The offshoring model may still have tremendous scope but there are some serious concerns worth keeping an eye on.

Sridhar Mitta, Managing Director and CTO, e4e India

I want to look at the Indian IT industry in a much broader context than just software programming. So while it started with wage arbitrage and process improvement, there are many other things happening today and many more things possible tomorrow.

Even in a narrow definition of IT, which means programming and product development, the offshorable component of the US market alone is some $ 300 billion.

India is doing around $20 billion. So what we have done is little compared with what can be done -- reason enough to believe that the industry will grow.

It is possible that existing IT companies may saturate, but the market is far from it. Wipro [Get Quote] or TCS [Get Quote], touching $2 billion in revenues, have their size as a strength, but may become victims of their past success if they don't identify new niches and new business models.

In the next five years, just like Wipro and Infosys [Get Quote] did 15 years ago, a new set of small companies will come up bringing with them new business models and new ways of doing things.

They will expand the market way beyond what the existing companies are doing.

Today, we are in a peculiar situation where just the top five companies are bringing in a significantly large portion of India's IT revenues.

We see the business through their eyes rather than the potential in the market and our ability to meet it.

But the new players will not stop at wage arbitrage or process improvement, but may even start with applying technology and business models.

The old way was about cost reduction, but the new way is to create new markets. I see brand new companies, young start-ups, new business models, new technologies all coming up and a market that is broad based compared with what we see as IT now.

In next few years, anything that is of economic value and amenable to conversion into digits is open to offshoring. Don't look at just what the large companies are doing today to determine where the industry is.

Let's look at the economy. Developing countries, too, need products; you can't simply make products for the US in India and sell them.

This is what C K Prahlad has been taking about: the needs of the Bottom of the Pyramid. Many things aren't luxuries any more but a necessity that makes for real business opportunity.

The key constraint is to offer the products at substantially lower costs. For that you have to apply technology. If you do it, the market is big and not just India and China.

Some four billion out of the 6 billion world population need these products. You can't sit in the US and make these products for the Indian or Chinese market. It has to be done here. The top 20 per cent companies need to change. Whether they will or not, I don't know.

The larger firms are into general-purpose software services but that is okay. The smaller ones, the remaining 80 per cent, will have to specialise. This has started happ-ening and they are getting into tie-ups, merging or getting acquired.

Further, can several small players get together and offer customers something more than the sum -- like what Taiwan did with the PC.

No single company did everything when Taiwan became the global leader in the production of PCs. It was through the cooperation of many companies; something similar can be done with software or IT.

One good example is the India Semiconductor Association, where a group of firms, large and small, have decided to "do something together which they find difficult doing individually".

There will be other collaborations, auto components and IT, healthcare and IT, to get to new markets. Indian companies are capable of the IBM-Mayo Clinic type collaboration, too, although they aren't doing it yet.

Two years ago, I would have said, we too need to make products. Now, products delivered as a service is becoming important.

Take Salesforce.com, a small start-up that delivers as a service, for $ 30 a month, many things that were so far only available as expensive license package.

But this needs a data centre, delivery and support and crucially, a model with high service content. This is precisely what the Indian way of doing business in IT is amenable to, unlike the American product or licence model.

The implication is huge. There is no reason why salesforce.com couldn't have started in India, sitting here and delivering support in the US.

Such things are going to happen. Wipro or Infosys already have what it takes to do something like this.

Will they deliver? Others will for sure.

Shankar Sharma, Director and Chief Global Strategist, First Global

The Indian IT services companies have a "unique" business model...but not in the manner you think. (These companies are not technology companies, by the way. India, strangely enough, has not even a single technology company in the listed space, despite producing millions of engineers).

The Indian IT services model is a high fixed-cost, highly capital-intensive model.

Put simply, for every dollar of new revenues, the sector has to add new bodies and new infrastructure. No new bodies and capex, no new revenues.

Hence, the business enjoys virtually no operating leverage, that is, a change in revenues does not lead to a more than equal change in earnings.

This is where the "uniqueness" of the business model lies.

Typically, capital-intensive, high fixed-cost industries enjoy tremendous operating leverage. For instance, cement, steel and automobiles.

Once breakeven revenues and profits are achieved, every incremental dollar of revenues is virtually pure profit. Hence, you witness tremendous margin expansion (that is, high operating leverage) in such sectors in boom times.

The Indian IT services sector, on the other hand, enjoys all the disadvantages of a high fixed-cost business, but enjoys none of the upsides of the same.

Costs remain fixed on the way down, but are variable on the way up. Hence, if you were to take a purist's view of their model, you can only conclude that the model is intrinsically flawed.

Much like Sehwag's technique. So what makes the sector tick, just like what makes Sehwag click?

Very simple: youth and growth. The Indian IT services model is built only for a single business scenario: growth. Change the scenario to moderate or negative growth, and the warts will begin to show.

The huge, virtually inflexible fixed-cost structure will begin to bite. Just like in Sehwag's case, once youth wears off, and hand-eye coordination worsens, the lack of a sound technique will begin to hurt.

Nothing at the moment suggests that growth for IT services companies is about to fall off a cliff.

But that's the whole point. Smart management requires that one starts modelling for this very scenario, and make subtle alterations to strategy.

A good case in point is Taiwan Semiconductor. The Rolls-Royce of semiconductor contract manufacturing, it enjoyed go-go growth through the 1990s.

Basically, it did the dirty fab part of US semiconductor companies, who turned into "fabless" design companies, outsourcing the fab to Taiwan Semi.

It enjoyed margins of 45-50 per cent then. Now that's down to the teens and low-20s. It could have re-engineered itself as an actual high-end chip design company somewhere along the way.

But it didn't. Once you get stuck in the "outsourcee" mind-set it is very hard to change tack, and upgrade skill-sets.

That's because outsourced revenues are a lot more painless, on the face of it, than innovation-led revenues. But none of Dell's sub-contractors enjoy the return on equity that Dell does.

Margins for Indian IT companies have shrunk in the past four to five years. At some point, the Taiwan Semi syndrome could be witnessed by our IT companies as well.

And that's the scenario that should be keeping us awake at nights.

Infosys employs more employees than Tisco [Get Quote]. Will it have the courage to lay off if and when growth slackens or turns negative, without abandoning its "industry statesman" status? The truth is: it will be difficult.

So the only solution is to chase growth. Organic growth has a shelf life.

Inorganic growth doesn't. Hence, some bold thinking will be required. How to improve the low value-addition per employee? Moving up the chain is easier said than done.

For one, the tag of an off-shorer is hard to shrug off. Perhaps, our companies would have been better off being American companies with an India-leg than the other way around.

Cognizant's growth and valuations have been far better than that of Indian IT companies.

A key factor has been it being perceived as an American firm. Second will be to build scale. To our mind, a very logical way is to merge among themselves.

A Wipro-Infosys combined company will be a giant, with a market cap of $ 25-30 billion. Use this heft to buy out the low-price/sales companies like Cap Gemini, EDS, even Accenture. "Winfosys" will have huge bargaining power with customers, as well.

Out-of-the-box thinking, American-style is what our companies need. The softly-softly approach has run its course.

As told to Harichandan A Arakali