‘There is a lot of headroom for Indian IT'

So says Pankaj Ghemawat, in a free-wheeling interview to Vivek Kaul and Satish John. Ghemawat is a full-time professor at IESE Business School, Barcelona. He is currently on leave from Harvard Business School, where he taught from 1983 until 2006.

Ghemawat earned his degree in applied mathematics from Harvard College and a Ph D in business economics from Harvard University. He then worked as a consultant at McKinsey & Company in London before joining the Harvard faculty in 1983. In 1991, he was appointed the youngest full professor in the haloed B-school’s history. Excerpts from the interview:

When Thomas Friedman said the world is flat, he probably exaggerated a bit, being a journalist…
Well … You know some journalists exaggerate and some don’t. He certainly got a lot of mileage out of the exaggeration. Remember what Ambrose Pierce said: “People who stand in the middle of the road get run over.” Just like that, I am more middle-of- the-road kind of guy. Let me exaggerate in order to gain attention.

And that’s the sense in which the Friedman book has been helpful because given how widespread this exaggeration is, and given this context even a middle-of-the-road view can be newsworthy.

So when I look at how the Washington Post carried the story - ‘Academic says world isn’t flat,’ it is relatively hard to imagine that headline…

That headline was everywhere.. after it appeared in DNA Money
Yes ... when people ask me how I feel about the Friedman book apart from concern of social consequences which I personally find very handy, because it is a readymade marketing tool.

Without that book, I would have to invent some kind of a strawman. Well I don’t have something up there like a big bull painted with an extreme view. It’s interesting how people seem to oscillate between the extremes, rather than exploring the middle ground.

Which is why I focused on the middle ground. Somebody was telling me that there’s somebody coming out with a book on globalisation, which is on the other extreme: everything must be localised to the individual customer. I guess if we read history, which I do, such swings in the pendulum are inevitable. But I don’t think progress is made by oscillating from one extreme to another.

A lot of Indian companies are going abroad. How does it fit in with the argument you are trying to make?
The point is not about going abroad. Otherwise I won’t write a book on global strategy.. (Laughs).. The point is if you go abroad with the notion that the world overseas is going to be just like the world at home you are in for a really nasty surprise.

Now most business people I know are a little bit too experienced. (Laughs again). A little bit too wordly-wise, as too many fingers burnt over the years at a corporate level to really take literally this notion that differences don’t matter anymore.

But I think for early-stage companies you know this is an automatic thing to do. The thing I always stress is, borders still matter. This is why you need to evaluate things properly rather than just treat globalisation as an act of faith.

I ran a survey for private business online for the Harvard Business Review where they polled readers on such propositions as globalisation as an imperative, rather than something that must be evaluated. 88% of the respondents agreed with the idea of globalisation (as an imperative) and that to me is the height to the managerial irresponsibility.

Indian companies find it more at home acquiring companies in UK. Is it because of a common language, a past colonial relationship?
Yeah it is. Having said that, a couple of caveats: One is the importance of the actors in the industry contingents. Common language doesn’t matter much in cement. It matters a great deal in television programming. For instance, to make stuff useful for business audiences, ideas have to be customised to the industry one is talking about, because reality plays out very very differently in different industries.

Second, there is a limit on the extent to which internal experience can substitute external proximity. A good example of that is Spanish multinationals investing in Latin America. Between 1997 and 2001, about 50% of Spanish foreign direct investment went to regions that barely account for 5% of world GDP.

The Spanish companies were emerging out of a highly isolated era under (Francisco) Franco. They figured this was easy. But if you look at where Spanish FDI is going now —- its heavily tilted on the European Union.

The third point I would make is that this doesn’t mean that language has become irrelevant. If you look at BBDA, which is one of the two largest Spanish banks, its north American strategy is big but that strategy is essentially focusing on the Hispanic speaking market within the US. Or, similarly, Tata Consultancy Services.

TCS now prides itself on no longer being just an English language-capability company but if you look at how the operations are actually organised, its Latin American delivery centres are a mix of Latin America and Spain and Portugal. Budapest is meant to serve the German speaking areas, Morrocco is being looked at as a basis of Franco-phone offshoring.

So it’s not that TCS is a one-language company. But in software services, given how much linguistic interchange matters, that is a sensible way of organising delivery capabilities.

Immigration laws being tightened, such as in the UK, flies in the face of the world being flat…
Exactly. This is not so much in variance with the facts, because very few people bother to actually go look at the facts. But it’s in variance with the data that we are bombarded with everyday.

For example, I spend more of life than I would care to at Heathrow. So do many of the other people who swallowed this gospel of globaloney. Surely, waiting in immigration lines might give you sometime to reflect, on whether it has actually become harder or easier to move across borders in the last ten years.

There is a sense that people don’t tend to bring data to debate. There is a almost a suspension of personal judgement as well based on direct personal experience that would seem to contradict this argument.

I spend a little bit of time, especially as I am thinking of writing a pop book on globalisation, thinking about why this disjunction between reality and perceptions. It looks like there are a couple of things going on. First, it is not based on rational analysis, it is based on emotions.

“People believe what they desire the most or what they fear the most,” —- I quote John de la Fontaine in my book. So, on the desire side, for young MBA students, what could be more tickling to their fantasies than the notion that the world is their oyster? Or, for the people who hang out at Davos, the cosmocrats. Again, there is something really seductive about the idea that people there are deciding the destiny of everything without borders.

At the other extreme, the worst fears are job losses etc, so this exaggerated integration stuff keys on their fears. That’s a part of it.Another part of it is the desire to be cool. And then the third and most important thing is the techno-trance stuff.

George Orwell wrote about it sixty years ago about the number of accounts he had seen back in the 1940s of the world becoming one. This is not a new idea. People confuse connectivity, which certainly has improved, with the notion that this means perfect integration.

Take software services, the example that Friedman uses and overuses in his book. If you look at the McKinsey Nasscom study, the 2006 version, on what fraction of currently offshorable IT services are being offshored. The estimate is 11%-14%, which is a lot closer to my 10% idea.

Ah, but the whole point is that with connectivity, it is perfectly feasible for an Indian software services provider to service the world. If you look at the Indian software services industry and the moves to establish more of a presence in a deeper way than just body-shopping onsite, or in geographies other than India, even in software —- which is Friedman’s chosen example —- the currents seem to be running in the opposite direction.

So what happens? For example, if the rupee keeps appreciating, will the jobs go back?
I have been working on the TCS strategy for the last eight years and this is exactly one of the reasons for building up a presence in Latin America, thinking about Morocco etc. It would take a massive collapse of the dollar for a lot of back-end work and basic programming jobs to flow back to the US.

As Indian software companies move up the value chain, quite apart from trying to diversify geographic risk, the reason why jobs are being put up in the US is because of more face-time with clients.

That’s taking more of a proactive role, rather than saying “we will just do your programming according to your specifications”. We need more feet on the ground there and not just of the bodyshopping kind, but people who can act as business partners.

But is that really happening? Are companies really moving up the value chain? The impression that we get here is, in order to protect their margins because of rupee appreciation, they have been offshoring more work…
There are two kinds of businesses. There is the bread and butter stuff, which they have done historically, and with the rupee appreciating you know its like the point you make.

Some of that business is countercyclical, so the worse the pressure clients face in the west, the more they look for cost reductions. So that’s a huge help.

In terms of new areas, there clearly is a requirement to think of things that aren’t simply driven by least-cost location. I went into TCS archives and found the best example of upgrading. That’s relatively hard to debate —- revenues and profits, divided by the number of employees. In rupee terms, TCS has gone from a revenue of about $12,000 per employee per year, back in the early 1990s, to $50,000-$55,000 now. Accenture and IBM Global Services operate at $120,000-$180,000. I see headroom even now.

IBM is building a lot of their business in India. Maybe in a couple of years they will as big as TCS in India?
I know IBM well. I have spent a lot of time talking to Sam Palmisano. He says IBM is looking to narrow the cost disadvantage vis-a-vis Indian software companies.
This is something one can infer from the compensation surveys.

The last one that I saw had the MNCs leading in pay; their entry-level pay is roughly twice what Indian companies offer. So clearly IBM needs proximity on the arbitrage dimension, but how is IBM going to try and beat not just TCS, but also Infosys and Accenture, as it has hardware, services and software?

The current IBM strategy is very much an aggregation: How do we exploit economies of scope across three different businesses? To offer clients something other players can’t? I think obviously, there is some pressure on the wage pool, on the labour pool because of the fact that the MNCs are expanding so rapidly in India.

Salary differentials are large because a lot of people have a perception of glass ceiling. Some people still have doubts about some of the MNCs’ long-term commitment to India. Some people have doubts about how much training you get in an organisation. That goes from a standing start to one having a lot of people. This is one factor.

The second has to do it with basically the MNCs themselves having realised that they don’t want to play the same game. So they are playing a different game.

The third factor has to do with internal MNC constraints. While Palmisano is trying to build an Indian operation, he has more than 200,000 people in advance geographies. That in some sense are adding more value to the company than the Indian operations are. It is a very difficult tightrope to walk.

But the way Indian salaries are going up, will that iron out the differentials?
This is always one of the theories that I have used. OK, say, Indian salaries are 1X the US salaries and are growing at 15% a year, while US salaries are growing at 3% a year. Does it imply that in Y years the Indian advantage is over?

This, to me, is the basic arithmetic of the industry. Narayana Murthy (Infosys chairman) points out that wage costs are at about 15% of the total cost structure. So 15% multiplied by 15% looks a lot more formidable than 15% multiplied by 100%, if the non-wage components are not escalating at such a high speed.

Point number two is that there are some absolute availability constraints in the west. Through medium term forecasting, you know how many enrolled in the first year of computer science programmes in the US. At least you have a pretty good idea what the labour pool is going to be four years from now. If one looks at Europe, where the demographics are even worse, you know it’s not just salary escalation alone, it’s that coupled with availability constraints.

The third thing is that these straight-line forecasts basically ignore supply-side responsiveness. When I was growing up here, maybe this wasn’t true in general, but I still remember, out of curiosity, I used to read the matrimonial ads for personal consumption data at that point in time.

You know, what were people looking for in terms of their spouses for their sons and daughters? Tax collectors, customs officials etc. What do they look for now? Software engineers.

These are the best jobs in India, so the notion that there are zero supply-side responsiveness —- and that all excess demand is going to be met just with salary increases —-  I think ignores the basic principles of economics. It’s not just demand and supply that matters, it is the intersection of the two.

Just to go back to the TCS example, the managerial way of thinking about it is not that salaries are growing at 15% per year, it is that how quickly do I die as a result? It’s like, can I keep the topline growing faster than the cost-line? Over the last 15 years, TCS’ cost per employee has quadrupled in dollar terms, but its revenue per employee has gone up by a factor of five. And this is why TCS makes several folds more profit per employee now than back in the early 1990s.

How is it when compared with IBM and Accenture?
Comparable to Accenture in profit per employee, despite having a revenue per employee, if I recall the numbers correctly, at less than half. The quality advantage of development work done in India is under-appreciated. I see a lot of headroom as the brand perception of Indian companies further improves.

Which goes back to where we started this conversation. I mean, for a long time, software guys were just seen as writing code, they can’t deliver business solutions. Clearly ACNielsen doesn’t think so any longer —- or else they wouldn’t do the billion-dollar-plus deal with TCS.

This is one of those businesses where perception lags reality. There is significant evidence that for certain kinds of development projects the quality of work done in India is better than work done in other parts of the world. There are some challenges associated with change in perceptions. Indian majors as a group have gone from being perceived as just the back-end boys to they have something more that they can bring to the party.

Do you see an Infosys, TCS or Wipro acquiring companies like Capgemini? There has been some talk….
That’s like saying that I have this canoe which is quite buoyant, and so I am going to strap a leaky canoe to it and improve buoyancy further. You buy revenues, but you buy very, very troubled operations.

They are troubled not only because of cost structures, but because of the employee base in their overseas operations. These are employees who don’t expect to move, they stay put. These are employees who still have cultural stereotypes about Indian managers. And these are employees who are not used to working offshore.

My sense is that programmers in India have made a virtue out of a necessity and are much rigorous and much more willing to be compliant with step by step processes, which is why quality levels on maintenance projects are better.

Imagine trying to retrofit that onto cultures. Instead, companies on their own are migrating to larger deals.

I think where you would want to do deals is where there are some complementary skills —- things you don’t have in-house, rather than just as a way of supplementing what you have in-house. But that, too, poses its own set of integration challenges and may not work very well.

Do you feel that the barriers are falling because of certain disruptions? Were it not for the US subprime crisis, investors from the Gulf would not have been able to buy a huge stake in Citibank…
There was an awful lot of this going back before World War I. Particularly with Britain running huge pound sterling surpluses and making massive investments in prime assets in a lot of emerging economies.

This included not just colonial economies but places like Argentina, where they were never really the colonisers. They built the whole Argentine railway system, which at one time was the key asset in that country. That was all British-financed. A lot of this talk about global financial crisis being something new - there is no evidence that we are seeing a greater incidence of global financial crisis now than we did 100 years ago.

Barings Bank first went bust in 1890 as a result of problems in Argentina. Argentina itself was responsible for three major global financial crises in the 19th century. So all these people who look at current shows in the system and say “Wow this couldn’t have happened before,” should read some history. We can’t rely on personal history.

Where there is cash, it needs to be invested. So where there are global imbalances it will end up crossing borders. Different countries are responding differently.
An article in the International Herald Tribune recently mentioned how the Gulf is now investing much of its money in destinations other than US. This suggests that sources of capital still matter significantly.

So wherever we see anything across cross-borders, somebody will rush to say, Aha.. the world is flat. My point is not that there is no cross-border activity. My point is that if you are going to do cross-border activities thinking that you can ignore the borders, you are going end up in a big load of trouble. So I don’t think that when the Gulf guys are investing in Citigroup, I don’t think politics is absent from their calculations. A big part of a challenge for companies in this region, given 9/11 etc, has been, how do we paint ourselves as friends of US.

This was part of why, Emirates (the UAE airline) has been so careful to make sure that it doesn’t just order Airbuses. It actually splits its orders between Airbus and Boeing, as they need political support on both sides of the Atlantic.

k_vivek@dnaindia.net
j_satish@dnaindia.net