Mobile device OS upgrades: How hard can it be?

Until the iPhone’s arrival in 2007, upgrading the software on a mobile phone was a rare experience for users. So rare that effectively it was not done. Few people were bothered though since they did not see the product they used as a software product.

This was even true for Windows Mobile and Symbian which were licensed platforms. Microsoft tried several times to offer upgrade paths, but more often than not the device vendors did not push out updates or the process required to perform an upgrade made it the reserve of either those who were paid to do it or those who enjoyed the challenge.

In the era of the modern smartphone, upgrades are more common. Certainly with the iPhone the process is easy enough that opting out of an upgrade is more challenging than opting in. But it’s still not as common with other platforms. Even with all the resources and experience behind them, Microsoft is still stumbling with Windows Phone upgrading.

UPDATE 1-Microsoft explains phone software update delay | Reuters

But is it really a matter of blundering or is there evidence of nominal partners working at cross-purposes?

Upgrading a mobile device is technically challenging. There are many things that can go wrong. Integration helps, but it is also possible for a modular solution to be implemented. There are harder things to coordinate.

However, things go wrong and as the linked article above shows, fingers get pointed. Device vendors who have no stake in a platform can easily interpretan upgraded OS as one fewer hardware upgrade. Platform vendor’s lament that device guys want to proliferate unique versions and act with passive aggression. Device vendors see the software upgrade cycle as increased support and engineering costs with no upside. Consumers don’t know whom to blame and often turn to the operators since that’s who they pay every month.

The problem of upgrade failure is a symptom of a deeper dysfunction inherent in immature modular business architectures. It’s not just that the brightest at Microsoft or Nokia or Google can’t make an upgrade stick. It’s that the upgrade is not universally beneficial to the value chain. To remedy this, licensors have to resort to contractual obligations to ensure upgrades, but enforcement is non-trivial and can lead to aggravated relationships.

Unless all licensees of a platform find compelling value in shouldering the burden of upgrades, they will continue to be really hard. This leads to increasing fragmentation and a corrosion of platform value for licensed platforms vis-à-vis their integrated alternatives.

  • TomCF

    The argument doesn't seem that different than the argument for/against better quality assurance.

    • Companies can easily increase QA within their companies. It's much harder when you are trying to QA an ecosystem that you don't control.

    • asymco

      I'm vaguely familiar with the argument for better quality assurance but what's the argument against better quality assurance?

      • newtonrj

        Constant Vigilance – If an organization lacks sufficient QA but has constant updates where micro adjustments are normal to the user community, then the quality of the product is same/same with high QA. In point, some ecosystems desire more micro movement and less macro movement with high QA. The comfort level for customers to migrate increases with repetition to the level that the developer is able to overcome problems while also introducing new features.

        An organization with high migration comfort is able to overcome subordinate macro movers of bad migrations. Such an organization would be much more comfortable with what resources are needed to complete migrations, back-out faulty code and bridge services around faulted elements.

      • TomCF

        I should have stated that I was talking about what I've seen in the software industry: companies skimping on QA to get the software out the door. It's sometimes seen as a cost with no revenue attached to it. It often isn't sufficiently budgeted. Those companies that fail to plan well for QA also put out mediocre (at best) products. It's short sighted. The value of QA (in software at least) is down the road you get happier customers. The problem is that it doesn't add anything to this quarter's numbers.

        I didn't mean to imply the argument against QA is a valid one. 🙂 But it seems like that isn't stopping carriers…

    • chandra

      Compliance assurance is key. Quality is another issue altogether although important. But there are software updates (QA fixes) and there are upgrades which expand device capability (i.e. adding cut and paste for example) and that is not a QA issue.
      In any modular architecture system, there has to be a compliance roadmap that all partners sign up to. Look what Adobe did to Apple in the past and you can see the many risks of failure to satisfy the end user. Without collaboration, failure is inevitable. Microsoft will see it as failure through fragmentation. And if MS is slow to innovate with OS evolution, in consultation with its partners, they will be held back from introducing new hardware features until they are supported in the OS.
      I'm not surprised HP bought WebOS. In the future, collaboration will become a true bag of hurt for all concerned. In that game it is rare to find every player singing from the same song sheet. They are wanton and they are fickle and, as MS has serially demonstrated, they are far more likely to knife a partner than to operate congruently with them in the common pursuit of profit.
      There also has to be a lead partner – in the case of LG, Samsung and HTC etc it's Microsoft, for now. in the case of the MicroNokia partnership I'm not clear which is the dog and which the tail. There are two partners each contributing one of the two parts of the whole widget, but which is the lead partner and how much authority and control can they wield?
      A lead partner can set the terms of business and police compliance. Microsoft clearly dropped the (conductor's) baton on this and they reap the predictable outcome. Unless they can secure compliance, Windows phones will remain a minority interest. IF they get it right, there are enough dweebs on the MS and the Nokia camps to make it a contender for second or third place.
      I remember a case study on Airbus Industrie where it was shown that because there were many nations partnered in the aircraft building venture, different parts of each plane were built in a different country, by a different company. Software systems were also contracted out to specialist companies in different countries. It sounds like a recipe for a dangerous and lethal nightmare, but it worked. Split engineering in the hardware and the software and yet there were strict rules on all contractors.
      It's possible to roll out modular solutions but congruence and compliance are what make it possible. The trouble is that both MS and Nokia are used to having their own way in everything. And that is not a good place to start out from in any partnership.

  • Eddie

    Maemo5 on the Nokia N900 is an even rarer breed, as it's now receiving officially sanctioned updates developed by the Maemo Community[1] independently of Nokia. These updates are released every two to three weeks and fix bugs while also adding functionality Nokia failed to deliver.

    The most recent update[2] on 11 April (the eighth such release in 3 months) replaced the Camera UI with an open source clone (actually, a better version as it has more features but uses the Nokia original as a starting point), bug fixes and feature additions to the email client (Modest) plus bug fixes for various other standard components, while also adding Portrait mode support wherever possible.

    This, to me, is how it should be – updates drip fed as they become available, PC-style, not monolithic updates once or twice a year.

    Such a shame Nokia will be losing this kind of capability as they hurtle head-long into their closed source, hardware ODM future.



  • BlackBerry appears to have updated OSes quite regularly. Carriers do their testing and do an official release. The OS is then available on for the user to download. The user connects to BB Desktop and the phone gets updated with the new OS.

    BlackBerry also have regular OS "leaks", which are OSes released by BlackBerry (they carry RIM's digital signature countersigned by Verisign) but not yet released by Carriers.

    • asymco

      So what does that say about the quality of upgrade experience for integrated vs. modular device solutions?

    • RIM is proof for this argument not against. Blackberry is different because they charge for ongoing service revenue (so have motivation & reward to updgrade existing handsets) and their own NOC and platform to deliver over. The testing and validation required for their upgrades is no easier–witnessed by the often lengthy delays in releases–but they have different rewards and incentives than other OEM and OS partners mentioned here.

  • I've never had a Nokia update brick my phone like some WP7 users had and out of all the OSs, Symbian is the easiest with small over-the-air installs or individual components updated OTA instead of huge 400+Mb installs that have to be done via a USB cable on the iPhone.

    I'd echo Eddie's praise of the Maemo update process also. It's just easy although if you enable the maemo-testing repository you can be updating software every time you switch your phone on.

    The problem with Nokia though isn't the technology, it's the process. For instance, the PR1.1 update for my C7 started rolling out in early February, which was late as Nokia had announced it for the end of 2010. But then my carrier had to approve it first and they took 2 months. The generic C7 PR1.1 update for the UK actually took Nokia UK longer to approve than even my carrier. On Nokia's support forum, they had a spreadsheet listing the hundreds of carrier and national variants and their update status. Even now, some carriers in Australia still haven't approved the update. PR1.2 even came out BEFORE everyone had PR1.1. Utter madness.

    IME, the Maemo Community SSU rocks by comparison.

  • CndnRschr

    Perhaps by the time Nokia releases its Windows phones, Microsoft will have gotten its act together on OS updates? Nokia is certainly uniquely incentivized to be the class-leader in Windows Phone experiences and it looks like Microsoft is taking a pro-active stance on Nokia hardware development too (although this could be detrimental to overall Nokia design freedoms). Still, Microsoft's repeated inability to deliver seamless minor updates is more than embarrassing at a time when it needs to impress. Going for the long win isn't helped when you trip up over the first two hurdles.

    • asymco

      The point is not that Microsoft can't get their act together. It's that they can't ensure that the update is universally available to all users in a timely manner. They had trouble with Windows Mobile when it had thousands of SKUs and are still having problems with WP and a dozen SKUs.

      • CndnRschr

        That's really what I mean about getting their act together. They supposedly built Windows Phone 7 from the ground up (although it borrows a lot of code from WinMo and other frameworks). So why not get these fundamentals right on a second attempt? It's not as though they had *any* legacy hardware to deal with. It cannot bode well for the future if the first few wisps of the Windows Phone 7 ecosystem are so difficult to update.

  • r00tabega

    Apple's major innovation is cutting the carrier out of the upgrade process (with the exception of the baseband firmware which the carriers are very interested in).

    Seems that the major innovation for Apple is the force of will to negotiate and gain contractual wins that result in disruptive power and consumer friendliness.

    iPhone's complete ostensible lack of AT&T's branding and approval is one.
    iPad's contract-free data options are another.

    • carlitos

      "force of will to negotiate and gain contractual wins" interesting point.

      I can easily imagine an alternate 2006-2007 universe where operators were more prescient and both Verizon and AT&T balked at giving any ground to Apple's demands that we now take for granted as iPhone users. In this version of history the iPhone would launch on T-mobile (no 3G iphone in the USA for many years?) or Sprint(?) or maybe not even at all as Apple sticks to its guns and only launches in countries where they can dictate their vision of the user experience.

      Interesting to think about. Either way I am sure if AT&T had to do it all over again they would have done things differently (generally to the users expense I would guess)

      • Hamranhansenhansen

        The current AT&T did not exist when the iPhone was introduced. The original contract was with Cingular. They were trying to build a nationwide competitor to Verizon, and needed a flagship phone. AT&T won the lottery with iPhone. There's no way they would do it differently. Without iPhone, they would be just another T-Mobile.

      • asymco

        Launching outside the US would not have been a failure to launch but it would not have allowed Apple to leverage its retail chain.

    • mortjac

      "Apple's major innovation is cutting the carrier out of the upgrade process". Good point.

    • davel

      Not just cutting the carrier out of the upgrade process, but creating the end user relationship with the phone vendor and not the carrier.

      The phone functionality ( applications ) rests with Apple, media access rests with Apple, the OS fix is with Apple. Apple turned the relationship where the customer sees the carrier as the nuisance rather than the provider of a service.

  • LOL. I can remember screaming at Samsung to provide me with a WM upgrade from 5.0 to 6.0 a few years back. For some reason it was restricted to a couple of countries, which made no sense to me. Thank goodness for XDA developer forum.

    Upgrades are a part of a software platform though and they are part of forming the value proposition to the consumer ( Although only a handful of consumers will be actually demanding the updates, it will be a source of competitive disadvantage if manufacturers/OS suppliers can't get their act together. Even the non techie user will at some point wonder why his phone does not do that cool new thing his friends phone can do (for instance be used for payments etc etc etc), and will start influencing their choice of device when they upgrade the hardware.

    But it is difficult, even for more controlled environments like Windows Mobile, and even harder for open like Android. As an app developer, your headaches will continue (

    • Frank

      Android isn't open (and never has been, at least not in the truest sense of the word).

      The only truly "open" mobile OS is MeeGo (formerly Maemo) which is why community based OS updates are being provided OTA for that system, independent of, but sanctioned by, the hardware manufacturer. I doubt you'll ever see that level of support for any other mobile OS. Ever.

      • Sander van der Wal

        iOS. Not OTA, but who cares?

      • Rj.

        Over The Air

  • Hamranhansenhansen

    The thing with a good software update program like Apple has is you forget the phone was ever missing the newer features. I can't remember when I couldn't listen to background audio or couldn't Copy/Paste and so on. This is huge for user loyalty. They keep moving the whole platform ahead, and you want to continue to be a part of that. And when you are coming to the end of the life of a particular hardware device, it is not like you're getting rid of it, it's like you just need a new case for your Stradivarius. You don't think of throwing away the Stradivarius and starting fresh with nothing, you just go out and get the new case and off you go for another couple of years.

    • r00tabega

      Keeping the value of old devices has a multi-pronged beneficial effects for the product producer and customers:

      1) Customers value the new product higher, as quite a few are smart and factor in resale value with a new purchase (some even yearly upgrade and sell last years model for a decent price).
      2) Hand-me-downs are good marketing if they function well
      3) You can keep high margins on new product, allowing price-sensitive customers access to your ecosystem by purchasing last year's model.

      Apple does this with all their product lines, and it seems to reap great customer loyalty and profits.

      The ability to jailbreak and even unlock iPhones also increases their resale (and post contract) value.

    • mortjac

      You're right about the Stradivarius, but I had to admit I will considering a white Stradivarious, now that I know iPhone 5 will be late…

    • davel

      Yes. Apple's ability to add functionality to older gen's of hardware is compelling. They show you all the neat things you can do with this years model while telling you all the neat things you can do with last years model.

      They do this while giving you a relatively seemless way to upgrade. They do this of course so they don't balkanize their platform too much which makes it easier for them, but the consumer also benefits.

  • Walt French

    “It’s that the upgrade is not universally beneficial to the value chain.”

    I'd say this is the problem right in one short sentence. The value chain isn't.

    A better OS version (a happier customer) evokes a big “meh” from the carrier instead of the notion that every customer is getting more for his money at essentially a trivial cost to the carrier. The OEM, whose reputation for state-of-the-art products is theoretically on the line, instead sees it as a diversion of engineers' and testers' time from what could be devoted to future products, or worse, a cut in next quarter's sales.

    Only the OS author persists in this quaint idea of “value” for the consumer, yet it bears most of the direct engineering, testing and distribution costs while generating exactly zero additional licenses — in terms of short-term forgone sales, perhaps a negative number — in seeking to build its reputation with its customers (who, thanks to the “value” chain where consumers buy from the carriers who buy from the OEMs, might actually be several steps removed from being their customers.

    Back in the days when desktop software was a (fairly) large up-front purchase followed by free updates and paid, but discounted upgrades, I wondered why a firm wouldn't sell the software including a license to all new versions developed in the next, say, 24 months. You would never have to worry about buying v2.6 when 3.0 was likely out in 90 days. That's implicitly the deal that iOS customers have, and that Microsoft could explicitly bundle with their phones, totally eliminating the middlemen's ability to disrupt customers and a PR plus. I suppose individual Android OEMs could try the same, which would provide a competitive edge until everybody was forced to do it; right now Android seems the most WYSIWYG OS.

    • davel

      I wonder if this is the real competitive advantage of integrated systems vs. the modular ones. Each participant in the chain has their own motivations. The hardware maker wants to sell more hardware, the software guy wants to sell more software or get licenses to upgrade. The carrier wants to sell new product and keep the customers it has while milking more money out of them.

      For all the players other than the software guy the upgrade is a cost with no direct benefit. Apple has the benefit of constantly enhancing the experience of the platform while reminding the customer of the enhanced benefit they gain.

      IBM became IBM because of exactly this point. They were the first computer manufacturer to provide a new model computer that ran code from the old machine. This was a huge benefit and they took over the market.

      Apple's attention to detail and understanding of its value wins again. Google did not understand this even while Mr. Schmidt sat on Apple's board. Perhaps they did not give it enough weight or perhaps they are right. They are now putting conditions of their customers after they have achieved marketshare. I am interested in how that plays out.

  • Android is the 'disposable razors' of mobile OSes.

    • mortjac

      You are quite brutal here. Disposable razors cut me, makes me red and hurts a lot. Is it so bad?

      • Lol. I'm not trying to be brutal. I'm just trying to find a good analogy. If you look at the market for razors, disposables take the majority of the market. It's the mindset that I was trying to capture. Maybe I should have said the 'paper plates'?

      • CndnRschr

        But who wants to stick to the same crappy disposable razor (or paper plate) for two years? That's the problem here. I want to eat off china and use an optimal razor device if I have to stick with it for so long. One good aspect of the analogy is that razor companies rip us off as much as the telcos. Do I really need 5 blades… (think 5" screens)? Actually, no, thanks Gillette.

      • David

        But they make you "quite smooth."

      • Frank

        "Smooth" as in Samsung Tablet sales? 🙂

  • mortjac

    In a comment to this post I just wrote:
    "I’m so surprised that Google and Microsoft have such problems with getting their app an mobile business in order. Android payment system is still reported as weak, and Microsoft doesn’t yet have in app purchase. And we are hearing a lot of problems with the distribution of new versions, both for Android and WP7.
    The why on earth are so seemingly simple things so difficult to implement?"

    Thanks to Horace for throwing some light on the issue. But enlight me more please! I still somewhat in the shadows. I wish you could elaborate more around "upgrade failure is a symptom of a deeper dysfunction". This sentence is essential in your post I think.

    • newtonrj

      It means that no one really owns the O/S. Manufacturer, carrier, customer are all seated passengers on an OS aircraft heading somewhere glorious. Problem is that there are no pilots up front. -RJ

      • vinner57

        Ha! I like that analogy.

  • Andrew

    I have owned a total of 4 mobile phones over the last 15 years, living in Japan. The first was a Nokia, and I had to choose the phone based on the mobile number that was hard-wired to the phone. That is still my mobile number today. That was a blue Nokia – I could have chosen a beige Nokia with a different number. This was of course on the NTT DoCoMo service. I could connect to the internet with iMode over the tiny text-based monochrome screen, but this wasn't really a web browsing device.

    After that, I got a Fujitsu phone with a touch screen and a stylus. This had cameras that could supposedly do video calling but I never got it to work. Then NTT DoCoMo introduced a version of Motorola's RAZR for Japan. It was compact and could do video calling to my wife's RAZR but we never got it to work to any other phone.

    Finally we got rid of the Moto RAZR's and got iPhones. Before then, both the Moto RAZR's and the Fujitsu phone were apparently based on Symbian. I soon gave up on the older one, but spent some time trying to get the Moto RAZR to download Symbian software, a feature which NTT DoCoMo advertised but I eventually discovered that localising the device rendered Japanese phones incompatible with any of the downloadable Symbian software.

    NTT DoCoMo also touted the Moto RAZR's music playing capabilities. To load music on the phone you had to use iTunes to convert mp3 files into a different format, then copy them to the removable SIM card. I got the files onto my phone but could never get them to play.

    Anyway, my point is that for the past ten years I have bought mobile phones believing the sellers' promises that they will run software that I could download and install. Before the iPhone, all of those promises were lies.

    • Symbian on a Moto RAZR? I think you're mistaken there.

    • r00tabega

      I will have to say that, in my experience, Palm did right by the users on their Treo models. I remember be amazed to find that in 2005, I could run Bookworm on my Treo 600 and it worked great. Also, the Treo died a couple of times, and VZ would just ship me a new one and a box to send back my old one… the new one was up in 30m after re-imaging my profile.

      I was much happier when I upgraded to my iPhone that Apple copied most of Palm's good practices (ie, image based backup combined with software that worked and a functional address book). About the only thing I missed from my Treo was that the Treo could "localize" itself and prepend area codes based on configuration (i.e., your number starts with 310 but work in 408? you dial 7 digits, not 10).

  • To get the CSSU updates you have to enable the Maemo extras-testing repository. Only geeks do this. Once it's stable it's then moved to the stable Maemo extras repository which should happen far less frequently and shouldn't have phone bricking potential.

  • Frank

    It's a fair point. The frequency of the CSSU updates is largely due to the state Nokia left Maemo5 in, and also a reflection of the effort now being put in by the community addressing bugs and implementing missing functionality, without breaking compatibility with stock applications (which should reduce the need for developers to continually re-test with each release). The updates will slow down as the low hanging fruit is cleared up, and as formerly closed apps (Camera, Media Player, RSS Reader etc.) become fully open-source and move on in leaps and bounds from their more limited predecessors.

    The point is, however, that this is even possible. There is no other platform that permits OS-level updates to be provided by anyone else but the manufacturer. You get what they give you, not what you and other users would like to see fixed.

  • mike

    But then branding matters.

    If users just see their phone as a "Droid" (Verizon) or "Android" phone, then Motorola benefits little from increased "stickyness" – the consumer is more than likely to buy a Samsung or HTC or LG or Sony or ZTE or Acer or whatever next time. Ditto for any of those other companies.

    So ultimately Motorola is just investing time, resources, and money to improve a brand that belongs to Verizon or Google. Why would they do that? I think that is Horace's point.