65,000 BPO employees by 2016 or USD 500 million in revenues?

Posted on October 21, 2011 / 2 Comments

The Sri Lanka Institute of Chartered Accountants’ annual conference started yesterday. Big do, with 1,100 participants paying LKR 17k+. Perhaps one of the largest gatherings of professionals in Sri Lanka. I was asked to speak on the subject of BPOs and KPOs on a panel this morning.

I started with the government’s target of USD 2 billion in export earnings in 2016 from the IT and BPO sector. That’s roughly five times the level of export earnings in 2010, according to figures given to me by ICTA: close to USD 100 million from BPOs (13,000 employees) and USD 300 million from software (27,000 employees). To simplify I will maintain the 1:3 ratio.

That means that the BPO sector must grow fivefold from now to 2016. At current levels, this means 65,000 employees producing USD 500 million by 2016.

Before going into the question of whether Sri Lanka can produce 65,000 people to work in BPOs, given our constraints in people and skills according to AT Kearney, I raised a question I have raised before: the low revenue per employee in the sector. Using ICTA’s numbers I get an average monthly revenue per BOP employee of LKR 70k. Very low, I have argued. Too low to support quality operations.

So my suggestion was that we do not get hung up on the 65k employees issue. Instead we should focus on upping the revenue/employee so that we can still earn USD 500 million in export earnings with fewer than 65k employees. That way we can pay the employees better too.

Other related writing is in LBO:
4 July 2011
5 July 2010
1 March 2010

2 Comments

Kirigalpoththa
October 21, 2011, 4:39 pm
Yes, the 70k is not salary. it must compensate other costs of the organization as well. So your arguement is quite valid.

I think target should be high end KPOs like AMBA Research, where you earn more revenue per employee.

btw any idea on the figures on per employee revenue for other countries …say India?
Rohan Samarajiva
October 21, 2011, 4:44 pm
Sorry, no comparators. I suppose one could get the data from sites like NASSCOM