NDI’s Move to Google Apps

Last weekend NDI migrated off of the on-premise Exchange infrastructure we've used for almost a decade to Google Apps. It has been an interesting process leading up what we coined "Switch Weekend" - May 28-30 – Memorial Day holiday weekend in the United States.

As I write this post on Friday evening from 35,000 feet our 1200 staff in 70 countries need to start opening their browsers and navigating to our Gmail sign-in screen instead of double-clicking that oh-so-familiar desktop icon to access their email, the lifeblood of our communications and what I've always considered the "third rail" of NDI IT services. From an IT perspective, you can try a lot of innovative things at NDI but don't mess with people's email - at least that is how I've always viewed it. For that reason we didn't undertake this initiative lightly.

I'm not sure how it'll turn out, but I think we've done a lot of things right as we prepared the organization for the transition - strategically, technically and organizationally. We learned a few lessons along the way as well. I've got a great IT team here and they did an amazing job pulling together all the different components of this process - new territory for all involved.

There are several different pieces, the decision-making process, working through the various technology issues, bringing the organization along with strategic communications from the top all the way to staff around the world and starting with the tech team itself, and then preparing everyone to use the new set of tools and actually make the switch when their email stops landing in their current Inbox.

In many ways the decision was easiest - certainly in retrospect. I've come to the conclusion as we moved through the process that moving email services to the cloud is a no-brainer. No organization like NDI has any business running mail on-premise anymore. It's irrational in about every way you look at it - the high cost of managing the complex infrastructure internally being the primary one (we estimate that we get about 1.5 system administrators back to work on other projects by this move -and that is almost half our technical staff!). Better security, performance and availability being the others. The real question for us was whether to take the easy road and push out Exchange to the Microsoft cloud, or make the switch to Google?

Every instinct I had said we should move to Google because of the wide adoption of Google tools by NDI staff around the world, and by contrast and equally important - the lack of adoption of the majority of features of our existing Exchange platform. Go to where your users are, a cardinal rule in online strategy that we took to heart. There is a reason for many of our staff going rogue and using Google tools instead of Outlook and Exchange - they are a great set of tools for an organization with a global footprint like ours (as long as you've got bandwidth to use them). Google also promised a much better email experience for our staff given connectivity challenges of accessing our DC-based services as we felt they had the best global network infrastructure to accommodate our staff in some of the far corners of the globe – and did I mention that it is considerably less expensive given the non-profit pricing that we qualify for as a 501c(3) (free with caveats, for now)? True.

The biggest risks of moving to the Google cloud for NDI derives from Google’s strategic approach. While I think Google's vision to run everything from a browser sitting on a broadband connection is the best approach, Microsoft's strategy of linking heavy desktop apps to the cloud has merit for NDI because we have offices in a number of places where bandwidth is still a challenge and staff who spend a lot of time traveling and in transit - often offline but wanting to work. Storing mail and documents locally and syncing to the cloud when you're back online is still necessary for us in a lot of situations.

Working offline, as I'm doing now on a plane and many of our staff do for any number of reasons, would pose a considerable challenge if our staff rely on browser and cloud applications. We knew we could always fall back on Outlook running on a desktop with the excellent Google sync tool so this was fairly safe. We’re also told that Google’s HTML5 implementation will provide better offline functionality.

The other risks were around staff reactions to change - in any organization some like it and some don't. Introducing significant organizational change is a bit of a roll of the dice since you never can really predict reactions, outcomes or consequences. Working the internal politics and good communication are key factors here - not to mention a smooth technical execution and good luck. We had a fairly comprehensive 8-week communications plan that tried to address not only the possibilities of the new Google tools, but also the anxiety that many were feeling about making this kind of change.

We had plenty of time to sort out the technology challenges during a several week pilot program and gradual data migration process which made the actual switch weekend process manageable. In fact, our 3-day process that envisioned working through Monday went so smooth that our engineers were done with the migration by Saturday evening. Worked out a few kinks Sunday and Monday as our field offices in the Middle East and the rest of the world returned to work, and we’re live globally today.

Time will tell how this transition plays out during the next couple weeks. So far it seems to be going well. I’m very anxious to see where we are in a few days, a few weeks and most exciting will be where we are this time next year. My next post will outline why I think going Google will be a game changer for the Institute in the coming months and years.