Why are some companies reluctant to go offshore?
Well, there's more risk going offshore than onshore. It takes a tremendous amount of internal management to make this work. And because visa requirements are such a headache now, it's very difficult to bring your Indian suppliers on- and offshore. A lot of business people are afraid to go offshore because of privacy and intellectual-property rights. Also, some of the people that we've talked to are worried that the labor arbitrage won't be as favorable anymore, because Indian salaries are going up.
You mentioned that internal management issues could be holding people back. What are the main ones?
It takes a lot of coordination. Software development is hard enough when you do it in-house. It's hard when you outsource it to a domestic supplier, and it's even harder when you outsource it to an offshore supplier. The differences in time zones, culture, travel and communication can be conquered, but it requires internal management coordination to make all this work.
Despite these risks and difficulties, do you still think more companies should be trying to outsource offshore?
I think it's a viable option, and certainly, if I were in a Fortune 500 company, it would be an option that I would want to explore. This is still a very immature market. If you look back 15 years, we had the same issues when companies started outsourcing to domestic suppliers. There's a lot to learn at the beginning of an emerging market. The nice thing is that early adopters conquer that learning curve pretty quickly.
If you're interested in going offshore, how do you select an approach that provides the value your company is seeking, while also balancing costs and risks?
There are a couple of decisions you have to make up front. First of all, what do you want? Where are you going to go? Which supplier are you going to pick? But there are four main approaches for providing value: captive centers, joint ventures, build-operate-transfer and fee-for-service.
What are these four approaches, and how do they differ?
The first model that U.S. customers usually try is the fee-for-service model, because it has the lowest risk, and it doesn't require much up-front investment. In "fee for service," the customer pays a fee to a supplier in exchange for services. The fee may be based on a fixed price, time and materials, or even a retainer model. It's a low-risk approach; you're just testing the waters with that kind of a model. It's easy to disengage if it doesn't work out. The disadvantage is that in order to get the cost-saving, you have to get enough volume of work.
Click here to read a survey on how well CIOs handle outsourcing arrangements. So companies seeking to save money by using the most common model might not actually find it.
Not unless they conquer their learning curve and get their volumes up.
Can you give me examples of companies that have pursued the fee-for-service model?
We know a lot of customers that are satisfied with the fee-for-service model, but they might try out different suppliers before they find the right supplier. One Fortune 500 company we studied engaged four Indian suppliers, and had 17 pilot projects, in order to test proof-of-concept and who they really wanted as a partner. At the end of the day, they ended up selecting one large Indian supplier for a certain type of work and one small Indian supplier for another, and they're finally at the point now where they're achieving the cost-savings at the level of quality that they want.
How much are they saving, and how much has quality improved?
We've asked that question over and over again, and the savings are very difficult to calculate. Most of the time, all you're going to get from a program management officer is the differences in hourly costs. If you stand back and try to say what the total cost-savings are-well, it's very difficult to say there are any total cost-savings. The only people that give us any kind of benchmark are the research firms.
It sounds like there's a reason that smaller companies are reluctant to go offshore: You need a lot of volume to save any money. How much work do you need to give to the offshore firms?
If you want total cost-savings of 20 percent, and you're talking about application development, then between 50 and 100 full-time equivalents in the long run. I'm talking only about software development. Savings on call centers require a much smaller volume of work. Sometimes you get savings almost immediately. The easiest one to get savings on is call centers. If you do it domestically, it costs $25 per call. If you do it offshore, it costs $15 per call.
Let's talk about the other models.
The other models are also interesting. For example, there's the build-operate-transfer model. Here, what companies want to do, ultimately, is to have their own captive center, but they're afraid to do this themselves. They may not understand the culture or the politics or, for example, the graft they might have to pay to, say, the fire inspectors. So they hire the supplier to erect the facilities, wire the building and hire the employees; later, the supplier transfers both the ownership of the facility and the employees to the customer. We don't have any examples in our data of companies that have actually gone through the transfer; they said, "I don't know who to call when the lights go out, and the Indian supplier does."
The other two strategies are joint ventures and captive centers. We saw a lot of joint venture activity in the early nineties with domestic suppliers. The idea here is to take the customer's main expertise and combine it with the supplier's IT expertise and then service a vertical market. Those didn't do well in the nineties, so we'll see if it works with Indian suppliers.
Click here to read about Google experience outsourcing its billing.
What's the difficulty?
I think the customer becomes schizophrenic in these kinds of joint ventures. Let's say I'm a customer, and I create a joint venture with an Indian supplier. The first thing is, I'm their first customer. I have engaged the supplier because they're going to do some massive amount of work for me. But as a joint venture, I also expect the supplier to be able to turn around and get more customers outside of this relationship. Here's where it becomes schizophrenic: As soon as the supplier turns its attention outside, I'm likely to say, "Hey! What about us? You've got to deliver to us." It's very hard to play this game well.
What about captive centers, where a U.S. company owns the facility and all the employees there work for that company? Is that something that only very big companies could do?
It's something where you have to have a large commitment, a volume of work over the long haul. We know a lot of large IT suppliers are creating captive centers. We also see some organizations creating captive centers in places such as China, but their reason for going to China is that they ultimately want to sell their end product to the Chinese market. We know of an aerospace company that's setting up a captive center in Kuala Lumpur because they want to sell fighter planes to Malaysia, and one of the requirements of the Malaysian government is that you have to source in that country. So that's an example of selecting a location based on a business reason.
This article was originally published on 03-17-2005