As described in How to do 11 Years of Work in Under 10 Days, the Google Migration project kicked-off in January 2011 and was completed four months later, in May. The foundation for this success was a terrific team, and many good choices made along the way. Following are three key decisions that I believe made it possible for us to pull off such a large project in such a short timeframe.
1. Adopt a "lift and shift" philosophy
Whenever a large project like the Google Migration is underway, it's natural for enthusiastic staff to want to use it as an opportunity to add additional features or revamp existing peripheral processes (also called "gold-plating"). We knew from the outset that these sorts of suggestions could distract us from our ultimate goal and threaten the project deadline. A "lift and shift" philosophy, in which applications are moved to a new platform as-is, was adopted from the beginning. Whenever someone would begin a sentence with "wouldn't it be nice if..." the immediate response was "lift and shift!"
And we stuck to that decision. As the project proceeded, if a stakeholder requested a project change that did not fit our carefully defined scope, the team was empowered to deny the request. Had we pursued the tweaks and adjustments that were proposed along the way, we would have risked the success of the project.
2. Set clear goals and priorities
With the strong support of Project Director, David Ackerman, once the preliminary planning for the project in December 2010 was complete, we did not waver from that outline. This was largely possible through the aforementioned empowerment to avoid adding "extras" that could potentially derail us from the core goals. In my experience as a project manager, projects of this scale often change direction, so I was thrilled to be a part of a team that always knew where it stood. In addition, team members were aware that the Google Migration project was a top priority for ITS and the university at large, so they were able make time in their busy schedules to help the project succeed.
3. Conduct multiple pilot tests
Making time for not just one, but two pilot tests may seem to contradict the goal of avoiding "extras" but was in fact a key decision in the project's success. When it was determined that about 25% of the participants in the first pilot test called the IT Service Desk for support, the decision was made to run a second pilot, with an ambitious goal of a 3% call-in rate.
The second pilot gave us the opportunity to include approximately 1,000 staff and faculty from NYU Steinhardt and three NYU global sites, rather than the primarily ITS staff included in the first pilot test. This let us further refine our plans for communication, training, and support, ultimately achieving our goal of a dramatically improved 3% call-in rate. The success of the approach we used for the second pilot led us to ultimately use it for the final University-wide migration, which took place only three weeks later.
These three decisions, supported by a strong foundation of a dedicated team, and proper planning, communications, and project management, were the key to this project's success. See the NYU Google Apps website for more information about the full suite of services.
ABOUT THE AUTHOR
Monica McSharry is an NYU consultant and acted as the Project Manager for the NYU Google Migration project.