Can Web 2.0 be adapted to the enterprise?

Last week I had the distinct honor of listening to Harvard Business School's Andrew McAfee speak on Enterprise 2.0 at The New New Internet in Northern Virginia.  I've written several times here about McAfee's thoughts and work about the use of freeform, emergent, social software to enable ad hoc collaboration.  McAfee's work essentially applies and combines many of the key Web 2.0 concepts of social software, user generated content, and discoverability via search to the workplace. 

And given the efficacy of these things out on the Web, a number of companies have begun to seriously consider his ideas.  Some of them have even begun building products labelled as "Enterprise 2.0" solutions and we may be witnessing this as part of a larger trend of moving the consumer world of Web 2.0 into the enterprise.  We'll hopefully do a tour of some of these Enterprise 2.0 products in upcoming posts.

But getting back to the McAfee's talk, his keynote was the closing speech at the end of the event and turned out to be one of the highlights of the day.  McAfee made numerous interesting points but key among them was the the idea that modern Web 2.0 software (at least the succesful ones) introduce ideas that deal with classic user adoption challenges that enterprise software tends not to directly address.  In particular, that despite what they're given to use, users "vote with their feet"  and prefer software tools that are simplest, most effective, and most familiar.

Key to this discussion is that unlike the Web, users tend to have very few software options in the enterprise and are usually prescribed the tools to use to get their work done.  And when faced with a dizzying array of features and capabilities in their shiny new, sophisticated enterprise IT systems, they tend to default to the tools that are easiest for them, over which they have the most control, and are most familiar with.  These "comfort apps" are things like plain old e-mail and Microsoft Office.

As I listened to McAfee bring up stories and anecdotes that supported these points, such as the Pedia project -- a precursor to today's Wikipedia -- which found that the radically simpler the submission process was, the more contributions were elicted by quite a large margin.  Processing this, I was struck again by the similarity of this to the whole emergent vs. deliberate debates that we frequently find in the software and IT worlds.

The fact is, most procedures in the often bureaucratic world of business are seeking specific predefined outcomes that are often driven to completion without measurement for effect or optimization.  For example, software must have the features -- and only the features -- laid out in the original requirements list; processes must have the same repeatable, measurable results each and every time; and budgeted activities should generate only what was intended from the outset, with little expectation that significant unexpected results will occur and need to be managed, much less exploited.

What we need are examples of high-value emergence use of social software

Yet it seems clearer these days that highly general purpose software like simple e-mail, blogs, wikis, and other social software can enable and form the foundation of almost uncounted open ended and adaptable collaboration scenarios.  I often get asked for example of this happening in the enterprise, and while negative examples often come to mind first (pyramid schemes via e-mail, creative phishing attempts, and so on), some of the real emergent Enterprise 2.0-like scenarios I can cite are really back out on the Web.

My favorite real-world example of effective, useful, truly emergent collaboration is last year's PeopleFinder project (aka KatrinaList)  and shows the real potential that can be tapped if -- and only if -- you have seeded your organization with these tools and found a way to make them the easiest, default choice for the workers in an organization.  The PeopleFinder story is a fascinating example of the emergent use of the blogosphere and the story, which I often use in my Web 2.0 talks, goes like this:

In the immediate aftermath of the Hurricane Katrina disaster, tens of thousands of people fled the New Orleans area, surfaced in neighboring towns and cities, and began to report their location via e-mail and, you guessed it, their blogs.  Followers of the blogosphere looking for Hurricane Katrina news from the scene queried the usual feed search engines for posts about the disaster.  Along the way, they repeatedly located posts from survivors and the details of their situation.  Immediately, a pattern emerged of people indicating they were alive, safe, and where they were located.  Some of these volunteers soon perceived the importance of this information and got to work scraping these reports into an online database from the machine readable feeds of the blogosphere.  Within three days -- at virtually no cost -- and with a rich outcome essentially impossible without the widespread use of freeform, social tools, approximately 50,000 survivor reports were collected, centralized, and made publicly searchable so that friends and family could locate loved ones.  Emergent software helped save the day.

This is potentially a very compelling story of what can happen if you build a social software ecosystem in your organization and put some real elbow grease into hammering down the barriers to adoption.  But getting back to McAfee's talk, he made another important point about emulating the successes of the Web within the enterprise, and it was this: There are perhaps 5,000 active contributors to Wikipedia in all its forms worldwide (disclaimer: numbers on this vary), yet it's in 200 languages and contains millions and millions of entries.  This level of scale evidently seems to work quite well on the Web.  But McAfee pointed out that if you scale this back down to the size of the enterprise, it means you may have virtually no one contributing content.  The point: The scale of the audience on the Web may be an important success factor.

Does all this mean we're challenged for techniques to solicit contributions and emergent output from our workers?  Very likely; getting workers to move much of their work from private, undiscoverable tools to social, public, freeform tools will undoubtedly be a challenge; and ironically a challenge that most organizations won't be likely to take on without knowing the outcomes.

Few feedback loops and limited software choices creates poor outcomes in the enterprise?

Several things are clear from this, and I tried to make this especially clear in my diagram above.  One is that traditional enterprise software development processes have probably continued to impose too much control and predefinition in the two major stages of the sofware lifecycle, the first being the creation and/or procurement of software, and second in its actual adoption and use in the organization.  Like agile software development -- and to a greater extent Web 2.0 software development techniques in general -- has shown us that lack of active, end-user driven feedback loops results in software that just doesn't do what is needed and doesn't evolve gracefully when the local landscape inevitably does.

The second thing that seems apparent is that there is little to no software competition within the enterprise.  This is very different from the Web where users will flock to the easiest, most effective tools given good options that number in the dozens -- and often hundreds -- of potential, easy-to-switch-to solutions (think about why you use a given Web client for e-mail.)  Enterprises that don't learn from the co-evolution process of development that many successful Web 2.0 sites are using to "find" the right set of features and capabilties are likely acquiring software that few users will find useful or effective.  This may be the biggest legacy that McAfee's work gives us; that emergent, social software results in:

1) Richer, reusable information ecologies in our organizations ala the Web, and

2) A workable framework for co-developed situational software based on social software platforms that can readily adapt to what users need at a given point in time.

Please listen to the podcast of McAfee speech at The New New Internet yourself, and settle in for a front-row vantage point on some of the latest thinking in enterprise IT.