My personal opinion and experience is that there is no simple answer to how to implement Enterprise 2.0 practices and solutions. It all depends on what you want to do, why, where, when and with whom. An implementation approach inevitably needs to be tailored for each specific business context (type and size of enterprise, technological and cultural maturity, objectives, business use cases, scope, type of technologies...). Sometimes a pilot is good, sometimes it has no or very limited value. What we can do is to identify implementation patterns that work - similarities between approaches that work in typical business contexts. Some of these patterns will include pilots, some won't. These patterns can be used when designing the implementation approach for a specific enterprise. But, as I have argued before, I don't believe there is such a thing as a best practice when it comes to such things as implementing Enterprise 2.0 practices and solutions, even though there are good or even great practices (or patterns) that enterprises can learn from.

Michael Idinopulos wrote a similar piece (Enterprise 2.0: Skip the Pilot http://ow.ly/1C6fP) last year and there are some really good comments to that post as well.

Pilots yer or no? It depends! I mostly agree with Oscar but for a very specific reason. In his post Andrew says:

"The more I learn about and think about the value of emergent social software platforms, the more I suspect that the deep meta-benefit they provide is technology-enabled serendipity, defined as ‘good luck in making unexpected and fortunate discoveries.’ Serendipity is possible when we’re collaborating with our close colleagues on a well-defined project, but that’s probably when it occurs least often. It’s much more likely during wide forays and broad searches, the kind that are so easy to do with current technologies."

If Enterprise 2.0 was only serendipity, Andrew’s point would have been absolutely on the target and any pilots would not have been representative of critical mass needed to trigger and prove the exponential dynamics of interaction evident in online community . But, in my view, Enterprise 2.0 is not exclusively serendipity.

I've developed an Enterprise 2.0 Framework (http://www.socialenterprise.it/?p=105) starting from McAfee's bull's eye model that clearly shows how other than the strength of ties between participants, what makes a difference are the goals and needs a project is launched for: collaboration, connection, communication & sharing, collective intelligence.

If collective intelligence or emergent non-codified expertise are certainly two of the most original contributions introduced by Enterprise 2.0, allowing people to connect, stay in touch, collaborate asynchronously, leaving traces of partial outcomes and maximizing the reusability of these assets over time, should not be ignored at all. In my experience these are among the most frequent scenarios in small and medium enterprises (most of the Italian and European markets).

You can read more about these considerations and other criteria to properly choose Enterprise 2.0 pilots on my blog at http://www.socialenterprise.it/en/index.php/2010/04/23/enterprise-2-0-pilots-yes-or-no-it-depends/