As Enterprise 2.0 continues to evolve, the one question that continues to go unanswered is how to measure ROI.

The question came up again today at an E2.0 “town hall” at the Enterprise 2.0 conference.

My response wasn’t popular, but I think it’s worth repeating:

If you can’t sell more, buy less, or fire somebody, you’re not getting real ROI.

In private, a number of folks complained to me that Enterprise 2.0 has been too ethereal in the past. Perhaps this is why there seems to be a grassroots movement to rebrand the topic as “Social Business.” Yet simply saying that something (a technology? a movement?) addresses real business problems doesn’t make it so.

Dig underneath the surface, and “Social Business” still seems to be about things like discovering expertise inside the company, or boosting employee engagement.

One argument that came up at the town hall (and was more popular than mine) was that deploying Enterprise 2.0 technology should be thought of as a cost of doing business, much like email or telephones.

Much as I wish this were a commonly accepted fact, I suspect that this is a) wishful thinking, and b) not likely to go over well with your 2011 Budget Committee.

I keep coming back to a simple principle articulated by the pragmatic philosopher William James: “A difference that makes no difference is no difference.”

If Enterprise 2.0 has a business impact, it will show up in business results, e.g. net income. The impact might not show up immediately, but if it’s there, it will show up eventually.

Enterprise 2.0 has been around long enough that we should be able to detect its impact, much like astronomers can detect the presence of extrasolar planets by observing their gravitational effects.

Nor are the necessary experiments and observations hard to find. We have numerous case studies of large enterprises that have adopted E2.0 products broadly across the enterprise. Alcatel-Lucent’s E2.0 platform reaches 45% of all employees. 80% of Booz employees logged into their internal social platform last quarter. The list goes on.

We can examine the financial results for these companies, and compare before and after. If the recent economic crisis makes such temporal comparisons difficult, we can compare E2.0 adopters with comparable companies that didn’t try the tools.

And that doesn’t even count the thousands and thousands of departmental pilots that could be used as experimental data.

This analysis won’t be easy. It’s easy for me to make my arguments, but you don’t see me rushing out to perform the painstaking research that would be required to reach a provable conclusion. That’s hard work!

But that hard work might be the very thing that helps make Enterprise 2.0 a mission-critical part of every enterprises’ IT portfolio.

P.S. @RonTeitelbaum pointed out on Twitter that I’m leaving out improvements in employee retention and improved efficiency. I disagree–if those improvements occur, they should have a bottom line impact. Improved retention should reduce recruiting and training costs. Improved efficiency should allow you to reduce staff, or generate more business by diverting resource savings to selling and serving new customers.