Hidden costs of offshore software development
Rocky pointed to an article called “Considerations in Evaluating Onshore versus Offshore Software Development” which I wholeheartedly agree with. Having led a distributed team in India and two time zones of the US, I can tell you that its not all cake.
Some of the problems we came across ranged from the typical cultural differences to actual management issues within the Indian company.
As far as cultural issues are concerned, it would be typical that the Indian team was given an unrealistic deadline ( mea culpa ) , and everyone would sign up for it without much feedback in terms of how possible the deadline was in the first place. I’ve seen this happen with several organizations and as best as I can tell, there is a real disconnect in the feedback cycle. So when asked ” Can you get this done” the answer was always “Yes”. No matter how ludicrous the schedule.
Another important aspect is the preference of Indian teams for very structured and requirements driven development. I have seen only 2 out of 20 developers who were able to incorporate agile methods. For the most part companies thrive on CMM certifications and “big design / big process up front” development. It is so much part of the culture that I had people leave my team to join CMM Level 5 companies. In a way that is understandable, because being thousands of miles removed from the customer makes it hard to be agile, so big design up front goes a long way toward saving the programmers sanity.
Last but not least another cultural observation. This particular experience may have been due to our Indian vendor / partner having some pretty weak technical management. Whenever there was an unknown factor or an architectural decision, our team would debate as a group for hours on end to find an appropriate solution. That sort of discussion happened several times a week and impacted the schedule quite a bit. I wonder if that something only germane to our specific partner org .
In any case, after months of schedule delays I decided to move the production of a particular piece back to our US team and had made more progress in 2 weeks with 5 developers locally than a team of 20 devs and qa people had made in India in 2.5 months.
Makes me wonder if my experience was just an extreme case. If you have had successful offshore projects, I’d love to hear how it was done. And did you save money in the end ?
Comments
Comment from Thomas
Time October 29, 2007 at 7:06 pm
Seriously I am not kidding thats what happened. And you are right on the money in regards to the screening process. After it was all said and done we came to a similar conclusion. The company was a wholly owned subsidiary of ours and unforunately the people doing all the screening were VB6 / ASP coders, not .NET. Thanks for your comments.
Comment from Shankara Narayanan
Time October 30, 2007 at 6:25 am
Hi,
I have some thoughts on this on my blog. Do check it out. I had success in running some offshore projects. The success factor depends on various different aspects as you might be aware. But some of the points what you have told is true and I personally in my experience have come across the same situations and have also overcome the same. If you need more information – you can definitely contact me on my personal email id.
Your screening process for your offshore team should definitely be reconsidered.
Pingback from TechMount » Archive » Daily Friction #297
Time November 9, 2007 at 1:37 am
[...] Hidden costs of offshore software development – Having led a distributed team in India and two time zones of the US, I can tell you that its not all cake. Some of the problems we came across ranged from the typical cultural differences to actual management issues within the Indian company. [...]
Comment from What a crab is this?
Time November 24, 2007 at 1:47 am
ohh….internet is full of s..t like this…
some usefless article..i spent around 3 mins
waste of 3 mins + 1 min(this posting)
Comment from Heine
Time January 17, 2008 at 2:58 pm
I will agree with this article whole heartily. Having experienced this first hand, offshore development is more like Offshort development. Teams of people over there seem to over engineer everything. Give the example of the following product:
1 Small Bus
Color: Yellow
Decks: 1
Wheels: 4
Passengers : 30
Seatbelts: Yes
Emergency Exit: Yes
Send this requirement to India and you get this:
1 large overly complicated bus
Color: Well is depends on which hue and saturation you want.
Decks: A minimum of 1, however due to the necessity of the potential overload of the bus they built 5 levels on top of the first level without any stabilizers to hold it up.
Wheels: They determined that they needed 8 wheels to drive the bus. 4 was just not enough, so they created 4 extra wheels to handle any “unknowns” in the road.
Passengers: Since the original requirement was 30, they figured in 6 months the bus might hold 300 people, so they build the extra decks (see above) as well as compacted the space between each row to hold more passengers per deck.
Seatbelts: Yes, but each belt will hold 5 people versus 1, for more efficiency and it will be cheaper.
Emergency Exit: This was wrapped in exception handling which did nothing but stream out stupid data which has nothing to do with the bus itself, and is only there for show. The door does not even work.
The point here:
Don’t use offshore.
Write a comment
You need to login to post comments!
Comment from Adel
Time October 29, 2007 at 6:16 pm
India is dying… the culture is so diverse between US and India you should try off shoring as it will of course benefit you specially on the long run, but choose to offshore to a less distant country, will much alike culture and people easier to adopt your needs.
i encourage you to look for development partners in Egypt… you can contact me for that
also truth to be said your experience for the most part is extreme come on.. from 20 dev / 2.5 month vs. 5 devs / 2 weeks
you should consider more efficient screening process.