I was chatting with a friend at a mid-sized software company in Bangalore; he said that foreign clients were increasingly insistent that they would pay a fixed amount (typically $20/hour) for the projects he had been working on in the last year or so, regardless of the project member being a developer, manager, or tester. We were joking that software billing rates in another year's time would fall to McDonald's hourly wage. Anecdotal evidence from friends and ex-colleagues working on software projects for foreign clients in India seem to indicate severe reduction in hourly rates and head counts due to client pressure in the last one year.
Well everybody knew that it was a scam - the software boom in the nineties meant easy money for most outsourcing companies. I remember a cousin of mine mentioning the case of a Japanese client who was charged $8000 for changing the text on a dialog box. Not a single project ever ran without being overstaffed and the client being fleeced. There is enough
schadenfreude to go around. While computers and software are supposed to make us more efficient and reduce unnecessary man power, Indian companies ironically used the software boom to increase the number of people employed. A person's worth at an India software service company was typically measured by "
How many people report to you ?". Incidentally Sloka Telecom founder Sujai Karampuri also wrote "
Why do we have so many jobs in Bangalore ?", which explains the dynamics of outsourcing quite well, though it's not comprehensive. The less said about the daylight robbery that went on in the name of ERP, the better.
A recent
issue of Forbes (Indian Ed) covered the changes the management at Indian IT services bellwether Infosys is considering in the face of falling revenues. There seems to general agreement among the top executives that revenues based on project head count are getting squeezed and the company needs to find alternate ways of generating revenue.
...
Some indications of the emerging, new Infosys are already coming in. Ask Dhar. His elevation to the EC came just a year before the worst downturn in living history hit the world. It has been baptism by fire. And the decisions he has been taking have been unlike any at Infosys of the past.
Dhar found himself in a rather awkward spot with one of Infosys’ oldest customers — a telecom giant in midland Europe. The client was under serious pressure to reduce costs and the Infosys contract was in peril. Dhar knew he had to act fast. He was aware that a host of rivals had already offered to cut their price by 25-30 percent. If Dhar did not respond on time, Infosys would be edged out of other juicy contracts that were on the table.
He then did what would have been once unthinkable inside Infosys. Rather than reduce his per hour billing rate (which would have hit revenues and profits), Dhar decided to change the way the client was billed.
For a third of the value of the contract, he promised the client an upfront cost saving. Infosys would now charge the client on the number of technical problems it would solve instead of charging for the number of hours it worked in doing that. In sum, he was proposing an outcome-based model.
Now, the risks were considerable. To ensure that they still earned a profit from the account, Dhar had to bet on far fewer engineers than usual to use their native intelligence to pull off the job. But there was no guarantee that they would be able to accurately predict the rhythm of the job. In fact, in the first six months, the company might even lose some money working this way. But in the long run, Dhar believed it was the best thing for both the client and his company. “It’s always been at the back of our mind to move to this model, but we never had the incentive to do it. This downturn has given us the reason,” says Dhar.
...
Gopalakrishnan has put together a five point plan for getting Infosys ready for the next stage. This includes increasing the share of higher value services like consulting, delinking revenue growth from staff addition, pursuing large deals of above $500 million, improving efficiency and finding new locations for talent.
Update (23'June 2009): Hacker News reader edw519 makes a great observation in a
comment on this post: