) global banking unit.
In a wide-ranging interview with BusinessWeek Senior Writer Pete Engardio, Marovitz discussed the financial pressures that are driving banks to shift skilled work abroad, lessons he has learned about how to do offshore outsourcing right, and what it's like to work with Russian software developers. German-based Deutsche Bank is using a staff of 150 at Luxoft, a Russian information-technology services company, for 27 different projects, such as a system for managing client relationships. The bank also has a lot of back-office and does analytical work in India.
What are some the biggest pressures behind the push to offshore outsourcing?
One of the principle forces driving outsourcing is cost. But outsourcing also deals with the question of core competencies. In other words, on which products do firms wish to focus, and what kinds of knowhow can be used from external providers? As a complex investment bank, you have to run increasingly big, mission-critical technology enterprises, and do it cheaply and efficiently. This means they have to think very hard about what they do for a living. Some financial institutions realized that they were becoming technology entities as well as banks, which is not always desirable.
What are the cost pressures?
Other than people, it is [information technology]. IT can consume 10% to 15% of the revenues across large firms. That's a big chunk.
Isn't this really about cutting labor costs?
There is a window of opportunity right now to take advantage of labor arbitrage. But in 12 or 15 years, costs in India, for example, may be on par with London or New York -- the labor arbitrage story is temporary.
The real story is about recruiting talent. China and India are the two largest labor markets in the world. If you are only using people in London, Chicago, or San Francisco, you will lose out because these places are a tiny part of the population. You need to recruit in Pune, Harbin, and Shanghai.
Does offshore outsourcing give you a competitive edge?
The issue is that if you don't do it, you won't survive.
Do you expect that outsourcing back-office work will boost productivity down the road?
Measuring productivity in business processes is a difficult calculus. It's not difficult on the IT side, where we have a number of years' experience under our belt. Where it gets more complicated is in higher-value work, such as research, which requires a more qualitative assessment. Our numbers certainly look better. We are saving costs. But it will be interesting to see how far we can change the business model.
What are some of the major pitfalls to avoid?
Outsourcing a mess is a real recipe for disaster. If you have a process that is not working, and then decide to outsource, you are heading for a train wreck.
Any teething problems that you went through?
One issue, particularly in India, is that sometimes the level of procedural rigor actually is too high. All of the major Indian companies have CM Level 5 [a quality standard for software processes developed by Carnegie Mellon University] people. However, that requires a high degree of documentation and very formal processes. If you take a U.S. company that is run ad hoc, and then outsource to India, you find the Indians are so much better that there is no way to pass on your own processes. So you can lose some control.
Deutsche Bank has its own development center in Moscow. Why also outsource software in Russia?
I can't imagine any organization the size of DB not having a combination of resources onshore and offshore, in-house and outsourced. To do work in India, you must have scale. If you have a 25-person operation and put nine of those jobs in India, that's a bad idea. While the labor cost is attractive, the infrastructure cost will never break even. You need to have a group chunky enough to amortize the costs. In Russia, the team is relatively small.
What is the difference between working with Russian developers and Indians?
Part of why we liked Russia initially is that we were working with a small team of 30. We wanted to accelerate what we were doing developmentally and couldn't afford to do it onshore. We wanted to get more bang for the buck. We looked hard at India, but for a group that small we could never make it work. We wanted to start with 10 or 15 guys.
Because Moscow is only three hours from London, it was easy to get in and out. We could get a small group there without being burdened with huge costs.
In terms of skill levels, both Russia and India have strengths. The math and science tradition of the old Soviet Union is a plus. There is a real history of technical innovation in Russia, and there is almost a demand by the Russians we work with to be innovative and creative. In India, this is less the case. The Indians are incredibly process-focused. They work with great care, but with the plan that you give them.