The New Choreography

At Wednesday’s New New Internet conference Google’s Rajen Sheth gave a great talk.  As part of it, he drew a distinction between application deployment within the enterprise and on the Web.

I think he was being kind to the enterprise model when he described it as managers and technologists trying to figure out what users within the company want, then trying to deliver it to them.  He got the three constituencies right, but soft-pedaled the real goal of most enterprise IT efforts.  

They’re not about the users, even though they’re often positioned and discussed that way. They’re about the enterprise —  what’s best for the organization as a whole, what will make it more productive, efficient, analyzable, etc.  

There’s no guarantee that what’s best for the enterprise will be best for all or most enterprise IT users.  In fact, there’s a near guarantee that some people and groups will be worse off after a new enterprise system (ERP, CRM, eProcurement, SCM, etc.) goes in.  They’ll get an application that’s by definition less tailored to their specific needs than the legacy stovepipe system that’s being replaced.  They’ll have to learn new screens, transactions, and processes, some of which are going to be less friendly or efficient than the previous ones.  They’ll have to go through lots of training that takes them away from their jobs.  And at the end of the day they’ll be encouraged (read ‘forced’) to use an enterprise application that gives them  fewer places to hide and less freedom.

I want to stress yet again that  this is not a bad or unintelligent thing for companies to do.  The work redesign, standardization, and monitoring capabilities provided by Enterprise IT are powerful and in many cases vital.  But acquiring them entails a particular arrangement among managers, technologists, and  users.  To put it bluntly, managers negotiate with each other about how the system will work, collaborate with technologists to configure it, then deploy this system regardless of whether or not users want it, or like it.

This sounds like a heartless process, and one that no self-respecting modern day business school professor could advocate, but I actually think it’s the right idea a lot of the time.  Companies are simply not going to get 100% consensus or up-front ‘buy-in’ from all  parties affected by a new enterprise system, so trying for it is a waste of time and effort, and can actually be counterproductive.  A 1994 study concluded that “… a culture of consensus decision making at the top of the organization can delay [business process] reengineering efforts and even make completing them impossible.”1  So a fairly heavy-handed, top-down adoption process driven by management is often the smart approach.

On the Web, of course, this approach is completely infeasible.  There are no managers, for one thing, and there are no mandatory sites or tools.  As people from Google are fond of saying, "the competition is always only one click away."  So in this environment technologists have had to operate very differently if they want their offerings to be adopted by users.  They’ve had to make them incredibly easy to understand and use, because our tolerance for unfriendly technologies and sites is very low.  Web technologists have also often started by introducing a tool that did only one thing, but they made sure that it was an important or valuable thing and that the tool did it very well.  Once users were both hooked and educated the tool could incorporate other features, as long as they were also easy to use.  Del.icio.us and Google are  two good examples of sites that have become gradually and seamlessly more complex.  And as I’ve written here before, the technologists of Web 2.0 have often tried hard to get out of users’ ways, to stop imposing structure on them and instead give them tools that let structure emerge.

So so far within enterprises we’ve seen managers and technologists imposing technologies on users, and on the Web we’ve seen technologists offering technologies to users, with only the best surviving and not a manager in sight.  

And now, with enterprise 2.0, we’re starting to see something new.  The easy, intuitive, and powerful Web 2.0 technologies — like blogs, wikis, tags, RSS feeds and aggregators that attract users and foster emergence — are starting to spread behind the firewall.  Sometimes this is happening by stealth and sometimes it’s happening with the explicit approval of line executives and CIOs, but what’s really interesting is that in every case it’s happening in an environment where there’s a third constituency —  managers —  in addition to the normal duo of technologists and users.  

How will this play out?  As I said in my talks at both the New New Internet and Interop, it’s simply too soon to tell.  The appropriate choreography among managers, technologists, and users for deploying big enterprise systems is pretty well understood (it’s hard, but the steps are clear).  The choreography between technologists and users on the Web has also been rehearsed enough times that it’s starting to fall into place.  But when managers are added to Web 2.0 tool deployment, transforming it into Enterprise 2.0, no one knows exactly what to do yet.

I’m pretty sure that proceeding as if managers aren’t there at all is a bad idea.  They can help encourage adoption, for one thing, and they’ll also likely be hostile if they get blindsided by these technologies, so they need to be in the loop.  But it also seems clear that it would be fruitless and silly for managers to try to deploy these technologies the same way they’re used to rolling out ERP.

The research I’ve done and the discussions I’ve had with Enterprise 2.0 pioneers has convinced me that line managers have a large role to play in deployment.  I’ve seen and heard too many examples of successful managerial intervention, and seen and heard too many instances of Enterprise 2.0 deployments that failed because they never got the ‘activation energy’ they needed, and so just faded away.  I don’t see how managers can effectively dictate blogging, tagging, wiki editing, etc., but I can easily see how they can encourage these kinds of activities within their teams, and provide good coaching.  In short, there’s definitely a role for managers in the new choreography of Enterprise 2.0 deployment; we know they’re not center stage all the time, but they’re definitely part of the ensemble.

Another highlight of the New New Internet conference for me was listening Michael Arrington of TechCrunch.  He made some fascinating points about how much cheaper, easier, and faster it is to get an industrial strength site up and running today, compared to only a couple years ago.  He said that TechCrunch will soon have an enterprise section (or separate site).  I’m planning to read it faithfully.

And thanks to Dion Hinchcliffe for putting together a great conference.


1Bashein, B. J., M. L. Markus and P. Riley 1994. Preconditions for bpr success and how to prevent failures. Information Systems Management 11(2): 7-13.