|Outsourcing research and development work
||18 Oct 05
I came across this great post by Luiz Esmiralha in the pragprog list.
>> On Sun, Oct 16, 2005 at 11:48:34PM -0200, Luiz Esmiralha wrote:
>> > Maybe another question can help clarify the issue.
>> > Why pay 100 grand a year for a programmer just to make a guy in the US
>> > or Europe happy if I can get the same job paying 20 grand a year for a
>> > guy in India?
>> Well those aren't anything near the figures. I know someone who is
>> the IT director for a UK company that uses outsourcing. The financial
>> savings are around 15%. If he sets up a UK shop in one of the more
>> deprived areas of the UK he will be able to get to somewhere near that
>> figure - lower salaries, govt grants etc. As a result he is seriously
>> looking at doing just that.
In a recent survey, average savings from offshoring are around a bit
less than 10%. 5% of the companies polled acheved savings around 50%
(cases where they replaced 400 bucks/hour resources for 50 bucks/hour
The realistic target saving when offshoring operations is estimated
around 30% as companies become more mature in the way they conduct
20 grand a year for a programmer in Brazil is a very good figure. I
don't know the current average salary of a programmer in the US or UK.
The 100K figure was based on supposition rather than fact.
>> > Your analogy implies that Indian software lacks quality. Quality seems
>> > to be measured (or guessed?) in CMM levels by every company in the US
>> > and the rest of the world is following this trend. Given that fact, I
>> > would like to remind you of the many companies in India appraised at a
>> > CMM 5 maturity level.
>> > I don't believe in CMM, but the guys in India didn't make it up, they
>> > just happily jumped on the SEI bandwagon.
>> I don't believe in CMM either and I dont believe it necessarily produces
>> a high quality product. It has been my experience (of three different
>> Indian outsourcing companies) that product quality is variable, and
>> a lot of the developers are what I would class as juniors even though they
>> are not sold as such. Undoubtedly some of this can be tracked to
>> the fact that development of apps that distant from the users is not the
>> best way of doing things. Also there is a disinclination for the Indian
>> guys to question the spec if they think it is inconsistent or lacking in
>> detail and to raise questions with whoever is leading the project.
>> Some (maybe most) of these issues are often also present for the big US/UK
>> consulting firms.
I worked with the 'Big Five" in some projects and they present a
uniform pattern in their staffing that I call "War Movie Team".
These teams are usually composed by a seasoned sargent (consultant)
which marvels the spectator (customer) with his ability and 10 rookies
(juniors or trainees). The sargent is usually assigned to another post
in the middle of the movie and must leave the rookies to their own
In unrealistic war movies, the rookies will use their talent combined
with the experience they acquired with Sarge to overcome the enemy and
come home as heroes. In real wars and real projects, they are
slaughtered by snipe fire, hidden mines and disease and return to
South Dakota in bodybags.