- The facets of collaboration Part 1–Meet robot barbie
- The facets of collaboration Part 2–Enter the matrix!
- The facets of collaboration Part 3-The feature jigsaw
- The facets of collaboration Part 4 – BPM vs. HPM
- The facets of collaboration Part 5 – It’s all Gen-Y’s fault – or is it?
I have a friend, lets call her Jane (not her real name), who was a huge web2.0 fan. Seriously, if it was a wiki, blog, tweet or anything remotely sounding like RSS, Jane would wax lyrical about how it was the answer to all that was wrong with the silos of the old world and if only people would get with this new paradigm and embrace the social revolution, collaboration within her organisation would markedly improve. After all, look at the popularity of sites like facebook, wikipedia and twitter. Jane also had a very strong vision for what this new world would look like and had spent a lot of time and money investing in customising SharePoint to meet this vision.
Much to Jane’s dismay, her project failed miserably. Her organisation ultimately wanted something much more boring – a solution to help them manage their files better. All of this “new-age gen Y social crap “ was of little interest to the masses.
Janes story is actually a very common pattern with SharePoint projects. As it happened, Jane had committed probably the most cardinal sin of information architecture. She had projected her ideals onto an organisation that did not necessarily subscribe to her view. Therefore SharePoint represented her vision and little else. Few others shared it and to this day I think she blames the culture organisation she worked for. I say that she did not do enough to develop a shared understanding of the problem, and shared envisioning of the solution.
Now most of us know that SharePoint can be a platform for going gung-ho social if you wish, given that it has features like wikis, blogs, folksonomy and RSS. But it can can also be a platform for structured business process via features like workflow, BCS, information management policies, content approvals and the like.
So that raises a really important question. Why is it that a particular collaboration feature (such as a wiki) might be total nirvana for one situation and can be a project killing fatal flaw in another? Since SharePoint has a ton of collaborative tools in its toolkit that can be rolled up in different ways, when do certain features work well together and when do they completely suck balls together? Why do some people have such wildly differing views on the nature of collaboration and what potential solutions should look like? Why do some people look at the SharePoint feature-set and say “meh”, while others are totally charmed by it?
This was a question worth pondering and I did look into this in some detail while I was creating my SharePoint 2010 Information Architecture course. In this short series of posts, I am going to introduce you to a way of looking at the wide spectrum of activities that all fall under the guise of this term called collaboration. This mental model seems to help paint a more realistic picture of the collaborative world than the simplistic views that Jane and her ilk use. It goes some way to many of the questions that I raised above. Furthermore, the model has been very well received at my courses. Maybe there is something to this?
Introducing “robot barbie” – the yardstick for SharePoint information architecture
Say hello to Robot Barbie, my all-time favourite SharePoint metaphor. I use this picture in all of my classes and talks, such is its persuasive power.
I first saw this picture via a talk that Joel Oleson gave. The story behind it was that Joel came across this very real toy in a market somewhere in Asia. When Joel asked the storekeeper what the idea was behind the toy, the reply he got was along the lines of:
“Well, boys like robots and girls like Barbie. Therefore, if we put Barbie’s head on a robot, then logically both boys and girls will like it.”
Now I don’t know about you, but I can’t see that boys and girls would suddenly drop their respective robots and Barbies and start playing with this ugly hybrid. In fact, when you look at the result, it is this bizarre, somewhat disturbing combination of features that brilliantly demonstrates the folly of taking two things that work really well alone and expecting that by combining them, things will work even better. Instead, we have a combination that is much less than the sum of its parts and unlikely to satisfy anybody.
Robot Barbie is a very powerful visual metaphor for SharePoint information architecture because SharePoint is full of Barbies and full of robots. In fact the global SharePoint information architecture mantra should be “avoid robot barbie”. To that end, when I show this image to clients or conference attendees, I ask them the question, “So is your SharePoint implementation a robot-barbie solution?”.
Many people will admit to varying degrees of robot-barbie. How then to avoid it?
Research into collaboration itself
The only SharePoint person I have met who goes to a university library to research more than me is Erica Toelle. In this section I am attempting to make her proud – hehe 🙂
Since these questions around collaboration came to me while I was developing an Information Architecture course, I became curious as to whether academics, authors or bloggers had ever actually attempted to deconstruct collaboration itself into its core elements (in effect, performing an information architecture exercise on collaboration itself). Surely understanding the some of these elements would help us gain some hints or insights into how collaboration works and in turn, help us avoid Robot Barbie?
So I hit the journals and did some digging. As it happened, some academic research had been undertaken over a long period of time, trying to codify this lofty ideal called collaboration. As expected, authors looked at the issue from various angles. Some looked at it through the lens of the type of conversation taking place, some via the nature of the problem solved, some through “business activities”, some via the characteristics of group doing the collaborating and others examining the underlying purpose that drove the collaboration in the first place.
Some of the more interesting writing that I examined in developing my own model included:
- A Taxonomy of collaboration in support of information seeking (Golovchinsky, Pickens, & Back, 2009)
- Three domains of problems (Talley, 2003)
- Know the Four Types of Conversations (Banathy & Jenlink, 2004)
- The Art of Focused Conversation: 100 Ways to Access Group Wisdom in the Workplace (ICA series) <- A great book I already owned
- Understanding Groups’ Properties as a Means of Improving Collaborative Search Systems (Ringel & Teevan, 2009)
- Beyond Document Collaboration (Hollis, 2007)
I read all of these papers in detail and then converted the arguments made by each into IBIS and created a single (admittedly very large) issue map to try and draw out any patterns from them. I would group all of the various concepts in the map together in different ways to try and elicit some sort of insight. This proved to be quite a frustrating process because of the various different ways each writer tackled the subject matter. The map was huge and I was at first, unable to find a pattern that worked for me. All of them seemed to have some interesting elements, but they all seemed incomplete or over-thought in some way.
In frustration, I gave up and slept on it. During the night my subconscious must have kept working to unscramble the map because the next morning, I had my model within a few minutes. I looked at the map again and saw a pattern immediately. It was in essence, a combination of some work by Chuck Hollis and interestingly, Microsoft Research (The last two of the above references).
Meredith Ringel Morris and Jaime Teevan of Microsoft research wrote a brilliant paper called “Understanding Groups’ Properties as a Means of Improving Collaborative Search Systems”. For the record, anyone that is looking to leverage user profile data via mysites should read it. In this paper, Morris and Teevan divide collaborative groups into trait or task, based around the longevity of a group. They also use group identification in the form of group membership being implicit or explicit. For my own purposes, the implicit vs. explicit membership of a group was not overly relevant so I discarded it. But I found the notion of trait and task based groups fascinating.
Morris and Teevan characterised task based groups as short term and comprised of people with a shared goal and working together to achieve that goal. (They are describing every project team right there). Trait based groups on the other hand were comprised of users who were related through shared traits of long term interests. Here was the quote that really made me sit up and take notice however. They also suggested that trait based “group members may not be consciously collaborating on the same task, but may be highly likely to repeat or augment tasks already accomplished by other group members” (Love it – they are describing every discussion forum right there).
So I had my first dimension of collaboration. Task vs trait.
In 2007 Chuck Hollis of EMC wrote a really insightful blog post entitled “Beyond Document Collaboration”, where recognised that via the advent of social media, organisational collaboration was changing and that “different collaboration models are emerging, each one with entirely different value propositions”. Hollis identified three such models and labelled them transactional collaboration, document collaboration and social collaboration.
Hollis described transactional collaboration as “workflow, or business process management, or something else” and was characterised by “humans are largely automatons; repetitively processing the output of one function for input by another function. Not much spontaneous creativity interaction here, nor is it usually encouraged!”. Hollis then spoke of document collaboration and explicitly mentioned both Documentum and SharePoint. He finished with social collaboration, which he described as “It’s not predefined interaction. It’s not a structured workflow. It’s something entirely different than the other two collaboration models”.
After reading that post, I thought Hollis was definitely on to something but I felt his descriptions didn’t quite encompass what I was looking for. Like the Ringel/Teevan paper, I felt that elements of Hollis’s breakdown didn’t quite fit for me. After I had slept on it, it dawned on me that document collaboration was the odd one out. I liked his distinction between transactional and social collaboration, of structure and predictability versus a non predefined interaction because the social dimension to me was describing knowledge work. But document collaboration did not work for me at all. A document is simply a medium, as is a wiki or a forum. Structured, process driven collaboration can still use documents, and so can relatively unstructured social collaboration. They just use them in different ways. The same can also be said for whether the collaboration is task based (ie outcome driven) or trait based (interest driven).
So I discarded document based collaboration and in doing so had my second dimension.
Making a model
Out of all of the material that I researched, I found that these four dimensions or facets of collaboration (task, trait, transactional, social) helped me explain most collaborative scenarios and understand why Robot Barbie solutions occur. Of course, the great thing about distilling it down to four dimensions is that I then I got to do what academics live for. Make a 2*2 matrix!
Academics simply love doing this, because it helps them deal with over-analysing everything that moves and justify all that R&D that they do! :-). In all seriousness though, people love creating 2*2 or 3*3 matrixes to explain concepts for the simple reason that they make good mental models to help us understand reality. After all, one of the key purposes of information architecture itself is to help users create mental models of a site (“Don’t Make Me Think ”).
So here is my complete model. In part 2 I will explain how I use this model and the insights that it gives in preventing robot-barbie outcomes.
Thanks for reading