The smartphone wars: Google goes Taoist, Microsoft uses the farce

A few days ago, I observed that Google is not pushing back against cell carriers’ lockdown of Android phones as vigorously as I had expected in the wake of the Android 2.2 announcement. If this Twitter rumor is true, Google locked horns with the carriers and lost that confrontation, leading to the semi-discontinuance of the Nexus One. Does this mean the carriers have won the war?

I think not. The good news is that the announcement of the T-Mobile G-2, which will run un-skinned Android, suggests that Google’s longer-term strategy is still working. I have noted that the opportunity cost to the carriers of these unhelpful customizations is rising, propelled upward by increasing time-to-market pressures that force them to either drastically compress development schedules or run down-version releases of Android that sacrifice performance and customer appeal.

I still think the carriers will hang themselves, given enough rope – and that the third phase of the smartphone wars, in which carrier efforts to tame Android collapse of their own weight, is about to begin. I’ve argued this much in a previous post; my point now is that Google’s smart move now is Taoist. They don’t need to fight the carriers in any other way than by keeping the Android release tempo up and the time-to-market pressure on the carriers correspondingly high.

One of my commenters asked how Android customization by handset vendors (as opposed to the carriers) changes this picture. Not much at all, actually; if anything, the time-to-market pressure is worse on them than it is on the carriers, because they have more competition. Most places, the carriers’ markets look like oligopolies, but the handset market has more scrappy small players – and this makes complete sense, since the financial mass required to build and maintain a cell network is so much larger than you need to hire some engineers and rent fab capacity.

The difference in time-to-market pressure has consequences; expect the handset vendors to try to shed the cost of customizing Android faster than the carriers do. Notably, I don’t think HTC Sense will survive very long, especially not if the G-2 (which is expected to be an unskinned HTC Magic) does well. Crap like Sprint’s NASCAR branding will outlast it, because NASCAR will actually pay the carrier for that placement and offset Sprint’s development costs.

Of course, as the handset vendors stop customizing Android, the carriers’ cost of differentiating Android will go up faster. There’s only one way that game can end, really; carriers lose, and Google wins. The carriers can’t get off the tiger any more, not with Android’s new-unit sales growth utterly demolishing every other smartphone OS worldwide.

Which brings us to Windows Phone 7. Microsoft actually had the audacity to throw a shipping party that included a mock funeral for the iPhone. All that really needs to be said abut this is that the event was recorded by an Android phone.

I never thought I’d hear myself say this, but…I find I’m actually pitying Microsoft a little. They’ve got nothing, no hand at all in this game. Their desktop-computing monopoly not only won’t help them gain share, it’s an actual hindrance; carrier and handset-vendor execs understand precisely what Microsoft did to the PC market and are determined not to be Ballmer’s bitches. Microsoft’s previous mobile products have been a string of money-losing duds followed by an outright disaster – the wicked knock on Windows Phone 7 you’re going to keep hearing is “next of Kin“.

Brand power? It is to laugh; if you ask a random consumer in 2010 whether “Android” or “Microsoft” has more positive associations, Android will win hands-down (high marks to Google for effectively flooding that zone). Really Microsoft’s only asset is financial mass; if it’s possible for anyone to outright buy their way into the smartphone-OS market, they’re the outfit that can do it. Of course, this would involve forgoing actual profits, which is a problem now that their legendary cash hoard has largely been spent on fruitless acquisitions and stock buybacks.

No, Google has nothing to fear from Microsoft…and isn’t that an indication of how much the world has changed?

45 comments

  1. The thing that will really kick everybody in is when there are a ton of android2.2 devices that are no longer on contract because their 2 years got done and the carrier sent you another, faster machine. The poor old two-year old device, no longer tethered to a contract and GSM network, will still be a capable 1 ghz machine with touch screen, wifi, camera, music player, and all that jazz. They will also be dirt cheap because the average consumer will dump them on their 5 year-old to “play mommy” like they do existing phones that have completed their contract. But hackers will get these things and root them with impunity.

    Then, hello mesh networking with wifi.

  2. I previously stated that I don’t expect the oligopolies to fail until the granularity of customer choice exceeded the carriers’ capacity to control, specifically that the security model of the device changes to a provable micro-kernel with full disk encryption, so that the firewall security theater ends and the browser is free to control the device entirely. The point is that if that model of free market law proliferates on the desktop and the carriers block it, then mobile evolution will stagnate relative to the desktop and Coase’s theorem will destroy the carriers, e.g. WiFi mesh networking. Afaics, the proliferation/granularity of Android versions and handset choice does not trump the carriers’ control over the popular networks, because the carriers can choose which handset configurations to offer and the handset vendors must compete to be chosen. Apparently, Mr. Raymond acknowledged the theory that the carriers which control the popular networks do not need to compete to be chosen. The key distinction is that the Android and handset choice is not changing in real-time. The carriers won’t be able to cope with real-time change in configuration.

  3. > Google locked horns with the carriers and lost that confrontation, leading to the semi-discontinuance of the Nexus One.

    *semi*-discontinuance? Its all but over for the N1. Its only real life is as a dev phone, or in a few select European markets.

    > Does this mean the carriers have won the war?

    No, but they won that battle, while Apple continued to gain strength.

    Android is eating Microsoft’s lunch, not Apple’s. Android is a licensed platform for OEMs like HTC, Samsung, LG, Motorola, etc.
    iOS is not. The question is simple, if Windows Phone 7 doesn’t take market share away from Android, who’s going to make Windows Phone 7 devices?

    Microsoft was playing a sophisticated game with its mock funeral for the Blackberry and iPhone. It was telling the public, “Pay no attention to that man behind the curtain! Don’t look at Android, don’t think about Android. Concentrate on the death of the iPhone.”

    Meanwhile, Oracle is suing Google over Android.

    You’re so blind sometimes, Eric. So very fucking blind.

    Why you’re unable to see this is both unusual and questionable.

  4. So phase 2 was (Carrier) Empire Strikes Back, and now the Taoist Jedi at Google triumph?

    If I didn’t know better, I’d say Eric was having a funny with this…

  5. HTC is in the best position to benefit from this, but a clear way forward is to sell unlocked devices directly to consumers. The ONLY way for this to work in the American market is as follows:

    1) make a hybrid device with a radio capable of AWS and 850/1900mhz (frequencies to run 3G on both T-Mobile and AT&T). While you’re at it, why not make the frequencies programmable, and add CDMA and LTE hardware as well. Make a single device capable of running on any of the 4 big networks. I think WiMax 4G can be safely left out for now.

    2) Make deals with large resellers, like Walmart and Amazon.com, so that said devices can be sold subsidized THROUGH THE RESELLER!

    3) Make 4 different devices that have essentially the same internal hardware, but each with a different form factor: slate, portrait-slider, landscape-slider, and qwerty-board (Blackberry-esq). Give them all the same CPU, GPU, RAM, internal storage, etc. Make it so the ONLY thing the user chooses is the form-factor.

    4) Make 2 versions of each: a really cool, super-slick looking unlocked stock-Google device, and a slightly more boring version to sell to the carriers letting them do whatever they want.

    5) Make the carrier versions super-easy to root/unlock. Heck, seed some hackers information to make a simple one-click hack that unlocks, roots, and de-crapifies the device.

    1. >While you’re at it, why not make the frequencies programmable, and add CDMA and LTE hardware as well.

      Not gonna happen. Something I learned recently is that there’s a direct tradeoff between transceiver efficiency and programmability – trying to make a phone that flexible runs up the unit costs and hurts battery life, that’s why multiband phones are expensive specialty items. Your scenario of some handset maker cutting deals with WalMart seems plausible, though.

  6. Also, as far as Windows Phone 7 is concerned, I wouldn’t discount it completely, particularly with one demographic: hard-core gamers. The total integration with XBox Live and Windows Live services is the killer app. It’s how they best Sony even with inferior hardware that dies in less than 10 years — it’s the Live! network which is massively superior to what Sony can offer.

    This doesn’t mean it’s going to “win”. What I think it does is guarantee Microsoft a niche. I see the Microsoft eventual market-share to be similar to Apple’s eventual share — 5% of the market owned by hard-core gamers (instead of hard-core fanboys). My prediction for 3 years from now:

    1) RIM will be dead
    2) Nokia will have abandoned both Symbian and MeeGo
    3) Apple will have ~5%
    4) Microsoft will have ~5%
    5) WebOS will have 30% of the printer market, and probably 5-10% of the smartphone market
    6) Android will have most of what’s left over, but you’ll still have to root your device to have it not suck in most cases

    1. >My prediction for 3 years from now:

      Supposing we grant your premise that Live gaming gives Microsoft a toehold with the gamers, this all seems quite reasonable and much like what I’d predict myself – except I’m not clear why you expect WebOS to be have that much share. I wouldn’t mind if it did, mind you, I just don’t see a lot of strength there.

  7. Android is eating Microsoft’s lunch, not Apple’s. Android is a licensed platform for OEMs like HTC, Samsung, LG, Motorola, etc. iOS is not.

    Not entirely true. Microsoft used to have some share in 2 markets:

    1) enthusiasts who wanted full control over their phone (Android got this market)
    2) business users, due to excellent exchange integration (RIM now has most of this market)

    Most hardcore gamer-types today now have iPhones. This is the primary demographic they seem to be targeting now, and business users I could see them targeting with their next major push, as who is in a better place to offer better Exchange integration than Microsoft? RIM is dying quickly. Android is, as of now, really poor in this market — I had to spend $20 on an application to do proper exchange stuff on Android, and it still isn’t great. What happens when Microsoft adds some sort of feature like allowing admins to force laptops to authenticate through a tethered Windows Phone device, and use that device as its VPN? Many, many corporations would LOVE that sort of lockdown — only allow authentication against Microsoft ActiveDirectory over the VPN — they don’t do it now, because then business users would effectively have bricked laptops when given sales presentations. Make the whole experience seamless and centrally controllable, and Microsoft shops will buy these things for their employees.

    There’s also expanding market share without taking anything away from Google OR Apple in terms of actual units. The percentage of people using smartphones is going to go up. It’s a market that will be growing for a while regardless.

  8. I don’t know much about Taoism so I’m not sure what you mean when you describe this strategy as Taoist. But it does call Sun Tzu to mind. Sit by the river…

    1. >I don’t know much about Taoism so I’m not sure what you mean when you describe this strategy as Taoist.

      Achieving your goals by refraining from action is very Taoist. There’s a vulgar misinterpretation of Taoism that passivity is all there is to it, but the Chinese usually translated as “refraining from action” can also be read as “acting in accordance with the law of nature” or “minimum action”; the English idiom “going with the flow” captures the sense pretty well, I think. Act with minimum energy for maximum result.

  9. Supposing we grant your premise that Live gaming gives Microsoft a toehold with the gamers, this all seems quite reasonable and much like what I’d predict myself – except I’m not clear why you expect WebOS to be have that much share. I wouldn’t mind if it did, mind you, I just don’t see a lot of strength there.

    In my opinion, WebOS has the nicest UI of any smartphone OS, particularly when it comes to multitasking. It also, to me, appears to be the best UI fit for the 7″+ tablet form factor. Though it lacks a fair number of key features and a vibrant developer community, it is designed very well, and has a high degree of polish. HP, since buying Palm, is going inject a good bit of cash into it. I’m hoping (because more players in the market is better for everyone) that they succeed. In fact, my gut tells me they will.

    Of course, I could be wrong.

  10. Not gonna happen. Something I learned recently is that there’s a direct tradeoff between transceiver efficiency and programmability – trying to make a phone that flexible runs up the unit costs and hurts battery life, that’s why multiband phones are expensive specialty items. Your scenario of some handset maker cutting deals with WalMart seems plausible, though.

    Serious bummer.

    Do you, by any chance, have any references on this? I’d love to read up on this myself.

    Also, is this a hard physics limitation thing? Or is it likely to be solveable by Moore’s law?

    Finally, what if there was a successor to the SIM card which contained the transceiver on it? Would that be possible? The phone itself provides the antenna, but the network provides the transceiver and CODECs, etc. all on some sort of super-SIM card. I’m asking because I honestly don’t know…

    1. >Do you, by any chance, have any references on this? I’d love to read up on this myself.

      Reread the comment thread on Smartphone, the Eater-of-Gadgets, which featured Russ Nelson and me getting schooled on the fact that modern digital radio transcievers don’t work on the heterodyne principle.

      >Also, is this a hard physics limitation thing? Or is it likely to be solveable by Moore’s law?

      Alas, yes, and alas, no. But it’s possible my understanding is still defective.

      >Finally, what if there was a successor to the SIM card which contained the transceiver on it? Would that be possible?

      Don’t know. Don’t know enough about the technology in SIMs.

  11. re: radio programmability (@esr, @Aaron Traas):

    The obvious solution to the programmability vs. efficiency problem is to modularize out the radio part. Make a universal radio interface that will accept a pluggable radio. Probably not that hard to do, but IANAEE.

    1. >Additional thought: the carriers wouldn’t like that very much, would they?

      The carriers probably wouldn’t really care much if they still had control of the phone software. They don’t make money on handsets; we know this because they subsidize them in order to lock customers into term contracts. What they don’t want to see is large numbers of American consumers adopting pay-as-you-go sims – their opposition to unlocked phones is derivative of that. Whether the phone is single- or multiband is thus only a secondary issue.

  12. @AlanR
    > The fact that they think their competition is the iPhone only shows how out of touch with the market they are.

    There is that, but I also think there is still a great deal of institutional inertia against the whole concept of open source at MS, despite the history of the last 10 years. “Android is open source, which is laughable, so we’re not even going to consider it*”

    * At least not in public

    1. >“Android is open source, which is laughable, so we’re not even going to consider it*”

      It might surprise you, but I actually don’t think this is true. I think Microsoft’s internal culture got quite the salutary shock about that in the late 1990s. They certainly see open source as the enemy, but I think they’re long past dismissing it as a trivial one.

  13. “Of course, as the handset vendors stop customizing Android, the carriers’ cost of differentiating Android will go up faster.”

    Has anyone ever saw even one customer who actually gives a damn about that, or really about anything else in selecting a carrier than 1) price per minute or GB 2) available devices and their prices 3) reliability of service 4) quality of customer service?

    1. >Has anyone ever saw even one customer who actually gives a damn about that

      I think there is a relatively small minority of customers that do care about more than low price and whether they can get to a tech with a clue – for example, for me having a web browser with a decent display in my pocket all the time is really important.

      But I think your implied general point is valid. The carriers are mostly fooling themselves if they think they can differentiate on features, and one effect of Android is to make that reality brutally clear.

  14. >> While you’re at it, why not make the frequencies programmable, and add CDMA and LTE hardware as well.

    > Not gonna happen. Something I learned recently is that there’s a direct tradeoff between transceiver efficiency > and programmability – trying to make a phone that flexible runs up the unit costs and hurts battery life, that’s
    > why multiband phones are expensive specialty items. Your scenario of some handset maker cutting deals
    > with WalMart seems plausible, though.

    Jesus Christ, what you don’t know about radio would fill a 20 volume set.

    The original poster is wrong, but not for the reasons you give.

    “making the frequencies programmable” means putting a wide-band receiver architecture on-board, but these are less sensitive, and therefore, less performant.

  15. “I think there is a relatively small minority of customers that do care about more than low price and whether they can get to a tech with a clue – for example, for me having a web browser with a decent display in my pocket all the time is really important.”

    That belongs to my 2), “available devices and their prices” – display size is device-dependent and the web browser is part of the basic package shipped by the device vendor which the carrier might twist this way, that way, add bells and whistles which may slow it down, but I don’t think they ever dare to remove or seriously cripple it.

    My question / rhetorical question was about the customizations added by the carriers, that they aren’t a competitive advantage.

  16. >The obvious solution to the programmability vs. efficiency problem is to modularize out the radio part. Make a universal radio interface that will accept a pluggable radio. Probably not that hard to do, but IANAEE.

    Engineering wise, it’s (sort of) been done. When RIM wanted a handset for Nextel’s iDEN network, Motorola built a modem/radio card to fit their specs because they were the only folks with iDEN expertise. Of course, that was still a purpose-built handset–you couldn’t buy a different radio card and plug it in. There are some technical issues like antennas, which are band-specific and tend to be tightly integrated with the handset design (ex: iPhone 4!), but I bet those could be overcome by a determined manufacturer with a compelling business case.

    However, there’s probably a regulatory roadblock here. I’m not an expert, but my understanding is that it’s hard to get FCC certification for anything but a fully-functional handset. The various subsystems of a handset interact in strange and complex ways (consider that even the handset molds and plastics can affect thermal dissipation and radiation patterns) so the FCC usually demands emissions tests on the sum of the parts. I know the FCC will certify WiFi/Bluetooth/3G cards separately from the laptop/netbook/tablet they’re intended for, but you’re usually not holding those devices against your cranium. ;)

    There’s also the cost issue. I suppose a manufacturer could build their own handsets and a collection of radios and then certify the cartesian product, but I imagine that approach would add several dollars of cost (connectors, mechanics, n*m certifications) to something with precious little margin. You might be willing to pay a few extra bucks for the possibility of changing networks, but your carrier sure as hell won’t subsidize that model.

  17. @Some Other Guy: “Jesus Christ, what you don’t know about radio would fill a 20 volume set.” That condescension isn’t winning you any friends, you know? It doesn’t work for me when Eric does it, and it won’t work coming from you either.

  18. However, there’s probably a regulatory roadblock here. I’m not an expert, but my understanding is that it’s hard to get FCC certification for anything but a fully-functional handset. The various subsystems of a handset interact in strange and complex ways (consider that even the handset molds and plastics can affect thermal dissipation and radiation patterns) so the FCC usually demands emissions tests on the sum of the parts. I know the FCC will certify WiFi/Bluetooth/3G cards separately from the laptop/netbook/tablet they’re intended for, but you’re usually not holding those devices against your cranium. ;)

    The situation with the FCC is not quite as you describe, but close.

    SOGuy is correct, through abrasive about it.

    See: modular approval and SAR

  19. Something I learned recently is that there’s a direct tradeoff between transceiver efficiency and programmability

    I wonder if a transceiver module could be packaged somewhat like a SIM card, so that the user could simply pull out one and put another in as part of changing carriers. This would have the added advantage that if the FCC releases some VHF spectrum to the cell carriers, all you’d need is the module to exploit it, without having to replace the entire device.

  20. Well, here is the first step to your Walmart theory. This is from WSJ about Walmart’s new post pay, no contract, unlimited voice and text cell phone service. It runs on T-Mobile and there is one (sucky) Android phone already, the Motorola Clique XT. Unfortunately, data is sold in ‘buckets’ and looks rather expensive at $40/mo for 1GB. See the story for more details on the data rate.

    1. >It runs on T-Mobile and there is one (sucky) Android phone already, the Motorola Clique XT.

      Bletch. Android 1.5 with Motoblur and some undeletable apps. On the other hand, since T-Mobile was hawking it for $129 at launch, it’s going to be less than that – I wouldn’t be surprised if WalMart pegged it at $99.95. That would be within shouting distance of Gassée’s magic $89.

      Oh, the other carriers are so going to hate this. Heh. It’ll be fun to watch.

      I wonder if we’ll see a parallel deal with another major chain retailer soon. Sprint selling through Target, maybe?

  21. I wonder if it would be possible for the radio itself to eeparate from the handset, and connected by both wired and wireless bridge, so that radio could be in pocket and handset be in hand. I’m thinking something like Sprint’s mobile wifi access point. Such radio could then be switched out more easily. Thoughts?

    (typed on my HTC Incredible Froyo soft keyboard)

  22. esr>>Also, is this a hard physics limitation thing? Or is it likely to be solveable by Moore’s law?

    >Alas, yes, and alas, no. But it’s possible my understanding is still defective.

    In at least one way, this is a physics problem: the antenna. Anytime you make an antenna that covers a wider range of frequencies (a broader band), you get more resonant “leakage”, wasting more power out and in. Narrow-band antennae generally are much easier to design for efficiency, and are much more effective than broad-band antennae. To compensate would require a higher power transmitter and more sensitive receiver (generally diametrically opposed objectives due to heat-related issues within the radio), which will eat up batteries at ridiculous rates and make it harder to pass FCC safety qualifications for radiating devices.

    All that said, it’s not impossible, it just isn’t gonna happen tomorrow.

  23. esr> Sprint selling through Target, maybe?

    They already have a helluva campaign with Rat Shack.

  24. >If Steve Jobs were really skilled in ninjutsu, surely he would have been able to get those
    >shuriken on his own private jet without a problem.

    Such as:

    “I don’t have any throwing stars on me, I checked those in with my baggage.”
    “But Sir, you very clearly have them, they were found on your body and are right here before me as we are talking.”
    “Of course. I’ve appointed myself to the role of the check-in agent and baggage handler of my plane. I checked them in, handed them over to the baggage handler and just about to put them into the baggage compartment. Problem…? {trollface grin}”

  25. Interview with VP for Windows Phone 7. Development kit is free and runs on Windows PC. This won’t appeal to developers that want open platforms, but it will appeal to integrators. Windows still has a huge market share on PCs and in console gaming.

    I am leaning more and more to Nokia will adopt a multi-prong strategy to illuminate their areas of strength, with one prong being Windows Phone for some of their smart phones, to compete against Android with higher quality at low price points. I expect Nokia will leverage their economy-of-scale, and upsell higher margin offerings. I think they will continue to develop Meego as plan B to see if they can differentiate with it on any economic and market factors.

    The more I think about it, I see no chance that Nokia willingly adopts Android, that would be handing the market to their competitors.

    All this competition is great for developers and users. I prefer Android does not attain a runaway majority, at least until it is completely open sourced.

  26. The more I think about it, I see no chance that Nokia willingly adopts Android, that would be handing the market to their competitors.

    And adopting Windows is not handing it to Microsoft? Nokia would likely have even less control over the software.

    With the jostling now going on inside Nokia, I something like what you suggest just might happen. I think it would be a disaster, but if MeeGo isn’t ready in time, they are in a pinch of sorts.

  27. > And adopting Windows is not handing it to Microsoft? Nokia would likely have even less control over the software.

    Nokia has to choose between giving control to Microsoft or to Google. I say from their perspective, the lesser of two evils is to not give their competitors a popular open source. I also think it is in their advantage to delay Android’s complete dominance while they try to get MeeGo ready.

    Also they may be able to negotiate some concessions with Microsoft, such as ability to share revenue on games, some thing they can’t get with Android’s open model of app marketing.

    I hope Nokia does not go to Android, because I don’t want it to be a 2 horse race between Google and Apple. There is not much danger of Windows Phone gaining market monopoly, so I actually see it as a positive for competition. It might give a chance for Nokia to come out MeeGo phones later, along side their Window Phone phones. Or maybe even a dual boot option? Who knows, I just love it when there are more possibilities, makes it more exciting.

  28. Also they may be able to negotiate some concessions with Microsoft, such as ability to share revenue on games, some thing they can’t get with Android’s open model of app marketing.

    Not completely true. They could create their own app market which would give them the 30% that google extracts from each purchase from google market. If they went that route i’d suggest they provide a way to install their market app on any android phone, Nokia has enough of a name in the mobile space that i’d seriously consider publishing with them.

    What may end up being a killer app is a tool for app developers that allows them to publish to multiple app markets simultaneously. That way consumers don’t necessarily have to suffer for only being on one market.

  29. It’s not flash compatible yet although It’ll work with flash when Adobe has made it compatible. Heck, they’re just now rolling out Adobe flash player for Android. It’s saving the current state of the apps so when you open it one more time, it’ll be at the same state.No copy and paste. Hopefully Copy and paste will be replaced by smart linking. I hope there will be adding full copy and paste functionality in the next weeks.

Leave a comment

Your email address will not be published. Required fields are marked *