju

AT&T’s report on earnings reinforced some structural changes in the industry that Verizon’s report had already suggested.  One basic truth is that mobile services are more profitable and more fertile areas for growth than wireline.  Another is that mobile service gains and ARPU both depend substantially on broadband and smartphones rather than on voice services.  Finally, both carriers’ numbers show that you can’t even support a wireline business model on voice any more, and you can’t support it on broadband Internet either.  You either make money with TV or you don’t make money on wireline.

That’s an interesting counterpoint to the reports by the WSJ that Hulu is looking at becoming a kind of online cable operator, offering a premium service only and requiring payment for the service.  This has resulted in some suggesting that maybe it’s time for the whole cable TV industry to be subsumed into a broadband delivery/Internet model.  But if TV is the only profitable wireline service, where does that lead?  We all know the answer to that one, but that doesn’t address the question of whether an attempted shift to a pay-TV model for Hulu, coming after the significant growth Netflix has enjoyed, couldn’t create some serious stabilization issues for ISPs.

Democrats, you will recall, have proposed legislation that would not only codify the FCC’s framework for net neutrality but apply its terms to wireless and explicitly bar any paid prioritization of traffic or “fast lane” other than that paid by the customer.  As I noted earlier, this kind of bill stands no chance with the Republican House so strongly against even the current neutrality rules, but it does show that the political winds are at least blowing somewhat in the direction of accentuating a kind of “Internet-must-carry” principle that could have major impact on future services.

Buried in the details of the AT&T report is the fact that the company has fallen far short of its target for fiber feeds to its cell sites—less than half its goal for 2010 was met.  Now let’s be serious here, gang, nobody doubts that AT&T understands how to deploy fiber; the problem isn’t one of skill or technology.  It’s ROI.  The financial industry has noted that there’s a surprise boost to vendors who offer inexpensive digital-over-copper to buttress AT&T’s tower bandwidth; clearly the fiber shortfall is cost-driven.

The ROI shortfall shouldn’t come as a surprise because it’s caused by the same forces that have marginalized wireline—bit commoditization.  Customers want more applications online, but they don’t want to pay more for the capacity to deliver them.  High-end services at 50 Mbps or more undersell unless there’s virtually no price premium for them.  4G isn’t something people want to pay for; they just want to get it.  The public doesn’t understand how the business model of online services depends on the simple launch point of being online with capacity to access the service, and nobody in the media is interested in offending them with the truth.

I’ve been told that with the current trends in both wireline and wireless, and with no additional revenue streams, broadband Internet would be an unprofitable service for US operators by 2013 and that mobile broadband would be unprofitable by 2015.  Carriers like Verizon and AT&T who are gaining customers in the traditional postpay market can only gain so much market share because there’s only so many customers.  The “Hulu model” of broadband TV would exacerbate the problem by driving up traffic in both wireless and wireline networks, and worse it would compete with the traditional video services of operators—services that are making wireline at least somewhat profitable and that aren’t contributing to wireless traffic growth.

But it gets worse.  Online ad revenue per user from a Hulu model would be about 4% of what TV commercials now bring on a per-user basis.  If we saw migration to an online model, we’d either lose 96% of the money available to fund content development or we’d have to presume that somehow advertisers would agree to pay 20 times or more as much for each online eyeball, which sure seems unlikely to me.

What this adds up to is simple; we’re heading for a usage-priced model of broadband foreverything and if regulators try to stem the tide they’ll simply drive the operators out of investing in infrastructure.  One way or the other, the Internet isn’t going to be the “over the air” of the future, with no marginal cost to deter usage.

Cable’s business model works not because of voice or broadband but because of TV.  Fiddling with that golden goose could cook a lot of the players in the industry—and the vendors who support them, and the viewers who depend on them.  We’re facing a major disruption here, and the question now is whether we can, as an industry, come up with some logical way to save the health of the system that the whole of the Internet depends on.

Finding the Bucks, or Making Them

Earnings season is underway, and I think it’s clear that the results are generally positive and probably more so than expected.  That raises again the possibility that the recovery is proceeding more quickly than economists expected, and another sign that may be true is that the major parties in the US now seem to be vying to take credit for what will happen.  You can’t be responsible for good stuff by doing nothing, so you have to be cooperative to the extent needed to get at least something done.  Later then, you can argue over who gets the credit!

The State of the Union address was a typically political instrument, and so of course were the responses.  While I think it’s clear that the President is right in his assertion that America has to become more competitive in the core production of stuff instead of focusing on mowing each other’s lawns or earning money through financial frauds, that’s been clear all along.  What’s not clear is how to make it happen in a society that’s becoming more focused on entertaining itself than on getting anything done or learning anything.  The shift from search to social networking in terms of time spent online is a reflection of this; how much can you learn by Tweeting?  In a societal-good sense, not much.

In networking, Cisco has again indicated that what it calls “ambient video” (meaning user-generated content) is going to put enormous demands on the network of the future.  But like the political process, Cisco’s light on realistic solutions to the problem.  Sure you can argue that to fix traffic congestion you buy routers (a logical strategy for a router vendor to propose) but the real problem again isn’t what it appears to be on the surface.  We need to know how to pay for the routers, and ambient video has the smallest monetization potential of all types of video because nobody is prepared to spend much to advertise in that kind of material.  Our research says that only about 0.04% of all the video uploaded by consumers has any potential for ad monetization, and even with that there’s the question of how the ad money ever flows to the network operator to pay for those routers.  Absent a solution, the only near-term measure operators can adopt is to put price pressure on the gear to improve ROI even when the “R” part isn’t growing.

Up in the service layer, all’s not rosy.  Google is reportedly unhappy with the sales of Android apps even though developers are reportedly more favorable about the Android platform.  The problem Google has is that there are simply too many Android versions that are developing as the platform struggles to match features with Apple.  Since Apple can monetize its iOS better and faster, it’s able to put more back into development, and since it’s setting the feature standard for the space it can choose its fights.  I think that Android’s current problems are transitory; by the end of 2011 I believe that Version 3 will be out and widely accepted, and that will create a much more stable framework for development.  However, another important element in the picture is just what Google will do with its HTML5 “URL store” concept and Chrome OS.  A better way to create the platform APIs to allow features to be either device-resident or hosted might offer Android some real benefits.

Google has taken an important step in another area, allowing users to port mobile numbers to Google Voice.  This means that the features of Voice in call management could be available to mobile users more easily, and that in turn would accelerate the disintermediation of operator voice services from future feature opportunities.  The capability isn’t offered for wireline voice because of issues with E911, we’re told.  Operators will need to address this with their own set of richer voice features, but if they try to do that inside the traditional IMS envelope they may face price/cost problems that will work strongly in Google’s favor.  That means they’d have to block competing voice apps, and even the rumor that one operator plans to block Skype has been enough to create an FCC complaint under the new neutrality rules.

Earnings season will continue for a month or so now, and I expect we’ll have other comments and predictions based on the numbers that emerge.  I also think we’ll be seeing signs of how Apple and Google will operate in their new age of management, and that may be the most critical issue for the industry right now.

New Brooms?

The shape of the networking industry has long been determined by forces on the outside in what could be called the “on-net” space, and two powerful players there are undergoing management transitions.  Apple is losing (at least temporarily, though we hear management expects Jobs’ departure to be permanent) its charismatic CEO and Google is switching its politically connected “professional” CEO in favor of founder Page.  How much these changes will impact the companies involved, and the industry, will surely be the focus of much discussion but we’ve got to weigh in with our own views since both Apple and Google are truly seminal forces.

Apple has, more than any other company, transformed the relationship between users and networks—by transforming the instruments that connect the two.  Anyone who has worked with, or inside, Apple knows how much Jobs has shaped the company and how much his vision of the future has dominated Apple’s planning.  But his style has made it difficult for Apple to do any succession planning despite the state of Jobs’ health, and many inside Apple have suggested to me that charisma and determination have more often slipped into intransigence in recent years.  Apple’s vendetta against Adobe’s Flash, for example, have put the company into a position of supporting HTML5 when it’s clear that HTML5 is a benefit to the browser-based Google model of the future more than to the Apple app-based model.  In fact, the iPad/Phone incumbency is rendered meaningless if all portable apps are nothing more than URLs into an HTML5 world.

In the case of Google things are a lot more complicated.  Eric Schmidt isn’t a charismatic figure, and he’s generally seen by people in the Valley as a bit of a stuffed shirt, a businessman and not a real tech guy.  Brought in to add some “maturity” to a management team that investors tended not to trust, Schmidt championed a number of things outside the normal range of an online search giant—most conspicuously stuff like cloud computing and enterprise services.  He’s seen as having let social networking languish, losing the space to Facebook.  Some say he didn’t back Google Wave properly (others say he promoted it too much).  In any case, he’s now being replaced by one of the “infant” founders and there’s a lot of talk that this is going to prevent Google from “going Yahoo”.

It won’t, because Schmidt isn’t the problem.  Google is now a public company, a company that has to make money for its shareholders either through stock appreciation or through dividends.  I’ve said for years that there’s a fundamental problem with an ad-revenue model—the total value of all advertising can grow only at the pace of GDP, and gaining market share to show strong growth invites (as Google has already seen) regulatory scrutiny.   Google really needs to transform itself, and it’s not clear that Page is the guy to do that.

Enter Cisco Videoscape

Cisco took what could be a giant step for itself at CES with its new video ecosystem.  Called Videoscape, it combines in-home tools and software to centralize the mediation and management of video relationships, creating what’s probably the most architected video service layer available to network operators today.  Since Cisco was already doing well in the early content monetization project trials, Videoscape could be a real winner for the company.

But despite the positives, Videoscape still has some issues in my view.  Paramount is that Cisco is developing a content strategy in the absence of an overall service-layer strategy, or at least is creating the latter by simply assembling pieces instead of creating an architecture.  Most of the stuff in the Videoscape Conductor (the back-end) could easily be helpful in other missions, but it’s not clear how they’d be applied outside the video context.  There’s also a very strong push for video sharing and uploading, which generates traffic for operators and has essentially no potential for monetization.  That makes the product a bit of a risk in itself, but it also shows that Cisco may pursue its own aspirations (which are to generate so much consumer video traffic that operators are essentially forced to buy tons of Big Iron to carry it) more than support the operators’ business cases.

In the net, though, Videoscape is a strong achievement for Cisco because it plays to their strength—breadth in the video market.  The net effect of deployment could be a kind of “TV Everywhere”, and with Comcast pushing that very thing already, the timing couldn’t be better.

Economics and Networks

There’s more good economic news this morning; ADP’s private payrolls report gained the largest number of jobs in its history, which strongly suggests that hiring may be coming back.  You may recall that our forecast for unemployment for 2011 was considerably more optimistic than the official one, and I’m hopeful that the ADP data is validating that optimism.  Employment is the biggest barrier to a resumption of normal economic growth.

Qualcomm is buying Atheros, a chipmaker whose product line is more directed at smart appliances, including smartphones and tablets, in yet another validation of the consumer electronics craze.  The move comes as both AMD and Intel announce their own successor chip families, and the former seems directly aimed at the low-end market, including tablets.  Intel clearly has aspirations in smart devices too, and has been promoting its own Linux-based OS to gain some developer credibility.  With CES launching today we’ll certainly be hearing more about tablets, and while it’s obvious that not all of those announced will be market leaders, recall that the laptop market has plenty of active players.

Speaking of M&A, Dell has purchased a security company (SecureWorks) to buttress its enterprise services position.  What’s not clear yet is whether Dell will be applying the technology to the standard data center framework or focusing it more on private clouds.  My research says that enterprises are very concerned about the way that private cloud computing and hybrid clouds would impact security, and while this isn’t as large a market at this point as the managed security services market, MSSP is a security outsource strategy that might in the long run reduce the revenue Dell could hope to obtain from the acquisition.

Some financial analysts are posting positive comments about both the LightSquared “wholesale LTE” model and the vendors who are involved in it.  The basic idea of LightSquared is to provide a wholesale wireless network with national coverage, a host to MVNO relationships with players who want a wireless presence but don’t want to run a network themselves.  The idea has some appeal in that there are certainly companies (cable companies come to mind) who are likely to fit the customer model, but there are also challenges.  In fact, there are three.  First, wholesale profits are lower than retail, and the industry is already squeezed.  Second, the MVNO model has been tried by operators, including for cable MSOs, and hasn’t exactly sung.  Third, it’s far from clear that a satellite network hybrid will be technically successful and that terrestrial coverage will be ample where it’s needed most.

So here’s how I see it.  LightSquared has marginal financial options at best, and there are a lot of factors that say “best” won’t happen.  Thus, I’m not inclined to give the deal much credibility, or to assume that it will drive any benefit to vendors.

Reading the CES Tea-Leaves

The kick-off of the Consumer Electronics Show this year may be more meaningful for tech than usual because it’s a barometer of some critical market dynamics.  Tablets are set to take the hot seat at the show, even though (as usual) Apple isn’t attending.

The big question in my view is less whether we’ll see a zillion tablets than whether we’ll see tablets focusing on both a smaller (7 or 8-inch) form factor and WiFi-only connectivity.  The tablet as a satellite of a 3G/4G mobile service plan isn’t going to transform the market because the cost will be too high for most users.  Sure they might end up with value, but why take a risk?  On the other hand, an “uncoupled” WiFi tablet that’s affordable becomes the instant device of choice in hotspots, particularly hospitality sites.

Most users (83% according to my model) plan to use tablets primarily at home, at work, or in a setting that’s likely to have WiFi.  An even larger percentage would prefer WiFi access to a wireless subscription tie-in.  Tablets and WiFi would have a major impact on the ebook space, creating a device that could become a universal reader, devaluing the incumbency of both Amazon and Barnes & Noble, and making Google and even Apple happy by validating a more general model of device.  Google, who’s trying to get its own book program going, might be particularly gleeful, and of course as the Android backer they have some influence on market direction.

At least a couple of the tablets at CES will surely be WiFi; Vizio already says it will launch an 8-inch Android tablet with WiFi only.  If we see a lot of this sort of thing, it means that the appliance vendors are looking to drive the market.  If not, then it means that they’re not prepared to step on the older partnership with the wireless carriers, which in turn means they’re not fully confident about the tablet future.

CES this week will start the drive toward the next stage of the media/network relationship.  The more pressure created by new tablets and new integrated TV delivery systems, the greater the pressure on ecosystemic tuning and market consolidation.  It won’t revolutionize viewing, but it could revolutionize the industry that delivers it.

The New Year, the New Ad?

The new year is always a time of perceived change, though of course the simple transition between two calendar dates doesn’t drive change itself.  Rather than talk about the coming year in general (which I’ve done in our Annual Technology Forecast issue for Netwatcher in any event), I’ll focus here on the immediate “changes” the industry is dealing with.

The comment over the holidays that Facebook has overtaken Google in popularity is a potentially seismic shift, but not for the reasons that have been suggested.  Yes, this is “bad” for Google, and yes, it shows the “power” of social networking.  But the real problem is that it may show that our use of the Internet is getting harder to monetize.  Somebody searching for “HDTV” is very likely to be thinking of buying one, and thus there’s value in selling ads to them.  Somebody chatting on Facebook is another matter altogether.

First, there is no easy way to establish the commercial goals of a Facebook comment.  Not only is it hard to interpret whether a reference to “HDTV” is soliciting buying advice versus talking about moving furniture around, it’s certainly an intrusion if Facebook started scanning our text for ad opportunities.  Thus, it’s very likely that social-network advertising will be more like banner ads, which are less visible, less clicked, and less valuable.  We’re shifting users to a place where fewer ad dollars are likely to follow.

Second, social networking is the framework for viral campaigns, not for direct ad sales.  If you want to leverage Facebook or Twitter or whatever, you have to start buzz and then let social networks propagate it.  That could generate more clever YouTube commercials but it’s not likely to generate direct ad sales.

Third, it’s an indication that in an effort to find the Next Great Thing, we’ve left the “great” category completely, at least insofar as creating and sustaining the Internet ecosystem is concerned.  You can argue that search provides a value to the Internet overall, a value of organizing and finding stuff in the vast repository of Internet data.  You can argue it promotes education.  It’s hard to make either argument about social networking.  I’m not saying people don’t like it, or that it is culturally revolutionary, only that it’s not moving the Internet to a better place in terms of sustainability.

For Google, you can argue that social networking is something that you can lose in by missing out on it, or lose by capitalizing on it.  If Google contributes to a social-network war that adds further to the time spent on social networks instead of on searches, then Google is contributing to a shift of focus away from profitable search ads toward less profitable display ads, and moving away from its own strength.  Would it be better to sit this out and let nature take its course?  Remember that Second Life was a craze a while ago, and most have forgotten it completely.  Not to mention MySpace.  Twitter is used by less than 10% of the online population, according to a recent study.

But if you sit out social networking as a Google, you ignore the insidious truth here, which is that the “growth of the Internet” is growth in increasingly non-commercializable areas.  We are likely seeing the inevitable plateau in online ad revenue approaching.  It’s not that ad revenues won’t grow, but that they’ll not explode in new areas where people can make a ton of money, and that means VCs and growth companies like Google will have to look elsewhere to make a killing.