The Hole in Android and Google’s Double Pony Problem

Android is on fire and Gartner predicts it will be the number two mobile operating system worldwide this year, surpassing Apple and RIMM, but behind the seemingly immortal Symbian.  Google embraced the ubiquity strategy and it is working.  But they’re getting a free pass on whether it makes money on the assumption that Android handset volume will eventually drive material search queries, advertising revenue and pull other attached services.  Unfortunately, there is a big hole in that Android business strategy, shaped roughly like this:China

Google’s self-immolation of its China presence means they won’t see much mobile (or any) search revenue in China, the world’s largest mobile market (and home to the largest number of Internet users).  Google’s mobile search share in China dropped by 30% in the second quarter and they’ve already fallen to third place (bonus points: who is second?).  Android stalwart Motorola is using Baidu and Bing on its phones in China plus there are a variety of efforts by Chinese operators and handset vendors that fork Android.  Forking sidesteps the remaining Android constraints altogether and of course provides complete discretion for what services are integrated.  So you can cut Android’s expected revenue per unit by roughly 20% just based on China.

And where China goes, others may follow in decoupling Android from Google search.  In countries with strong domestic search engines like Russia and South Korea, it may be a simple matter of consumer preference.  The more dirigisme (I’ll just note it is a French word) may not be able to resist the opportunity to play with search defaults.  And in the US, Microsoft is persuading Verizon to use Bing for Android phones with what looks like just cash.  There is a real risk of further decoupling of Google search from Android.

Now Google may be content with not monetizing Android due to its other strategic benefits.  Android pressures Apple and Microsoft, significantly disrupts the traditional operating system business model (which we may soon see extended to tablets and netbooks, which will be really interesting to watch) and raises the capabilities bar for the mobile web.  But settling for non-monetary strategic benefits when the guys you’re outselling are making billions is a little embarrassing (admittedly they’re making it from hardware).  I know Google is monetizing Android on the sly around the edges (it turns out Android is not so open and free if you want the latest version and the Skyhook lawsuit suggests some other tying shenanigans) but it is a rounding error from the standpoint of a $25 billion company.

Google is stuck between two Pony problems.  The One Trick Pony problem and their need to find another material revenue stream beyond search looks more pressing as both their search share and their revenue growth flatten out.  Their heyday window to make hay by building additional businesses while on top of the world seems to be coming to a close (life at the top is getting shorter and shorter – we’ll see how long Facebook lasts in that position.  They could peak even before they become a $10 billion revenue company.  Deferring the IPO for as long as possible makes a lot of sense for them to maximize their window).  Android is one of Google’s better candidates for a revenue stream with lots of zeroes after it, but we are already seeing multiple examples where Google’s revenue link to Android is being severed.  This could be described as the My Little Pony problem (a Sun Microsystems reference for those too lazy to click through and parse the obscure video), wherein your free software doesn’t drive significant revenue directly or indirectly, even as others go to the bank on top of your efforts.  As Google’s core business matures, they’ll have less and less ability to make grand philanthropic efforts.  I suspect we’ll see free become less free and Google dare phone manufacturers to shift platforms once they have started down the Android path.

The good news is neither of these problems are mine to solve.


I’ve gotten multiple requests to dance on Sun’s grave, but Fake Steve seems to have the ceremonies well in hand.  I just need something from the wine cellar to accompany the festivities.  Perhaps a nice bottle of Churlish Chardonnay…


Mobile Photo Feb 7, 2010 11 29 55 PM (What ever happened to that Java ring?)

Mobile Photo Feb 7, 2010 11 46 53 PM

Mobile Photo Feb 7, 2010 11 47 03 PM

This was an Intel product if I remember correctly, underscoring how tough it is to pair chips and wine.


image Proposed stock symbol: SORE

Customer line: “I’m a SORE customer…”

Rejected names: Sunacle?  Too close to Unocal.  Orasun?  Sounds like a cloying dental product.  Orsun?  You could easily swap Ellison for Hearst in the remake of Citizen Kane…

Ok, so this isn’t a merger of equals.  It is the end of Sun.  Only Oracle survives.  But this deal surprised me.  I figured if Oracle wanted Sun, their well-oiled M&A machine would have swooped in a while ago rather than waiting for IBM to set things in motion, though there are rumors that Oracle made a joint bid with HP last year.  They certainly could have gotten a better price.  Did IBM get pulled in to play the second bidder?  The extra dime Oracle is sharing is not exactly the sign of a bidding war.

Quick thoughts:

  • MySQL – you’d never know Sun owned it from the Oracle press release.  Interesting market definition problem for the antitrust authorities evaluating this transaction.  With MySQL’s European heritage, maybe we’ll get some more novel legal theory from the EU.
  • Solaris – Oracle has been treating it as a legacy platform in recent years and it is hard to see Oracle being any more successful with Solaris vis-a-vis Linux than Sun. After all, Oracle’s Linux relationship is unbreakable…  But the acquisition means Solaris can reestablish itself as Oracle’s choice for high-end deployments.  Not clear whether Solaris remains a go-to platform beyond the database.
  • Sun’s Hardware Business – this definitely gets flipped to HP, Fujitsu, an Asian up-and-comer or someone else.  Sun doesn’t have the share and Oracle has too much invested with partners to go to the vertically integrated model (not even IBM can sustain the vertically integrated model and they invented it).  Oracle must already have a plan here (is there another shoe to drop with HP in the next few days?).  The hard part is how to minimize the atrophy during the long road to closing the deal.
  • SPARC – Oracle wants nothing to do with this.  Maybe Fujitsu will take it.
  • StorageTek – hope someone does the math on how much value got destroyed with this acquisition.  Sun bought this company just as disk became cheaper than tape.
  • OpenOffice – this deal could be a big boost for OpenOffice, as Oracle can’t resist an opportunity to tilt at Microsoft, especially with Microsoft driving more and more integration between the Office suite and the back-end applications that are Oracle’s bread and butter.  Oracle can provide OpenOffice a value proposition beyond price.
  • IBM – they got a good look at Sun, but in the end I suspect they’re going to regret not doing this deal.  They regretted giving Microsoft control of the software crown jewels for the PC; they may face similar situation now on the server.  Oracle can have a lot of fun making IBM choose between  open systems sanctimony and controlling their own destiny.  That of course assumes there is more relevant innovation to be had around Java.  IBM should have figured out how to get through regulatory approvals or how to cut the company up into piece parts to get what they wanted.
  • Integration Risk – this is a very different deal than the other big acquisitions Oracle has done. Sun customers are skittish as is and there is minimal maintenance gravy train here.  Whereas Sun offered IBM access to a bunch of net-new customers, I suspect Oracle is already in all of Sun’s accounts.  Serious layoffs ahead at Sun regardless of how it plays out for Oracle.
  • Consumer/Embedded Java – not exactly high on Oracle’s priority list.  Maybe it finally gets liberated in gesture of openness/misdirection by Oracle.  Maybe they can sell it to Google.  It fits nicely with the Android (lack of) business model.
  • Open Source – will be interesting to see if any of the projects Sun open-sourced get fork and/or critical mass.  Some argue Sun has already lost control of MySQL.
  • Timing – I need to go back and look at the sequence, but Oracle has been pretty good since their acquisition parade began at timing deals in such a way that they help their year-on-year comparisons.  Never forget Oracle is managed financially these days.  This might explain why they didn’t pull the trigger on Sun earlier.

What did I miss?

From an Undisclosed Location

Scott McNealy's current office at Sun MicrosystemsSomewhere, the team of guards that has been keeping Scott McNealy gagged and under lock and key in recent years has been doubled and they’ve taken to sedating him to ensure he doesn’t pop off about the prospect of IBM buying Sun and perhaps prevent Sun from being put out of their (and our) misery through acquisition.

The ever pointed McNealy had some subtle insights on mergers and acquisitions.  Two that pop to mind:

  • On the merger between HP and Compaq: “Two garbage trucks backing into each other in slow motion.”
  • On Unisys, a merger of mainframe also-rans Sperry and Burroughs, whose slogan was The Power of Two: “That’s about their stock price!”

I’m sure he has had some fine IBM barbs as well over the years but I can’t recall any (probably because we were to busy ducking his incoming fire at Microsoft to appreciate well-aimed shots at others).  Anyone remember any choice Scott lines about IBM?  An IBM takeover of Sun would undoubtedly be a long and protracted process so we need some choice color quotes to be repeated in every story for the months it would take to close.

A Little Sunshine

I’ve been meaning to post on Sun’s latest earnings, but have been delinquent.  It is harder to poke fun in light of their thousands of pending job cuts, but my post speaks to the root cause of Sun’s predicament: software.

I have often questioned Sun’s software strategy (here and here), going back to the fixation on Java as a competitive weapon as opposed to making it a good business, to McNealy’s “software is a feature of hardware” delusion to the “free, but we make it up on volume” philosophy that seems to be their current software strategy.

Despite repeated and lengthy lectures on the topic from CEO Jonathan Schwartz (who evidently is not among those losing their jobs), there just isn’t any discernable sign of a “virtuous cycle” to their software strategy, much less any financial evidence of it “expanding Sun’s market opportunity in the corresponding datacenters”.  If anything, there is evidence their software strategy is destroying their traditional datacenter revenue.

You can get the explanation from Schwartz, but honestly I find this to be at least as illuminating (and way more entertaining):

I have challenged Sun over the years to back up their software rhetoric by breaking out their software revenues.  Well, they finally did cast a little sunlight on their software business in their most recent financials.  It has garnered very little attention (hat tip to the Wall Street pro who pointed it out to me).  So allow me to commend Sun (feel free to pause and enjoy this rare event) for breaking these out, even if it is probably preparatory to the breaking up of Sun.  Here are the numbers for their most recent and previous fiscal years:

  FY07 FY08 Growth
Java $219m $220m 1%
MySQL/Infrastructure $198m $208m 5%
Solaris, Management, Virtualization $221m $216m -2%
     Total $639m $644m 1%

Software is less than 5% of Sun’s overall revenues.  And the absolute numbers are smaller than I would have guessed, especially the Java number.  For all the noise about Java, it is only a $220 million business?  I wonder what the trend line looks like over the last ten years.  Was it ever an appreciably larger business?  Obviously, there are others making much, much more on the back of Java.  Not much seems to have come from all those app server acquisitions (which I’m not sure I can even name any more, but there were multiple nine figure acquisitions).

It is also interesting to compare to other current software businesses.  Overall, Sun’s total software revenues are roughly the size of Red Hat, but the Solaris line item obviously is about a third the size and shrinking.  HP’s software business is over four times as big as Sun’s – that must rankle (HP = “House of Printers” to recall a memorable McNealy line).

Most embarrassing, the 1% annual growth was such outstanding performance as to put software into Sun’s portfolio of “growth” businesses.

So what happens next?  At one point today, if you netted out cash, Sun as a business was worth about $4 million (less with committed downsizing costs).  Maybe Fujitsu gets interested again as they unwind their Siemens partnership.  They would probably pay $4 million for SPARC.  But with the cash position, anyone could come along and cherry pick a very small asset out of Sun and walk away from the rest.  Maybe the econalypse (I am still looking for a good name for the “current economic situation” — leave any ideas in the comments) precludes this, but current Sun situation isn’t sustainable.

MySQL or My Bad?

I was wondering a few months ago how Sun’s MySQL acquisition would impact their relationship with Oracle:

How has the MySQL acquisition by Sun impacted the relationship with Sun’s biggest historical partner, Oracle?  I may be misjudging Oracle’s leadership, history and culture, but my guess is they view databases as their birthright and treat any real or proposed encroachment, even from a company with as poor an acquisitions record as Sun, as a serious matter.

The answer is becoming clearer:

  • Oracle has entered the hardware appliance business, partnering with HP
  • Sun’s stock is trading near a 13-year low
  • Sun’s CEO got a 44% pay raise

(Ok, the last one isn’t so clear).

The next question is what is the end game for Sun?  Do they circle the drain like Silicon Graphics (yes, they are still in business) or do they get consumed by the likes of Fujitsu or perhaps a rising dynamo from the developing world?  Leave your predictions as comments and maybe we’ll start a pool.

Utility Computing or Futility Computing?

Perhaps the network is not in fact the computer.  The Register reports that after 14 months Sun still doesn’t have any customers for their $1/CPU/hour computing service.


It is easy to pick on Sun, but they’ve made the same fundamental error that everyone touting utility, grid, on demand and other flavors of buzzword computing have made: the economics just don’t work in the wide area.  Sun seems to get the economics wrong more often than most (e.g. buying a tape company just as disk has become cheaper than tape), but they all suffer from the mainframe conceit that it is better to centralize processing even if that means the processing ends up far away from the task at hand.  That was once a good idea but not any more.  Timesharing has lost some of its pizazz.  Processing is relatively abundant (see Jim Gray’s Distributed Computing Economics paper again).  You better be doing a lot of processing to make it worth the round trip.  This is why it makes sense to interact with remote, high value services as opposed to shipping things out to be processed elsewhere.  The software behind the service actually does something useful and is close to its own data.  In the meantime there is more and more power on the edge of the network to consume and remix those services.


Sun says "In the long run, all computing will be done this way."  Keynes says we’re all dead in the long run.  Depending on your choice of strategy, the long run may come to pass sooner rather than later.