Businessweek Archives

The next step in outsourcing: Math

Tech services research underway in Bangalore |


| An opportunity for India and the US to come together

November 30, 2006

The next step in outsourcing: Math

Steve Hamm

If you want to find out what’s going to be the next trend in outsourcing, it pays to visit Bangalore. I dropped in on a 2-year-old startup that’s delivering analytics as a service. It’s Mu Sigma, with offices in Schaumberg, Illinois, and downtown Bangalore. Founder and CEO Dhiraj Rajaram, who spent time in the US at Wayne State (MS in computer science) and University of Chicago (MBA), and Booz Allen as a consultant, is in the process of moving back to India with the goal of managing his 120-person company to the 1000 mark in the next couple of years. His idea is that it doesn’t make sense for many companies to outsource their entire analytics department, but they can do a lot more analyzing at affordable prices if they augment their handful of high-priced math PHDs with teams of people in Bangalore who are adept at using the latest analytics software tools to spot opportunities in marketing, risk analysis, and supply chain optimization.

Rajaram calls it “information arbitrage.” His people support the client’s top analysts rather than trying to replace them. “We don’t do insight generation. We don’t have the knowledge of their industries and their companies to do that,” he says. “We’re humble enough to know we won’t go all the way.”

Right now, if a company taps one of the big consulting firms for this kind of service, they’re committing to hundreds of thousands of dollars in expenses. With Mu Sigma, projects can cost $75,000 or less. Put another way, instead of hiring junior analysts in the US for $120,000, they can benefit from labor arbitrage and get the same caliber of talent for $15,000 or $20,000 per year. His people typically have a masters in math or statistics or an MBA, but he has a handful of top people with PHDs as well. Rajaram says this really is augmentation work, rather than replacement of employees. So far, he claims, not a single client has reduced its workforce as a consequence of outsourcing work to his company. If he’s successful, I doubt he’ll keep that record intact.

In spite of its youth, Mu Sigma has quite a bit of credibility. Its customers include Microsoft, HP, Allstate, JP Morgan Chase, McDonalds, CDW, and IBM. Rajaram showed me a couple of case studies to show how useful outsourcing math can be. In one engagement, with a computer distribution company, Mu Sigma came up with a method for spotting customers who are in danger of being lost and taking just the right action to regain them. It also devised a method for the company to analyze customers and figure out when they are ready to graduate to a new generation or type of technology—so it can send in its high-priced sales consultants when the time is right.

Mu Sigma is a highly leveraged company. It has one employee in the US: Rajaram himself, and, of course, he’s leaving. He has been the company’s entire salesforce, but now he’s focusing on managing growth—partly by establishing a training program in Bangalore designed to prepare math grads with the domain knowledge they need to be really useful to clients. Mu Sigma will get new clients by word of mouth—a technique that has proven successful so far. The plan is to have just a few customers, but to develop deep, broad, and long lasting relationships with them.

Mu Sigma has already had one buyout offer, but Rajaram turned it down. He wants to see if he can build the world’s largest analysis factory first.

05:36 AM


TrackBack URL for this entry:

It makes lot of sense to outsource these services to those who can read numbers line "John Nash" since not many organizations have 1. dedicated people to do this analysis and data minning to get to meaningful information 2. these services are the ones that should be performed in lowest cost since they are not profit centers

Posted by: Salil at December 1, 2006 10:13 AM

Given this, please explain to me again why I should encourage my kids to major in Math?

Posted by: A Parent at December 1, 2006 11:21 AM

A interesting characteristic of MuSigma is that, that unlike other analytics firms they don’t use any buzzwords like ‘frameworks’,’solutions,’domain expertise’ ,’predictive’,’ transformation’,

’proprietery’ etc. It is plain vanilla top statistics brains on hire!

Posted by: Tharun at December 1, 2006 12:20 PM

If Americans retreat from doing anything Indians can do, it will be a sorry state of affairs.

Posted by: steve hamm at December 1, 2006 09:28 PM

Mr Rajaram has a damn good idea. Assuming that A Parent was not merely being sarcastic, here are three good reasons for students to Major in Math. One caveat is that I double-majored in business AND math. Having analytical abilities added to business skill is a very strong combination. First, it makes it easier to get in the door with top companies. Second, an undergrad math major will gave analytic capabilities comparable to many MBAs. This can present many career enhancing opportunities. Finally, someone on the inside has to understand the methodology, its uses and its drawbacks, and how it actually relates to the company hiring an outsourcing firm, or else the whole thing is just a big waste of money. Oh, and I think I read somewhere that his one and only saleperson was leaving the US. I estimate the chances that he will want a "math" person in this role to be roughly .9

Posted by: Sven at December 1, 2006 10:39 PM

I am wondering why I am spending money educating my son in the US since I know he wont get a job with Mu Sigma in the US.

Posted by: Radhakrishnan at December 2, 2006 12:01 AM

Too bad that 'Green Eyeshade, Inc.', there, can't devote its' mental energies to putting GM back on the path of fiscal righteousness etc...

Posted by: Bert at December 2, 2006 11:19 AM

To the Parent above make sure that your kid balances the analytical with management prowess and your kid will be fine.

Posted by: Van at December 2, 2006 03:29 PM

Notice that they are essentially doing the grunt work with little "insight"? I work in finance, and we outsource a decent amount of work to india, but believe me, they are not EVER going to take away the jobs that require legitimate insight. A math major is ALWAYS going to be in demand, not necessarily for the specific math skills that they have, but for their intellectual rigor, logical insight into NON MATH RELATED problems, amongst other reasons. Doing a math major is basically putting a sign on your head "I AM REALLY SMART". Math majors are hired in many non math related fields purely for this reason.

Posted by: Daniel at December 2, 2006 10:20 PM

Apart from statistical abilities which can be grown with education, India seems to have great mathematical capabilities, esp. in South India.

Posted by: Tirath at December 3, 2006 08:08 AM

One problem I have with the current outsourcing trends is that one day down the road, who is going to have the requisite skills to replace the Sr. Analysts still employed in the United States? Or - guess what position many of these Sr. Analysts had prior to being Sr. Analysts? Hmmmmmm...anyone? Oh yeah: Junior Analysts. You cannot get to a position as a Sr. Analyst if you never were in lower positions to know the ropes? At some point, Company's are going to have trouble finding Sr. Analysts, and once that occurrs, guess what company will be able to step in and provide that service? Mu Sigma. Boy - outsourcing is great.

Posted by: VT Biker at December 3, 2006 02:37 PM

what I could not understand is why somebody studying the same set of things, having the same amount of grey matter(noting that skin color difference does not cause any change in the IQ distribution curve anywhere in the world) would not be able to generate "insights" of the same calibre, given comparable exposure. The idea is probably not that some people "are not EVER going to take away the jobs that require legitimate insight" just because they happen to start working on simpler problems to begin their career with, but that a distribution of knowledge giving rise to competitive labor force only helps in generating more business innovation on the part of the challenged entity and thus ensuring that anybody good is always needed to implement newer things,regardless of which part of the world they come from. so smart people(notably americans here) will always find job, may be in newer fields at most, regardless of the amount of outsourcing at any level.

Posted by: kumaresh bagchi at December 3, 2006 02:53 PM

As a quantitative analyst I will tell you that Mu Sigma addresses a real need in US corporations. There is a tremendous amount of data that corporations are not mining due to lack of adequate quantitative talent. This is a cost effective way for corporations to implement fact based decision support in their organizations and leverage the rich data they are collecting.

Posted by: Zack Jamieson at December 3, 2006 07:31 PM

Steve, you comment: "If Americans retreat from doing anything Indians can do, it will be a sorry state of affairs."

Yes it will be a sorry state of affairs. But if potential employers have Indians lined up eager to do something for "$15,000 or $20,000 per year", then what other option do Americans realistically have?

I had a conversation a few weeks ago with a neighbor, a very bright freshman at a local university. When she mentioned that she liked math and was good at it, I choked back my cynicism and said this: "Math will always be a good thing to study. It's the lingua franca of science, and one of the primary bases of our civilization. Math is also a window into Truth and Beauty, and worth studying even for its own sake."

I hope I didn't give her bad advice.

Personally, I think that as soon as the factor-of-five disconnect between the rupee exchange rate and reality (i.e., why *do* Indian eggs cost one fifth what American eggs do? Are Indian chickens harder working? Better educated? CMM Level 5 certified? Is India a land of fabulous superabundance in chicken feed? Or is there smoke and mirrors with the currency?) collapses, things will get back to normal and Americans will be able to compete on an even basis.

Posted by: A Parent at December 4, 2006 10:58 AM

AParent , your comment on why bother with math is understandable given the hype surrounding the trend of outsourcing, but as a 7 year veteran of the indian software service industry i know the kind of work theat really gets done here (maintaince,enhancements) as compared with the work being done in the US(new product,technology development).

So atleast for the distance future tech skills, science and math skills will ensure a good

Now the flip side of this..

In the past few years the abundance of engineers in India meant that they would corner all the IT related jobs.

Non -engineering grads had a difficult time getting into the industry or they would get lesser pay for the same job.

Now non engineering grads in India can command good pay with a larger variety of outsourced work that is not related to coding.

Posted by: santhosh maruthi at December 6, 2006 06:31 AM

I have one question. Are Indian software and services to the US considered imports? If not, then why not? If yes, then shouldn't they be under review for some kind of import control like we have with all other imports?

Posted by: Steve Beihl at December 10, 2006 12:04 PM

Dear Parent,

I don't live in the US. I live in the Netherlands and I'm Indian. I'm a management consultant, who used to be an IT guy and before that, a chemical engineer.

I would strongly suggest that you tell all the kids you know to study Math, Physics and Chemistry. The reasons are possibly things you already know

1) They improve logical thought process development in things

2) They help kids understand how things work

3) They find it easier to pick up domain knowledge than others in industries as varied as Banking, Trading, Logistics, Farming, Retail and so on, as sector after sector becomes more and more data driven

I see this trend of "not studying Math" here in the Netherlands, and am worried by it. Kid here think that they'll sell stuff by looking good and do creative things. They can't even add up prices in shops in their heads any more. They need a calculator..

It's a sorry state of affairs, but I as a kid growing up in India used to think that the West was more intelligent at Science and Business than us Indians, because they had achieved so much - put a man on the moon several times, build such fantastic infrastructure and buildings.

I'd like the next generation to continue to feel the same way - while acknowledging that they too are building phenomenal dreams in India and elsewhere.

Thanks for reading this long,


Posted by: Arun Sadhashivan at December 14, 2006 04:36 AM

A Parent,

I believe your concerns about globalisation are valid to an extent in that they cause a temporary disrupment to local economies, with sometimes major impacts on individual lives. I also agree that some day the rupee-dollar parity will be achieved. So it is sad then that your understanding of currencies is so poor. Indian chickens do not live in air-conditioned comfort, nor are they fed genetically engineered feed, nor are they plucked and packaged by machines. With nearly a billion poor people desiring nothing more than two square meals a day, you would be surprised how much harder a dollar would work in India than it would in the US. Its hard to understand that if you have never been out of the developed world.

Posted by: Pippin Took at December 20, 2006 01:46 PM

"With nearly a billion poor people desiring nothing more than two square meals a day, you would be surprised how much harder a dollar would work in India than it would in the US."

And that is precisely the Indian Advantage. When hundreds of millions of people live on the edge of starvation, with no schools, no toilets, no electricity, it creates an opportunity to "benefit from labor arbitrage and get the same caliber of talent for $15,000 or $20,000 per year."

And those lucky few making the 15 or 20 thousand per year get to spend it all at the mall, instead of being taxed to build the infrastructure to bring their neighbors out of the Stone Age.

But you still haven't explained how it is that Indian chickens produce eggs *five times* more cost-effectively than American chickens. Have you seen pictures of American chicken farms? It would seem there's not a whole lot more you can *do* to a chicken than that!

Posted by: A Parent at December 22, 2006 11:48 AM

Sven's comment about "analytic capabilities (of) MBAs..." brought back a few not so pleasant memories from a while ago.

Back then, I used to teach at a top tier b-school in the Midwest and had a rather corpulent, poker-faced lawyer from New Hampshire in one of my MBA classes.

In one instance, after using MS-Excel Solver to tackle a rather messy non-linear product-mix cum pricing problem in a case we were analyzing, I asked my students what they would have done if they had to address the same issues in the good old days (i.e., before computers, MS-Excel and Solver were invented).

To close the discussion, I spent a couple of minutes on a slide or two, which showed the rather elementary principles of differential calculus involved in solving the linear-quadratic optimization problem (setting up the LaGrangean, taking first order derivatives, applying Karash-Kuhn Tucker conditions, etc.).

It wasn't rocket science at all, but it was enough to get the lawyer from NH (who had been cluelessly silent all semester long) all agitated and het-up.

He stormed out of my classroom, and headed straight for the Dean's office to complain. Apparently the Dean had promised him that there would be no "MATH" in the classroom, and since I had reneged on that promise, he wanted his tuition fees back. I got a memo from the Dean reading me the riot act that day and stayed away from trying to explain the mathematical underpinnings of anything to the students after that.

As long as the US high-school and university students are allowed to shirk away from mathematics, because they think its too geeky, or too complicated, etc., someone like Rajaram will always be able to step into the void and meet the need that US corporations have for the information, insights, intelligence,... that analytics could provide.

Incidentally, he ought to also open an office in Moscow or any other city in the CIS region – he might be able to get as good if not better mathematicians and analysts for one-tenth as much as he’s paying now.

And to A Parent, what do you have against Indian chickens, for crying out aloud? The day they get know more about foreign exchange rates, PPP adjustments, total factor productivity differentials, etc., you can bet they will go on “strike” (after all, they are Indian chickens, aren’t they?) and then you wont have anything to gripe about!

Posted by: No_math_in_the_classroom at December 23, 2006 08:56 AM

Hollywood Goes YouTube

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

blog comments powered by Disqus