Friday, April 29, 2011

TMI: The government needs to exit the software ecosystem — by Ditesh Gathani

The government needs to exit the software ecosystem — Ditesh Gathani

TMI: April 28, 2011

APRIL 28 — The government of Malaysia has had a chilling effect on the growth of the Malaysian software ecosystem, and needs to make an exit in order for the sector to blossom.

It’s hard to know where one should start commenting on the latest fiasco around the proposed 1 Malaysia email system. Perhaps more than any other single software project in recent times, the 1 Malaysia email proposal has highlighted the deep rot within the software industry in Malaysia. The public outcry has been loud and consistent. The subject even comes up during office meetings, as a form of collective outrage against yet another silliness we’ve been asked to endure.

For those unaware of the proposal, the gist of it is that Tricubes, a distressed public listed firm, won a bid to provide Malaysians with an email service. That’s right — to provide Malaysians with an email service. Oh, the best part of this proposal is that Tricubes would be charging the government RM0.50 for every email sent by the government to Malaysians.

If it sounds incredible in any zany sort of way, you must be new to this country and I must warmly welcome you to Malaysia, fondly known among its residents as Bolehland. Magical, amazing and outlandish things happen in this country which are entertaining in a tragic sort of way. So tragic that it’s worth creating a newspeak word to properly describe the tragedy.
tragedic. adjective. Recursive definition to describe tragedies so outlandishly tragic, that its tragedic

As tragedic proposals go, all things considered, this one had the distinction of being pretty dumb. Free email systems have been around for a while and the need for yet another email system is rather doubtful.

This particular proposal was, in fact, not even Tricubes’ little invention to make money off public coffers. Neither was it a deep nefarious plan by the government to spy on and control all Internet activity as the popular blogger Sakmongkol has alluded to. The proposal really came from Pemandu’s entry point projects, and much like all world domination plans, it had a simple beginning.

You see, the idea was mooted in the NKEA labs. It was discussed, tabled, approved and signed off. As these things go in the government, due process was followed and the idea got escalated up the food chain, punted over to Mampu to execute and subsequently bunted off to TriCubes to implement. Nothing extraordinary, just standard big organization bureaucratic SOP’s at work.

For the sake of fairness, it must be noted that some chaps politely attempted to call out the silliness of the idea whilst it was being deliberated in the various meetings. But they were far too polite to strongly drive the point across that it was a really dumb idea.

You see, there is one golden rule about big organisations: You simply cannot call out ideas as being dumb, not publicly at least. Nobody is quite sure why you can’t do so but rumour has it that doing so would lead to an unpleasant and uncollegial atmosphere, and as can be expected in such situations, a complete breakdown of law and order in society would be the result.

Oh, you would also be branded as being dickish for calling out dumb ideas. Such social conventions have ensured that politeness has almost always triumphed over intelligence in big organisations, and incompetence has continued to comfortably rule the roost.

Anyway, I digress. Let us get back to the topic at hand.

Who has suffered as a result? Tricubes’ reputation has certainly taken a beating but all things considered, they were already in the process of getting delisted anyway and as such, not exactly flying high. The news of them winning the RM50million 1 Malaysia email bid jacked up the firm’s stock price and the company had an increased market cap of RM40million at the very least. Not too shabby, really, for a dumb idea.

The Malaysian government too has come out well by cleverly distancing itself from the entire affair, with our prime minister punting the buck to the private sector (read: Tricubes, the not-so-failing public listed firm).

Mampu, on the other hand, has been fiendishly quiet so far, pursuing the well-known strategy of burying their head in the sand and hoping for the best. The strategy seems to have paid off in that they have been spared much of the public wrath. So, that really just leaves only Pemandu out in the cold.
Incidents such as this one don’t help Pemandu’s reputation. They have been tasked with re-engineering the national economy and the best we’ve seen for the software ecosystem industry is free web email? Oh, come on guys. Surely you can do better. Borrowing the slogan from Obama’s campaign “We can do better” may be a suitable new direction for Pemandu.

The government, even by generous accounts, has had a marvellously Homeristic tragedic track record in re-engineering Malaysia’s software ecosystem, and the 1 Malaysia email fiasco is the latest in a long line of failed government software ambitions.

MSC Malaysia, the key strategy vehicle for the government’s software ambitions, was set up in 1996 with grand plans of replicating the success of Silicon Valley in South East Asia. In this period of time, we have seen two Internet booms and countless wildly successful Silicon Valley startups. Google, YouTube, GroupOn, Twitter, Facebook, Blogger, PayPal, so on, so forth.

Silicon Valley has not been merely successful in the period of time but successful beyond comprehension. Countless millionaires were certainly created, but a large number of billionaires as well with fortunes larger than the national GDP of a number of countries.

In this interweaving period, MSC Malaysia has successfully missed out on riding every single technology wave. That includes online video, search technology, social, payment systems, biotechnology, group buying … and, let’s not forget, web-based email. The consistency in failing to ride the wave is, in itself, shockingly impressive.

For purposes of comparison, Hotmail was a free web-based email system, launched in the same year as MSC Malaysia, and the first real big Internet phenomenon responsible for kicking off mass market usage of the World Wide Web. In this light, our technology ambitions not only as old as Hotmail but, in fact, identical to Hotmail because after 15 orbital years, the best technology project our economic-reengineering-overlords have come up with is free web-based email service… powered by Hotmail technology. The irony would have been entertaining if it wasn’t so painfully tragedic.

Conservative estimates of technology advances put it at a non-linear doubling rate every 18 months. In layman’s terms, that simply means that for every two years of innovation that happens outside the technology industry, the technology industry manages to produce three years worth of innovation. This is a non-trivial rate of growth that has held true throughout the history of Silicon Valley and has been the primary factor responsible for the tremendous innovation in the industry.

Successful software technology firms go through an entire full growth cycle from startup-level-maturity to world-domination-maturity within 10 orbital-year spans. The classic textbook example which illustrates this observation is, of course, Google.

It was launched in 1998 (two years after MSC Malaysia) and within the span of a decade, it completely dominated the tech industry, leaving its competitors dazed and incoherent. If the 1990′s belonged to Microsoft, Google certainly left their mark on the first decade of the 21st century. It is still not clear if Microsoft will survive Google’s onslaught — most are betting on an Apple and/or Facebook dominated decade.

But I digress.

The important thing to take away from this is that MSC Malaysia was set up, by conservative estimates, 23 technology years ago. The government, through Mosti/Mimos (generally) and the MSC Malaysia vehicle (specifically), have taken on themselves to engineer a software ecosystem in Malaysia.

Not just a normal software ecosystem but a world-class software ecosystem — one that is revenue generating, profit-making, being a beacon of progress and responsible for catapulting Malaysia into the 21st century as well as showcasing possibilities that can be achieved in a non-Western environment. As far as masterplans go, the MSC Malaysia masterplan was pretty good — timing and strategy wise.
However, the devil, as they say, is in the details. The government ministries and agencies responsible for achieving the MSC Malaysia plan implemented the following recipe for success:

•    Identify technology trend.
•    Write up blueprint and masterplan to ride and monetize technology trend.
•    Determine key projects, KPI’s and budget.
•    Outsource key projects to private sector.
•    Provide grants to private sector for R&D, explore commercialization opportunities.
•    Review achievements, congratulate self, publish press release.
•    Lather, rinse, repeat with new technology trends.

This recipe was a refinement of an earlier recipe, used in the era before the web became big:

•    Identify R&D trend.
•    Determine key deliverables from grants.Provide grants to private sector for R&D.
•    Review achievements, congratulate self, publish press release.
•    Lather, rinse, repeat with new R&D trends.

Remember the DAGS grant? The MGS grant? The TechnoFund? The Science Fund? SMIDEC grants? The plethora of grants and grant money available for the cunning entrepreneur propped up the bottoms lines of a number of companies.

The quality of the grant proposal didn’t quite matter because there was a lack of depth in regulatory oversight. There was nobody to call out bad proposals, nobody to identify hackety software implementations and nobody to push for quality over quantity. It was a numbers game, pure and simple.

The general thinking was that with millions of ringgit of grant money handed out, surely someone and somewhere would be enticed to build great software. Nobody sought to question this line of thinking and as a result, the system was quietly gamed over the years.

Allegations of corruption made its way into coffee shop talk and became common lore in the tech industry within a few years. In fact, there were companies specializing in corrupting the disbursement process.

Knowing the right people became more important than having the right idea to execute. Companies were setup to game the system and multiple levels of outsourcing were established. Business plans were flimsy, as one would expect them to be, because the considerable effort was being put into building up connections instead of building up companies.

This was a game played equally well by all involved parties. Companies would manoeuvre to get the grant money and/or software projects, outsource the actual development to a third-party company, show progress by delivering half-baked products and upon completion, repeat the process again with another grant or software project.

Monitoring and approving officials, on the other hand, would simply check off items on their list and be able to meet their KPI’s in delivering the grant money to intended recipients and/or completing the projects. Everybody wins, right? Not quite, because no good technology comes out of playing this game.

In recent years, this system was replaced by a refined system of handouts in the form of smaller sized grants and larger sized software projects. The government’s approach in spearheading technology direction was achieved by identifying key growth areas and funding the implementation of software projects by the private sector in these areas.

The private sector was also asked to demonstrate the viability of the key focus areas through grants. Again, the idea was that this would stimulate the private sector by having them focus and invest in the key areas. That too did not quite materialize as everybody simply focused on gaming the system by investing minimal effort in actual technology development and technology management.

Does anybody remember the four flagship applications launched under the MSC Malaysia umbrella? What happened to the Telehealth initiative? Are Malaysians enjoying improved standards of healthcare? What about the MyKad project? Has it taken off in the private sector as a secure ID platform as originally envisioned? RM7 billion and more was invested in the smart school projects.
Where are our smart students?

The rot goes deep in all directions. I remember MDeC pushing the private sector to adopt CMMI methodologies to reach world-class software development standards. Whilst this initiative was being ballyhooed in the local media as the new new fix for all that plagues software development initiatives in Malaysia, elsewhere in the world eXtreme Programming (XP) and Agile Development methodologies took hold. Needless to say, the CMMI initiative has not succeeded in its attempt to build world-class software.

Remember the pre-seed grants? Where are our world-class start-ups? Remember the eContent grants? Where is our digital content? Remember the initiative to retrain graduates? Where are our world-class graduates? Remember the numerous training programmes aimed at providing skilled labour workforce? Where is the skilled labour workforce?

It’s not just a question of cluelessness on the part of bureaucrats. It’s also about politics and bad decisions getting the way of a progressive technology culture. The previous Mosti minister back-pedalled on the use of open source technology just about when open source technology was being used by every single major software powerhouse in the world.

Politics and unbelievable pressure from certain quarters was responsible for the minister’s decision, and the resulting stultifying impact on the industry was powerful. The one opportunity we had for a progressive technology culture went out of the window in that instant.

The very same minister took liberties in overriding decisions of Malaysian standards committees in an ISO standard deliberation by forcing the Malaysian committee tasked with reviewing international standards to submit a “No comment” instead of their unanimous (unfavourable) technical findings. Again, politics and pressure overrode good sensible decisions made by those attempting to do their jobs.

There are many more stories of government missing the boat on driving the software sector in the country. Think about it — 23 technology years. That’s a really, really long time to fail to produce a single world class company. In the parlance used by computer scientists worldwide, the intersection between the set of their masterplans and the set of resulting world class companies is the null set.

Failed technology management. Failed technology development
One follows the other and that’s the key insight here. Compared to the private sector, governments (and any big organizations) are horrible at technology management, and as a corollary, at technology development. They should be the last people tasked with building the software sector, because as history has shown, they will flub it at every possible opportunity.

Intuitively, the reason for this is obvious. Technology management and development is a difficult science. It requires a rare combination of deep technological knowhow, product management and an oracle-like ability to marry market analysis with technology trend spotting. Even the most prescient investors and technology pundits in San Francisco often get aspects of technology management wrong — what more Malaysian bureaucrats on a government salary.

Let’s take Facebook’s News Feed feature to illustrate the difficult in building and managing technology. The success of Facebook can be traced almost entirely to the Facebook News Feed feature. Remove the News Feed and all you’d have on your hands are photo, video, blog components with the ability to make friends. In other words, you’d have MySpace or Friendster instead of Facebook.

The quintessential aspect of Facebook is the News Feed. Facebook’s genius was in pulling your friends activities (photos, videos, blogs etc) and pushing them to a continuously updated personalized News Feed. The value of a News Feed seems obvious today, but before Facebook made it ubiquitous, nobody understood the value of such a feature. The very foundation of Facebook’s success was in the invention of the News Feed — it was sufficiently advanced technology, one that allowed Facebook to make MySpace and Friendster irrelevant.

The million dollar question is simply this: Would have any Malaysian technology bureaucrat realized that the future was social and invested accordingly in it? Would they have written up masterplans on going social and funding “social entrepreneurs”?

Would they have understand the impact that social would have on the world? Would they have understood the value of the News Feed feature, much less the incredible engineering effort it takes to build such a feature? Would they have had mentorship programmes in place to train technology entrepreneurs going social? Would these programmes have been successful?

If you didn’t think they would have bet on social, you’d be probably be right. The thought that government bureaucrats would be so technologically prescient strikes us as being ridiculous because intuitively we understand that building technology roadmaps and funding technology requires understanding the said technology. And understanding technology requires the ability to hack on technology, which in turn, requires an engineering mindset, not a policy or political or bureaucratic mindset.

All of the wildly successful technology companies were driven by entrepreneurs with deep understanding of technology. Not a single successful technology company was driven by direct government technology policies or government roadmaps. Not a single one.

Imagining that governments can drive the software sector is simply a proposition that does not compute. The best the government can do is to stay out of the software industry and let the market thrive on its own merits.

There is a simple way to visualize the benefits of no government involvement. Instead of one entity, the Malaysian government, making big bets on the next big thing, we could have had a hundred thousand technology entrepreneurs furiously attempting to make money from the varying shades of one hundred thousand different ideas.

When thinking of it from this perspective, it is little wonder the Malaysian government has failed in building Malaysia’s software ecosystem. We kept pooling our eggs in one basket and every single time we flubbed it.

For every wildly successful Google, Twitter or Facebook, there would have been 99, 997 failed entrepreneurs. Which is fine because the three that succeed, succeed wildly beyond belief and end up raising the economic tides for the failed entrepreneurs.

Because we’ve never succeeded in this regard, we’ve had to suffer economic woes and our poor showing on the world economic competitiveness index is reflective of this.

The great government experiment has failed as it has in every single instance around the world where big governments have been in play. Capitalism simply does not favour big governments. The best the government could do to boost the software economy is to get out of it.

Observations and analysis of Silicon Valley’s success indicate that the free market works pretty damn well. We would never have seen the likes of 1 Malaysia email in a free market environment. Capitalism rewards ingenuity and hard work, and punishes the lazy and the underperforming.

To help us move forward and build a working software ecosystem, the government needs to withdraw its grants and big projects. No more key focus areas and handouts. No more handouts and training exercises. No more junkets and technology conferences.

Just let the ecosystem be. Let them compete on their own merits, as they may be, and find their way. Let them make their mistakes and learn from it. Trust me, with the pressure of having to succeed, they will learn fast.

And who knows, they may even be wildly successful beyond anybody’s expectations. — gathani.org

* Ditesh Gathani is an open source expert.

No comments: