Until now, the Indian IT services business gained ground on the back of a simple premise. You make a tonne of money by replacing a $125,000 resource with a $40,000 - $50,000 Indian resource. I would never have changed the model at Infosys, it was too successful there. Here, the premium the company commands on the market comes from the predictability of its revenues. And that predictability comes from adding, say 25,000 people to the business every year.
My belief is that the future of sourcing is going to change. In your industry do you pay a freelance journalist Rs 1,000 by the hour or do you pay for a finished article? So why is it different for corporates? Here employees are asked how many hours are you billing, not how much outcome are you producing and how quickly?
Ask any Indian IT services company what their Holy Grail is and they’ll come up with a pedestrian response that essentially amounts to this: Each day, you run faster than you did the previous day just to stay in the same place. In the case of services companies, this translates into exponentially adding people into the business to keep growing.
But I would like to argue that large companies have a premium to be inefficient. They bill clients by the hour. Therefore, the more employees a company has on its rolls, the more billable hours it has at its disposal. In an environment like this, the incentive to redesign or automate any process is close to zero. This model will not work in the longer run. It is killing innovation.
Eventually something’s got to give. Which is why, for almost 20 years now, Indian companies have aspired to create intellectual property (IP). The hope being that IP will eventually translate into packaged software products and subsequently free their fortunes from the tyranny of a linear business model. But here’s the plain truth. Barring the odd success story over the last 20 years, software product companies from India have failed.
To build products, you have to be natively ingrained in the culture of buyers who are going to use the product. That is an insight many Indian companies are coming to terms with.
I went back to the drawing board to take another stab at breaking the linearity of the business model we have traditionally worked with. I have come to believe the current model has to morph into one where a client starts to pay for a business outcome and not just the effort a vendor puts in. For instance, if I am offering my services to a mortgage provider, I get paid for each mortgage that originates through my processes — not for throwing 50-100 bodies at a problem. Once you start looking at the world from this perspective, there are two or three ways to reduce costs both for the client and yourself. The most obvious one is to redesign processes continuously until they reach optimum levels.
The other not so obvious one is through shared services. This concept has a lot of complex elements and involves sharing the same set of resources across multiple clients.
Consider a payroll application for instance. If I have multiple customers who want their payrolls processed, why should I create separate teams for each customer? Instead, if I can automate the process and create a set of people within my organisation to do the job for multiple clients, I stand a pretty damn good chance of breaking past the current linearity.
I used to think pricing was an art. I am now convinced it is a science. I’ll explain that. I grew up in a model that charged on a time and material basis. There services were priced based on a gut feel. When the circumstance changes, gut feel doesn’t work. In my case, I was moving to a shared services model and pricing had to be based on service level agreements. It’s much more complex and has a lot to do with statistical modelling. So imagine I run 100,000 transactions for a client this year. Next year, it may go up to 200,000. In the older scheme of things, I add 100 percent more people. In the shared services model, I can get the job done by adding perhaps 20 percent more people to my rolls.
You have to choose the right processes and the right customers. Traditional wisdom suggests that whatever process generates a competitive advantage is not outsourced. Because if you do, it gets packaged and commoditised in the future. Given this background, it is difficult to get a large customer to consider the shared services model.
But mid-sized and smaller companies are amenable to the idea. They understand that they can achieve the scale, size, operational efficiency and cost structure of a much larger entity by adopting a shared services model. As a vendor, I can take efficiencies, best practices, cost structures and everything else in between to a smaller entity at SME prices. Therefore, the choice of the customer has to be very different for me. Like I said, the benefits to my business are immediately obvious. In case of the client, I can tell them that in this scheme of things, I can keep the price flat, perhaps, even reduce it by 1-2 percent.
I see three problems here. People’s mindsets are difficult to change. Rightly or wrongly, we have defined customer service as doing whatever the customer wants. As a country, we do not have the cultural firmness to tell somebody they ought to look at things differently. The shared services model does not work if you cannot go up to a client and say “this is what I am going to give you”.
Leaders identify themselves by the customer they work for — JP Morgan, American Express, GE, etc. — and the number of people reporting into them.
This has to change. In the model I am arguing for, power derives not from the number of people or kinds of clients being managed, but from the outcomes a leader generates. If it can’t be seen that way, this model fails.
And finally, customers are not always rational. I cannot understand why a manufacturing or financial services company ought to be worried about how their procurement or accounts payable processes work. The problem is, the security departments at these companies want captive offsites with access provided only to people working on their project. Customers are actually shooting themselves in the foot by not evaluating what processes demand high security and what doesn’t. If they weren’t so irrational, I could manage my resources better and pass the savings on to them.(As told to Mitu Jayashankar)
(This story appears in the 14 August, 2009 issue of Forbes India. You can buy our tablet version from Magzter.com. To visit our Archives, click here.)