The previous issue of Knowledge at Wharton featured the first part of a conversation between Wharton professor Ravi Aron and Raman Roy, who helped American Express and General Electric set up back-office operations in India and then went on to launch Spectramind, an outsourcing start-up that was later acquired by Indian software giant Wipro. Towards the end of that discussion, Roy spoke about that challenges that international companies face in setting up captive service centers in developing countries. Now read on:


Aron: Another alternative open to a cash rich company — say, a large insurance firm or a retail bank — would be to buy a company, run it jointly for one year, and then in-source it — the buy-operate-transfer (BOT) model. What would be the problems in buying an existing BPO firm and in-sourcing it within 24 months?


Roy: That is a strong possibility. A lot of people are trying various combinations of BOT models. Those models are in operation, and people have set up joint ventures with local partners. Some of these ventures have succeeded, and others have failed because of the lack of a cultural fit. That takes us to the question of what it takes for an alliance to succeed — whether it is a joint venture, a BOT venture or a direct acquisition.  


Each of these captive center models requires a huge, up-front investment as you need to plan ahead at least for three to five years. You will, for instance, have to create a 1,000-seat facility on day one. You will have to put in a leadership team and have a long- to medium-term perspective from the outset. All this requires a significant upfront investment. In contrast, a third-party service provider offers its clients a so-called pay-by-the-drink option. If it doesn’t succeed, the clients have no financial or other liability.


Simultaneously, one of the things we observe as students of this industry is that Wall Street today is challenging the headcount that companies have on their books. Wall Street doesn’t distinguish between headcount out of India and headcount in the U.S. It just says “You have 23,000 people on your books; what are you doing about it?” So the leadership of the company commits to cutting headcount by 10%. When you deal with a third party, that headcount is on our books. So, the combination of ‘pay by the drink’ and the pressures from Wall Street makes the third party alternative an attractive proposition.


Add on top of that what a company like Wipro-Spectramind brings to the party. Let us now come to a key issue – the learning from past mistakes and the repository of learning that we carry around in our heads. Our management team has made its share of mistakes – I have bruises and wounds from the mistakes that I’ve made.  But thanks to the rich learning that came out of these experiences, Wipro-Spectramind is yet to have a major failure, primarily because we are able to bring our knowledge and learning to the party. It isn’t easy for a company to build that up from scratch. Remember, the deeper the experience, the scarcer the talent becomes at the top — and to find leadership that has the expertise of operating in India and doing international work out of India is a very high barrier to scale. This is what makes the third-party option very attractive.


Aron: The downside of dealing with a third-party BPO provider is the degree to which the user firm gives up control. Does the lower level of control lead to a greater sense of risk for user companies?


Roy: It does. That is one of the basic aspects of the relationship with which a lot of companies come to us. As I see it, several issues are involved. One is the aspirations of the local manager. The local Indian manager for these companies sees India as a small market on a global platform. So the moment that a company decides to use the Indian marketplace to tap the intellectual capital for BPO and other activities, the local manager invariably aspires to run it, so that his career can grow. The second aspect is the international managers whose operations are going to be executed in India. If you were to get such executives to think about the initiative as, “These are still my people, and they happen to work in India,” it is an easier sell than to get them to “outsource” to a third party. So, the overseas managers get to say, “I now have an arm in India, and I now have 300 people there who report to me.”


Local managers, international managers, and then comes the third aspect – the consultants. As soon as the consultants realize that their client is looking at these options, they will come in and say “We will help you decide – you guys don’t understand India, we do, we will help you make the decision.” In this format, some of the models that I’ve seen consultants come up with are exceedingly complex.


So, that is the mix of factors that has to be taken into consideration whether it is the BOT model or the joint-venture model. All these factors, when taken together with the control aspect that you mentioned, raises the fundamental question for the user firm: “How do I control it?” As I look at the past and the present and into the crystal ball for the future, I think that companies need to overcome their control mania and find ways of handling local managers and their “empire building” aspirations as well as the desires of Indian managers that they have as a part of the team.  


One more factor generally adds to the complexity of the situation. Many international companies have an expatriate Indian (or a non-resident Indian) as part of their management teams. Now Indians are everywhere. But Indian managers who left the country 20 years ago don’t admit that their knowledge of the country may be outdated because they left India so long ago. Instead, they usually claim to be experts on India. That complicates matters further.


For all these reasons, I think the perfect model has yet to emerge, because there is no tried and tested formula that specifies what is the best option. Companies are trying out various things. Some companies will find a hybrid solution that gives them some control. For example, they may run a captive center with 200 people, but they may go to a third party provider to add another 500 people. In other words, some companies have decided to go part captive, part third party. And there are some other companies that have set up pure captives. There are people negotiating BOT deals, and we have customers who have said, “We will give you business, but we don’t know how to set up an operations unit in India. Will you, as Wipro-Spectramind, help us set up an operation in India? We will commit to continue business with you while we set up our own captive capacity.”   


Aron: Since we’re talking about captives versus third-party scenarios, this is a burning issue for many large corporations: Would a third-party provider be able to offer higher productivity or quality than captive centers?


Roy: The short answer is ‘yes’. Let me give you my insights. Having run two captive centers, I know that the tendency among captives is to run them as cost centers. As a cost center, you tend to take some costs for granted, and even after spending a lot, you can still show tremendous gains of off-shoring to India. Third party centers can also use the same assets (like seats/PCs etc) across multiple clients in multiple processes and multiple time zones allowing the cost to be distributed across multiple clients. I have gone on record with some of the captives, and told them that Wipro Spectramind is willing to do business at the cost structure that they have to date. And I will do fine on that, because their costs are aligned to the markets they serve – internal organizations. On productivity also, we mostly match or exceed the captive centers’ numbers.


Aron: Why?


Roy: Consider productivity as a cost per transaction. Captives tend to be cost-plus units. But we are not just a profit center, we cater to the uncertainties of a free market and still have to be competitive and therefore we tend to focus a lot more on training and our asset utilization for productivity. It is the basic distinction between a cost center and a profit center. Some world-class companies will be able to do as we do even with their captives. They challenge their management teams to meet the same norms as that of a profit center, but there are other companies who say, “Yes, our captive costs us more than what it costs you, but we have greater control of the captive than we have of you.”


Aron: When you took the plunge to start Spectramind, you were not only starting a new firm, but you were setting up a different kind of industry. How did this play out when you tried to hire people to join Spectramind? How difficult was it?


Roy: The spiel I used at that time was to show them our bank statement. We had motivated our investors to put the money in our account up-front. And so I would say to potential employees: “Listen, I have enough money. Here is a statement showing how much money we have in the bank. I have enough money to pay your salary for 16 to 18 months even if we do not get a single customer. If we do not get customers, we will all go and look for jobs two years later. But if we get customers, that will bring in revenues, and that will give us more money to pay salaries.” That is how we hired the first 50 people; they became co-owners in the company, and our success meant success to them. Fortunately, today we have delivered on that.


Aron: Did you plan Spectramind purely as a voice-plus-low-end-processes provider, or did you expect to see it execute as rich a set of functions as the American Express and GE captive centers?


Roy: Our desire was to do value-added work. It was very aspirational, but we saw clearly that it would have to be built up. It would not happen overnight; the customers would have to develop faith in us, and they were not going to give us high value-added work from day one. So it was a combination of what they were willing to give us and our will to achieve our aspirations in small, baby steps that resulted in the first few deals and the growth that we achieved thereafter.


Aron: But you did think that at some point you would move to the value-added-service end of the spectrum?


Roy: Definitely. That is where the skill set of this management team was. If a company were trying to set up a captive, where would they get this amount of depth in the management team? That was our value proposition, and we saw that it would be a compelling argument in favor of transferring value-added work to us.

Aron: One of the questions that C-Level executives wrestle with all the time is this: Can I ever give away – that is, outsource – to a third party a strategically important process that I am now executing internally? This translates into a question for you: Do you see a third party company in the foreseeable future execute the kind of processes that the GE captive centers in India are executing?


Roy: Yes, most definitely. It is important to understand that this is not a routine vendor relationship. For the success of our customers as well as our own, we have to architect a partnership. And as a part of that partnership, getting their core processes and their value-added processes outsourced is integral. I will give you an example. For a genomics knowledge management company in the U.S., we actually hired PhDs in mammalian molecular biology, who populated and maintain a genomic database, which is a core part of our customer’s business.


Aron: You had hired about 50 employees and soon after that you were looking for your second round of funding. That was when Azim Premji, who heads Wipro, came to visit Spectramind. What was the purpose of this round of funding, and what roles did HDFC, ChrysCapital, and Goldman Sachs play in the transaction?


Roy: When we came up for our second round of funding, we were running about one year ahead of plan in terms of our growth as we went into our second year. We were running out of space at our first facility, which was in Delhi, and we wanted to set up a second facility. Based on the success that we’d had, our investors were happy to invest more money, but our thinking was to see if we could bring in an investor who could offer more than the existing investors had already brought to the party.


We had HDFC as an investor, and that was done to get some credibility in the market, because HDFC is a big name. And we had ChrysCapital, which at that time was a small venture fund relatively unknown to the world, but they had given us the money, and what we liked about them was the quality of their management — they had fire in their belly.

We engaged Goldman Sachs’s local arm to help us raise the second round of funding. We jumped through the hoops of designing our information memo and meeting multiple companies. As a result, we uncovered some great options. We found companies that wanted to offer us some very decent valuations and competitive rates at which they were willing to invest the next $10 million – which is what we needed for the second facility.


Meanwhile at an office in Bangalore, Wipro was evaluating its own options to enter the BPO arena; and they thought we sounded interesting. And one fine day, Azim Premji showed up here. (Ed note: Wipro in 2003 has annual revenues of more than $900 million, and its chairman Azim Premji is often described as the richest man in India.) For us, Premji is a demigod. The mere fact that he was going to visit us, a small venture-funded company, was a great thing. All of us put on our best ties and our best shirts to meet Premji—-not knowing how much fact-finding Wipro had already done. As a parting comment, he said, “I know you are looking for second round funding, and we may be interested.” That is what started it. He put us in touch with some of his people, which eventually ended with Wipro taking 24% of the company.


Aron: Before Wipro acquired Spectramind, what was the compay’s culture like? Was it hierarchical? Were you personally involved in running the day-to-day operations at a tactical and operational level? Or did you entrust it to lieutenants while you did business development? How was it managed?


Roy: To answer your second question, I was fully involved and continue to be deeply involved in running day-to-day operations. As my team keeps telling me, I have specific knowledge in my head, and I have to use it for the benefit of our company.


But the first half of your question is incredibly interesting — the aspect of culture. One of the things that Wipro checked before investing in Spectramind was the cultural fit. For Wipro it was hugely important that the two cultures must fit, and for us, too, the cultural aspect was crucial. We came from diverse backgrounds, there was a set of values by which we set up Spectramind, and by which we continued to run the company. The leadership team at Wipro spent an enormous amount of time with me and my management team between the time that they acquired 24% of Spectramind initially and then increased their stake to 100%. Part of the reason was to find out whether we could be part of such a huge corporation and whether we would fit in. That was a very important issue in their mind too.


Premji once said something that I was not smart enough to see till he said it, but which I later validated. During one of his visits, he stood on the shop floor at Spectramind, and he turned to me and said, “You know, Raman, I feel as if I am in a Wipro company.” Until then I had never looked at it like that. But later, when I visited one of Wipro’s campuses and walked around and met people, those words kept coming back to me. I thought, “It feels like I am in a Spectramind office.” That is what it means to have cultural fit between two organizations. We have continued to build on that strength, which we think has been hugely beneficial for both Wipro and Spectramind.


Knowledge at Wharton: You said earlier that you have bruises and wounds from mistakes you have made. What were those mistakes, and how would you now do things differently?


Roy: The major thing that I would do differently is something that I implemented in GE after learning my lesson at American Express. Back then, we let a lot of the folks in New York make decisions about which way we would run certain things in India. We would have these people visit us in India, and then they would make decisions in conference rooms in five-star hotels, rather than walking the streets with us to make those decisions. I think I was remiss in not opposing some of the wrong things that were decided. These errors played a major role in causing some delays and our ability to showcase what could be done out of India.


We also made mistakes in the way we handled customers, the kind of equipment we bought, the way we trusted suppliers and vendors, and also in training and hiring people. We ran a 24/7 operation. I made the mistake of hiring some top-notch people who wanted to work only during the day from 10:00 AM to 6:00 PM while much of our business with clients in other countries was done when it was night in India but daytime in the West. You may have to compromise on the quality of some of the people, but you do need people who are willing to work when it is time to do business; you can’t bring in business to suit the time when people are available. So there were multiple aspects to my mistakes, and I learned from them. It taught me how to transition the business, choose the right technology options, accept certain types of work and decline others, and so on.


Aron: What was the principal objective with which you started Spectramind, and to what extent has this been satisfied?


Roy: We had very clear objectives in setting up Spectramind. We wanted to demonstrate what could be done out of India. We were all big believers in the capability of the Indian workforce. I wanted to be able to tell my grandchildren, “Your grandfather played a role in creating this company.” I’m not trying to say we weren’t trying to make money—that was a driver as well. But that was not the main driver. It was the idea of creating something new in India, and the ability to say that we contributed to its creation. That is why partnering with Wipro was a movement in the right direction. There were some things we could do as a start-up – but this business requires deep pockets. It requires something more than business acumen, and Wipro has played a big role in bringing that to the table.