We did a "soft launch" and let it grow organically by word of mouth before doing a formal announcement. Our fear was that doing a big splash with no content or activity already going on would cause a lot of people to poke their heads in briefly and then go back to work as usual, with nothing to keep them coming back. So we focused on a couple of key groups we wanted to help, but let anyone who wanted get into it and create a group. We launched in April 2012 and by August we had enough happening that we felt comfortable having our CEO mention it in his quarterly employee meeting (he even did a live demo!) along with posts and stories by our Corp Comms team.
I think that when it comes to internal deployment, a marketing effort is a good idea if it is used as a boost to an already organically growing deployment, but might backfire if done too soon. When we deployed I found some graphs on the internet that helped me argue for this kind of deployment strategy. If you look at this chart:
It looks like Google+ was way more successful than Facebook or Twitter. But we all know how that turned out. Facebook and Twitter both grew organically from relative obscurity in small start-ups, while Google+ started with a big publicity splash and "invitation only" access that everyone wanted just because it was Google, regardless of whether they needed (or even understood) what G+ could deliver. So the growth charts ended up looking like this:
And Google really should have known better because they had taken the exact same deployment approach earlier with Google Wave and got the exact same result:
I wrote up some of the thinking behind the approach we took in a series of 3 blog posts if you are interested. (The titles were not my idea -- InformationWeek wanted to do that numerical clickbait kind of title, which I really hate. In my mind these are three parts of a single topic.):