the problem of software development in Japan

The NY Times has an interesting article up on Why Japan’s Cellphones Haven’t Gone Global.
Techmeme and Twitter shows a decent amount of interest in the article, which is news for most people outside of Japan.  It’s all the more interesting because NTT DoCoMo have just launched the first Android phone in Japan, and the iPhone 3GS has been somewhat successful for SoftBank.  Probably not as successful as SoftBank would have wanted, but with NTT tied up with Google, SoftBank was smart to tie up with Apple, leaving AU looking disheveled and weak with no significant smartphone in it’s lineup.
So the key question is what should Japan Mobile Inc. (i.e. the companies that make Japanese cellphones) do?  Sadly for those Japanese businesses, it’s too late to develop a new mobile OS and to try to compete on that because none of the Japanese companies are strong enough to develop a compelling mobile platform on their own (or they would have done so already).  Tie-ups such as Sony-Ericsson have been less than compelling.  So if there’s no reasonable way to compete on the OS level (try to convince me I’m wrong), and competing on hardware manufacturing is a losing game due to cheaper production in Taiwan or China, then there’s only competing at the application level.  Sad reality.
Time and time again, Japan’s lack of prowess in software development (with the lone exception of console gaming platforms and their games) creates a barrier for Japanese businesses to be successful beyond Japan. Takeshi Natsuno’s solution for the Japanese mobile industry is to:

“focus more on software and must be more aggressive in hiring foreign talent, and the country’s cellphone carriers must also set their sights overseas.”

This is not a reasonable solution. Let’s pick it apart.
How can Japanese mobile hardware manufacturers focus more on software when they are tied to a schedule where they are releasing new models many times a year? The current staff can’t do anything more than what they do now, which is add a feature, add a new piece of hardware, and then recycle everything else. How can the existing developers, who do not make compelling software today, suddenly make compelling software tomorrow?  This is not a solution.
Regarding “hiring foreign talent”, until Japanese phone manufacturers can pay globally competitive rates and can draw developers away from places like Apple, or Google, or Nokia, or Samsung, they can’t do what Natsuno suggests. I submit that this is a hollow solution- toothless.
For those who are in Japan, and know how domestic (i.e. not international) the 3 main Japanese carriers are, the thought of them effectively competing in the consumer space outside of Japan (especially KDDI) is ludicrous. Japanese carriers are competitive globally in a few areas, KDDI in submarine cable laying via their KOKUSAI CABLE SHIP CO.,LTD. (KCS) subsidiary, and NTT via Verio in the US and other telecommunications subsidiaries around the world, but these are not consumer-facing businesses. These are B2B.
So Natsuno’s “solution” isn’t really a solution in my opinion.
So what is a solution?
I don’t have a quick answer, because it is clearly not a simple problem.  But I do believe that whatever future path for Japan’s software development, be it for mobile phones, or other devices, investing more in open source software development is a critical path towards a) getting Japanese developers to a global level of skill, and b) gaining prowess in software projects that will underlie the key devices of the future. 
That Japanese developers already contribute around 15% of the Linux kernel code (self-link) is a great step, and Japanese businesses and universities should be encouraging more of their developers to join key OSS projects. 
I welcome your constructive criticism and ideas on how Japanese companies can be more competitive.  This is not an issue only in mobile software development- this is an issue across all of software development in Japan.  Desktop software and Internet services all face the same problems.

5 comments on “the problem of software development in Japan
  1. Well, as I see it, Gen, it’s not about developers. I have no doubt that Japanese developers punch their weight: I’ve seen some truly impressive things developed locally, and I bet you have to.
    No, as ever, it’s about a holistic understanding of user experience, which was the selfsame thing that always seemed to be missing from my adventures in Tokyo Web development circa 2001-2003. With a few notable exceptions, everything I saw in those years always seemed to be less than the sum of its parts, because development shops (and especially nervous client-relations type anxious to keep that mythical “foot in the door”) would let any and every kind of prerogative come before user-driven measures: arbitrary release schedules, flavor-of-the-week functionality, shacho’s ego, “plus alpha,” you name it.
    And this, in turn, gives the lie to the notion of the “more advanced” and “more demanding” Japanese market. Whether it’s Web sites or, more to the point, handsets, time and again I’d see development organizations let any old crap hit the market, because they knew people would buy it. Up-spec compared to the North American or Western European markets? Sure, of course. But easier or more pleasant to use in any meaningful sense? Please.
    So what’s under this all? It’s thick and knotty and entirely overdetermined, of course, and I’m only one of the blind men patting the side of the elephant. But I’d wager that one of the causes is that Japanese handset producers have allowed their market to expect and demand a release cycle that resembles fashion more closely than anything you see elsewhere in technology. (There may not be anything that can be done about this at this point, but I don’t think this was what “kaizen” was supposed to mean.)
    It’s also true that a lot of what Natsuno is saying in that article applies outside Japan, not least to my own employers. But that’s a matter for another day – I have other thoughts, too, inevitably, but I don’t want to choke your comments section with a dissertation.

  2. The article focus on the ability of Japan of being successful on overseas market for cellphones handsets. (from a Japanese point of view).
    There are markets which are quite successful if we look at them: cars and cameras for example. Both have little software with a UI for human. They are usually good designed products on the physical form factor.
    Japan may need more interaction designers for software development. The level of computing curriculum in University is not that great.
    Though is it really true, if we look at the Game market where Japan is very active.
    After my 3 years in Japan, the two most important issues on working for an international context were:
    1. Japan is different. (This is deeply into the mind of every Japanese. A bit like there is the world and there is Japan, without accepting the fact that everyone is different and then that so many things are in common. It is a self exclusion.)
    2. Not *engaging* in international discussions. Sometimes not because of the level of the English, but just because there is no notion of arguing a point. (when) You listen, you do things your own way.
    There are competing neighbors: LG (Korea) and HTC (Taiwan) which created Touch phones with overseas market.
    Another point in the article which is not completely mentionned: foreign handsets are not that successful in Japan. I would love to know about the rest of Asia. I have seen a lot of LGs and HTCs in south east asia. But I would love to see hard data.

  3. Phil says:

    I doubt the makers have even figured out they have lost at the software level yet.
    Look at the response to the iPhone, a bunch of hardware form-factor clones, didn’t seem to occur to them the hardware was as much an outlet for the software platform.. (and associated 30% cut of sales)
    As for “hiring talent”… i removed the “foreign”, i don’t think they hire “talent” or value adders foreign or otherwise. They would hire who they can get and assign them according to who is a available, with a few talented/overproductive sempai’s here and there the cajoling and herding the underproductive, undertrained remainder. The ones who could add value and build platforms are weighed down simply making things work. (Don’t confuse that with me saying there is no talent, there is)
    … actually i imagine they largely outsource the software as an afterthought to marginal companies that work as i described above.
    (… not that there is anything wrong with outsourcing software as an afterthought, that is exactly how Motorola developed much of their handset software and look how that worked out for them 🙂
    )
    The talent is there, it’s just underutilized, and gets locked up in sempai/kohai relations ships rather than real teams. That’s one reason pay scales are a lot lower in Japan than the states.

  4. RMilner says:

    It is all about the interface design.
    Japanese companies don’t seem to have realised that modern gadgets all either need to be programmed in some way or connect to a computer to transfer media. Thus, the users interface is as key to the whole user experience as the hardware.
    Japan still designs great hardware but it is badly let down by third rate interface design.
    You only need try using a Japanese phone to see what a rotten interface it is.
    The same thing applies to loads of interface software for connecting devices such as cameras or music players to your PC and managing the media. In every case, the iLife software in Mac OSX is much better. It’s better thought out and much easier to use.
    I don’t have a solution to offer.

  5. shu says:

    btw Gen, you keep saying “except japanese console games”. but most of my game industry friends, as well as the online community, agree that over the last few years japanese game developers have fallen behind their western counterparts. both technologically as well as content-wise. japanese game developers themselves are starting to admit as much, like Hideo Kojima of Metal Gear Solid fame.
    http://www.1up.com/do/newsStory?cId=3170630