I recently had the opportunity to host a "town hall" debate at the excellent Enterprise 2.0 Conference in Boston. I asked a polemical question for each of 10 hot topics; I then invited the audience to select the topics that interested them most.

The audience zeroed in on issues concerning business case justification and tactical community management, as well as-somewhat to my surprise-questions about tool vendors. Here's how the debate went down on four key topics.

Can social software consistently bring real ROI? Sure, you can find some well-known case studies of documented financial ROI, mostly around reducing support costs via expert peer forums or increased sales for consumer goods and services.

However, the reality is that most social software adopters build first and (maybe) justify and measure later. Like many content-oriented technologies, there is no simple way to prove an ROI for, say, an effective intranet community. I think the most successful adopters recognize that social software-especially when deployed internally-is simply a cost of doing business, much the same way email has become essential and, before email, telephones moved from luxurious to ubiquitous.

You can still concoct dubious spreadsheets with mythical man-hour savings. But the better approach is, as one participant from a major consumer goods company said, "We support social computing because it aligns with our corporate values."

So which category of tools is best: platforms, suites, or best of breed? This is an age-old question in the software world. But you'll find some surprising answers in the social computing space. First, the question elicited some grousing among audience members who were customers of platform vendors (Oracle, IBM, Microsoft, Google)-dissatisfaction that mirrors what we've found in our own research.

In general, dedicated social software "suite" vendors came off best. But participants cautioned about the need for outside expertise around strategy and adoption in any case. In this space, as others, many vendors employ more consultants than product developers; you'll want to budget accordingly. Finally, some of the best case studies at the conference showed combinations of best-of-breed tools, including some innovative mashups.

When asked whether to socialize existing applications or invest in purpose-built social software, the audience made a good case for socializing existing applications. A core weakness of the current crop of social software vendors is that they focus on creating social spaces rather than socializing existing work processes.

Everyone agreed you want to try to catch people where they work. We can all criticize email, but one way to catch professionals is within Outlook and Notes-tools that can plug into those environments offer a leg up. Similarly, we can disparage the old-school software interfaces found in traditional financial and human resource management systems, but many people work in those tools all day.

In the end, we reached this conclusion: Enterprises need both socialized applications and dedicated social space. You can't depend on your enterprise resource planning vendor socializing its platform because big software vendors simply aren't innovating rapidly enough in that area. Perhaps more importantly, employees need "safe," usable places to network and collaborate outside the confines of individual line of business applications.

A fascinating discussion ensued around this question: Should community managers be responsible for information life cycle management? Many agreed with the notion that socially generated information should not be "managed" at all-at least not from the standpoint of setting retention schedules and getting rid of outdated stuff (the traditional purview of records managers). Instead, the community determines ongoing relevance and value, with nothing ever going away. After all, your search engine might find that certain nugget from years past that you still need today.

This may place too much faith in the discriminating power of contemporary search technology. Also, it doesn't deal with regulatory and legal problems resulting from storing information forever. We clearly need to rethink traditional records management, not the least because most records and compliance managers are overwhelmed dealing with traditional information repositories before they even turn to newer social software applications.

Unfortunately, intermediary retention steps-such as rule-based archiving-go missing in nearly all social tools that CMS Watch evaluates. So, better technology could surely help here, even if that's never the whole answer. There's a life cycle to all information, whether official enterprise content or emergent social discussions. The trick is applying the lessons of the former to the expanding world of the latter.