Friday, November 01, 2013

The Unanswered Offshore Question, Pt 2

So, how do you know if your offshore team is as productive on onshore, and therefore saving money (as is typically the primary reason for offshoring)?

Work from one backlog (combine backlogs if working from more than one). Without specifying which team would do which story, have both onshore and offshore teams participate in sizing, or estimating, from one larger pool of user stories. You may do it together, or start by doing a few together and the rest separately. In the end, sample enough of the stories that both teams agree on how many story points various stories are.

At or before the planning meeting(s), randomize the stories or alternate each team getting whatever story is the next one pulled. At the end of the planning meeting, you'll have an estimated amount of work for each team, which gives you an idea. At the end of the sprint, you'll know for sure. Even with various issues (no product owner, newer/smaller/etc team), you'll have a number.

If you can't have both teams estimate and pull work without upfront planning or designating due to resource, knowledge, or other silos or barriers, I think you have other problems that Scrum will fix (if you want it and let it).

Thanks for all the great feedback from the first offshore post (and all the spam comments for offshoring ;-)

1 comment:

Kirti K said...

In scenarios where I have multiple dev teams with no historical data on their performance or strengths, I would typically let the teams pick equal velocity story and run. The idea being to fail and fail fast so as to capture each teams capabilities. At that point the team members can be shuffled around to match/hedge strengths/weaknesses ensure repeatable and statistically predictable outcome.