It's a rut that's easy for anyone who evaluates software to fall into: emphasizing how an application looks and operates, instead of how it helps people work together. Software UI design is the first thing we see, and the overwhelming number of SaaS sites and mobile device apps means no one has time to do a comprehensive evaluation of every collaboration tool that sounds interesting--snap judgments are a necessity in this age of information overload.

So, let's step back and remember the point of collaboration: to achieve common goals and complete meaningful work--meaning, work that's of higher quality, since it incorporates a greater diversity of thought and opinion (the wisdom of crowds), and that's done faster, with software assistance to speed the collection and assimilation of diverse information sources.

This distinction between form and function--which in reality is more a conceptual decoupling of form and interface from goal, process, and structure--crystalized during interviews with Ryan Nichols, Podio's VP of apps, about the nature of work and the collaborative tools we use to improve both process and results.

As Nichols expands on in a recent blog post, anyone tasked with helping employees collaborate needs to address two questions: What is the center of work? And how can software best bring diverse elements--tasks, people, documents, and data--together to facilitate the collaborative completion of that work?

His answer, perhaps unsurprising given his corporate title, is the app: a lightweight, easily built collection of structured data, process workflow, forms, and unstructured information designed to facilitate a particular work product--everything from evaluating job applicants to managing a product development project. While this collaboration model moves the metaphysical ball forward, I don't think it really captures the essence and nuance of collaboration.

If Nichols' model is an output- or results-oriented view of collaboration, self-styled collaboration strategist Michael Sampson proposes decoupling discussion of intent and scenario from any particular collaboration technology into what might be termed a "community-oriented classification" of collaboration. Sampson prefers to view collaboration products through the lens of the collective, the gamut of those working on a task, by breaking the software genre up into tools for teams ("a team working towards a deliverable"), groups ("a group that shares a common interest or practice"), and organizations ("creating [systemic or omnipresent] opportunities to collaborate").

The problem with this partitioning is that enterprise workers are simultaneously members of many such subpopulations, typically working on one or more projects (teams), doing tasks comprised of different disciplines (groups), and part of one or more organizational units (organizations). Who wants to use different collaboration software for each? No one. Heck, 87% of the 452 respondents to our 2012 Social Networking in the Enterprise Survey have these tools, but most only see small pockets of use among employees. Users just aren't showing up for even one system.

Furthermore, many collaboration needs, such as unstructured comments, information tagging and metadata, and user profiling, are common across all subgroups. Here's where what ECM expert Jed Cawthorne describes as collaborative management comes in, functioning as something of a meta-collaboration space, coordinating the various project- or process-specific tools, or apps, in Podio-speak. This indeed is what extensible platforms like Podio and Chatter seek to do, and what I suggested Facebook could become if it gets serious about enterprise collaboration.

Circling back to Nichols' original question about the nature of work and the best collaboration software model, I submit that Cawthorne, in proposing a five-pronged schema of collaboration tools--messaging, content, conversation, process, and management--captures the complexity and nuance of the issue. Any software that attempts to force users into a single paradigm, whether centered on documents (like SharePoint) or free-form comment streams (like Facebook), is bound to be suboptimal. Each of us is constantly juggling personas, objectives, and tasks as members of different projects, work groups, organizational units, and the overall enterprise. We're simultaneously working on longer-term one-time projects, short-term business process tasks, or routine clerical duties, each of which has different deliverables and work products.

No wonder people are rejecting the current crop of collaboration tools.

If you want the system you select to get used, make sure it's flexible enough to handle a multiplicity of work products, to add structure where needed to tame inherently unstructured comments and feedback; to provide a platform that facilitates unstructured comments and conversation on any work activity; and to act as a portal to unify our online collaborative experience. Podio's apps, Asana's task-centered approach, and Basecamp's project-centric design are steps in this direction, but hardly the last word. We expect to see an amazing amount of innovation and churn in the next-generation enterprise collaboration space over the coming years; it should be fun to watch.

Kurt Marko is an IT pro with broad experience, from chip design to IT systems.

Federal agencies must eliminate 800 data centers over the next five years. Find how they plan to do it in the new all-digital issue of InformationWeek Government. Download it now (registration required).