Productivity, greed and the long term health of IT

To start with heres a few good links

http://www.iht.com/articles/2005/06/12/bloomberg/sxmuk.php
http://www.thehindubusinessline.com/ew/2006/08/21/stories/2006082100170300.htm

Though its toughie to measur the Indian IT industry's ouput, various sites point to numbers between $23k to $32k p.a. How the 'revenue' generated by the engineering and support arms of MNCs is included is a grey area - if its just the budget allocation for the headcount, that would skew the numbers terribly. The Oracles, IBMs and Microsofts of the world make a lot of money per employee, and that needs factoring in somehow.

In any case, in this backdrop, the auto industry has been seeing massive growth in productivity. A friend quoted a number of 30+ lakhs - thats about $55k per person - I have not been able to substantiate that, but have found papers refer to 40% growth in productivity bet 1993-2003, and some which state a figure of 13Lakhs or so - about $25k. This used to be a very labour intensive industry till recently.

Of course, the big story underlying the data is the wage difference between the two! IT almost exists in a different economy, and has created an economic island around it wherever it exists - driving up cost of real estate, services and everything else. This is turn has created a constant demand for the 20%+ raises year after year, and earlier-unimaginable starter salaries.

Sooner than later, we'll get to a place where we're so out of whack with reality we'll need to consider some questions.

  • How soon must salary growth get in synch with the rest of the economy
  • How do we see a multiplier effect in productivity, so brand India does not seem overpriced
    • tech upgrades
    • The industry grew up in Univs not too long ago - and 35% of the work never gets 'done', and a lot more never sees the market. Its got to start working more like the auto or consumer goods sector and working back from costs and market needs.
    • Build services, organizatiions around standards (and get involved in supporting/promoting them) and longer running "basics" - OSs, protocols will help. Too many, frequent upgrades, a multitude of standards can only slow things down, and create more learning curve issues.
    • Stop focusing on 'time' as a productivity enhancement tool.
    • Better time-management skills and tools.
    • Transparency, collaboration and wikis.
    • Some good knowledge management tools.
    • Industry needs to poke its nose into education wayyyy more - projects, seats of learning, sourceforge.net initiatives, curriculum.
    • Its possible to over-processify
      • Agile methodologies - in spirit, not letter (actually XP etc cannot be 'followed' - goes against their idea)
      • Measure the value of each step, and judge if its become 'bureaucracy' ?
  • Cater to the domestic market - and stop existing in the 'dollar' world. Or, work back from costs and price/develop for what a long term, sustainable market will happily support. Develop pricing models for the domestic market - its a very different one, and a much larger challenge. Once arbitrage opportunities end or migrate away, the base of the pyramid will have to be the domestic market, with highend tech work bringing in the high-margin revenue. A robust local market will also help push branding/products, but pricing vs value wise its a chicken and egg issue that the industry must take the first crack at.
  • Crazy - but for two years the hikes of all 3+ years experienced people get used for a managed upgrade of tech institutions and education.
  • Help smaller tech entrepreneurship for experimenting with region, niche specific products etc. Why do we have to ape the 'large tech product co' model ? India traditionally has been a very fragmented market, with customers keeping costs low and keeping monopolisitc situations from happening inherently (not counting the legislated monopolies).
  • Evolve new work models to use combined experience of individuals. A lot many people have made it financially, and are not willing to give all their time to work, which seems to have become the norm. Some of these would not mind taking either part time lowered effort roles, or even combining forces with someone with complementary skills to fill in a job! HR needs to innovate beyond team building, and management styles even more so.
  • There's been a lot of lip service to this - but team accountability needs to be way more important than individual responsibility (not replace it, but be the decider).
    • As a corollary, the hiring of 'clones' i.e. multiple of the 'stated' ideal person will reduce - cause the best teams always have a mix of complementary skills.
  • Telecommute when/where possible. The number of hours saved alone will improve output if used
Updated:

As one of my friends pointed out over email discussing this post, the whole using the raises to better institutes idea is too far fetched and 'socialist' - I take that back.

And, on the other hand, http://www.rediff.com/money/2006/sep/07jobs.htm.
One crore IT jobs! Wow. I don't know what that number is based on, but its staggering.

No comments: