So Much For Taxonomies

Well, in the last post I was optimistic that we had finally cracked the code here.

We were wrong.

And, in the process of watching what happened, we learned a few things that are probably worth sharing here.

The Background

There's a certain element that wants a nice, neat, orderly taxonomy for all of our communities and discussions, even though they're inherently chaotic and disordered beasts.

We thought that we had come up with a more attractive approach, centered around "verbs" rather than "nouns".  A few members of our team had organized all the existing discussions into a handful of verbs that were intuitively more appealing than other approaches.

So we opened it up to our extended community, and the fun began.

A Passionate Discussion

Remember, by now, we've got lots of proficient users that think of this as "their" platform.

The commentary was heated.  I had to turn off email notifications on "... another comment was posted" since I couldn't keep up -- at least 75 comments in under 24 hours, many of them long, or passionate, or both.

Some people actually liked the chaos the way it was.  They knew that there was no taxonomy, and felt very comfortable poking around, searching, following tags, etc.  Why change things?

More interesting were the community owners who felt a sense of ownership and pride in "their" communities (a very good thing, I'd offer) who had a lot to say about how they were being categorized, who else was in the category, how the various categories appear, and so on.

Now, if you think about it, none of us especially like being externally categorized by others, do we?  The same thing applies to communities and community owners, it turns out.

Even more interesting was the dynamic from the centralized team.  Our uber-community manager (Jamie) found herself in a defensive position, patiently arguing her points, over and over again, and not making much headway into the comment blizzard.

It almost felt like -- forget what the community wants, here's what we're going to do!  Not a good dynamic in a social media environment, I'd offer.  Hints of command and control were starting to be evident.

I"m arguing for a retreat from this topic -- I hope people agree with me.

Time For A Different Perspective

Usually when I get into situations like these, I usually take a step back and ask the question "what problem were we trying to solve here, anyway?"

This turns out to have three components.

The first component is the newbie who's presented with an alphabetical list of communities, and has no idea where to start, or get engaged.  The thinking is that -- by having a clean, orderly taxonomy -- the platform will be more engaging, and we want to drive user adoption, don't we?

The second is proliferation -- we're adding new communities at a steady rate, and it's not hard to imagine 200, or 500, or maybe 1000+.  If you've looked at Clearspace's "home page", you can imagine the list of spaces scrolling down into infinity .... yuch.

The third is human nature.  There are certain people who just don't like disorder in their lives.  It's not right unless it's neat, tidy, etc.  And, when those people are the ones who are running the platform, they tend to impose their views on the community at large.  I'm not blaming anyone, it's just human nature at work.

Solving The Newbie Problem

If you think about it, there are lots of ways of orienting newbies, right?  It's pretty clear from the feedback that experienced users don't have any issues navigating, so it's only the newbies we need to worry about.

Clearspace 2.0 has some nice features for "attention management" from the home page, including "most popular discussions", or perhaps "new communites forming" or other mechanisms.

We could also offer up a newsletter (blog!) about "What's Happening On EMC|ONE This Week" that could highlight interesting discussions that were going on, and target these to our newer users.

Or, perhaps an expanded directory (wiki?) with descriptions about each community, and focus this on the new people.

I don't know what the right answer is, but rewiring our entire top-level taxonomy in face of disapproval by our community probably isn't the optimal answer here.

Solving The Proliferation Problem

I'm noticing that communities tend to co-evolve in a very natural way.  At the outset, we had perhaps 5 or 8 communities start up around different aspects of what EMC does with Microsoft.

All sorts of agita around confusion, etc.  I strongly suggested to let'em do whatever they want.  And, now, several months later, it's pretty obvious to everyone that you could create a space called "Microsoft" and put all of 'em in that same space (still with their own subspaces), and nobody would argue much.

This sort of tactical, bottoms-up consolidation could take the number of top-level communities way down, but it'd have to be a natural process, and not forced.

We also have the opportunity (I think!) in Clearspace 2.0 to have the subspaces appear on another page entirely, which give us all sorts of expanded presentation options to help newbies navigate.

Solving The "It Looks So Messy" Problem

This is actually another problem entirely -- when control of a platform rests in a few hands, the natural tendency is to run the platform around what works for the owners, and less perhaps around the needs of the community.

This is a slippery slope that I don't want to start down.

As an example, if your users are comfortable with the clutter, who are you to say how it's supposed to be organized?

And I think this sort of power struggle will become more prevelant -- and more important to keep an eye on -- as we move forward.

Comments

[Steven]

Have you considered organizing communities into role-based, project-based, department-based, and business process/activity-based groupings?

[Chuck Hollis]

We did -- briefly -- but discarded it. The high value seems to be coming from interactions that span traditional definitions.

Your model works well for document-oriented collaboration (we use a varient of it), but -- conversations -- well, that seems to be different.

[Oliver Marks]

Oh boy, I've been down this path before...I spent what seemed like an age doing content taxonomy planning with the American Management Association.

Like catching a greased pig, it works for one group of people but not for others.

While doing horticultural pruning of Sony PlayStation's wild west of freestyle created, 'guideline free' mass of legacy content I opted for the 'divide into big buckets, then identify smaller areas within them' approach.

Not exactly a tree model since so many things are inter related and associated with multiple applications but a start.

So much of organizing groupings is dependent on the end users needs. The PlayStation users are divided into technical and non technical - techies like lots of macros and tolerate ugly ui's while non techs need a well thought out UI and group/content logic before they will dip a toe in the water and then get in and swim if they like it.

An 'index' page (as in last pages of a book) approach as well as search field seems to work well with people also...

Great post Chuck!