Monday, December 18, 2006

Falling off the offshore bandwagon

I head the IT function at an early stage media/ advertising startup. One of the reasons I got hired,I presume, was the fact that I had worked on both sides of the offshore equation and was an Indian.

I am a firm believer in specialization and core competencies. So, I was not looking to build a large IT department. I am happy to outsource. And, offshoring certainly figured in my and our management's plan. But, my experience in the last three months have been really bad.

Our first round of vendor selection removed the most obviously incapable guys. We narrowed our list to about three vendors (all CMM Level 5 etc...)- the range of costs were between 200k to 800K. It made us wonder if our requests were too vague or if vendors were signaling what they wanted the project to be. Mind you the hourly rate of these vendors were almost the same - +/- 2 bucks/hr.

And, as if this variance was not enough there were lot of humorous errors in the proposals. One vendor offered to staff culturally sensitive french speaking resources on our project here in Midwest US. And, other, in response to a question about their employee turnover rate, indicated that their retention rate was one of the lowest in the industry.

At this point we started looking at a software package with the help of a local software reseller. And, at the same time we decided to define the scope of the implementation better in an effort to reduce the variance. The reseller's quote for the implementation was about 1000 hrs. We then proceeded to seek a revised quote for the project (which now was a product implementation). One of the three offshore vendors decided to drop off because they did not have any experience in the product. Another said they will partner with another firm since they did not have any experience but wanted to build expertise in that area. We said that was not ideal but we were willing to hear.

The one guy who had experience and expertise was the guy who had quoted 800k earlier. They are a 30000 person / $1 b company. This time they said 235k. At that price the cost was higher than the local provider. And, the number came with no information on the effort estimate or resource allocation. When we pushed them hard they gave an effort estimate of 6250 hrs - over six times that of the local vendor.

Meanwhile, the other guy who was going to get a partner got one but did not know the partner's name. Yes, that is right. They got the partner's name wrong during the conference call. And, instead of trying to build our confidence in them the vendor did the exact opposite. They would not return our calls, turn in the proposals on time. We decided to give up on these guys and seek a proposal from another small vendor with offices in US and India.

We tried to get an understanding of the other quote. We asked them if they had accounted for our resources participating in the effort. Initially, they said they had. Then when we told them their effort estimate was six times that of the local vendor they said they had not accounted for our resources. So, they reduced the effort by about 1250 and offered us a new price of 185k. In the meanwhile, I opened Excel and tried use the numbers, rates and the onshore/offshore effort ratio they had provided to make sense of the numbers. And, what I found was that the numbers they quoted were mathematically impossible for the rates, effort and onshore/offshore effort ratio they had indicated.

Maybe, the local provider was being aggressive and had some of the effort padded in their hourly rate. The offshore guys lead with their low hourly rates and pad in the effort estimate. I am guessing the guys in the big corporations just look at the hourly rates and claim savings. But, do cost savings exist? In our case they did not. Not only that, none of these vendors provided cost, effort and resource estimate. It looks like these vendors have grown complacent and don't bother to check the proposal for cut and paste error or if the numbers add up. And, the numbers did not add up.

No comments: