The wild west of work media: A deluge of streamed, unstructured data

In the recent past, we’ve seen an explosion of innovation in the enterprise software marketplace. Perhaps I should soften that to the “business software marketplace,” since many of the innovators involved have opted for a consumer-style model of adoption. Instead of marketing to corporate IT staff, these new products are being marketed like Twitter or Foursquare.
Part of the innovation in this new generation of products is that they are — largely — built on a Software-as-a-Service (SaaS) foundation, and getting up and running can be as fast as buying a book on Amazon. And because these applications are social at their core, they can be very viral. One member of a small company’s marketing team decides to manage a project using Yammer or Podio, and she immediately invites four other team members to get involved. This means that the tools are quickly adopted, at least in the small group level. (Note: I’ll talk with Soonr’s Martin Frid-Nielsen, Mavenlink’s Ray Grainger, and QuickOffice’s Alan Masarek about this topic onstage at Net:Work, on Dec. 8.)

The new “work media”

However, the more important side of this social tool innovation is that they are based on activity streams. Users’ activities within these applications are not simply captured in the metadata of directories or the states of information in databases: these activities — such as making a comment, adding a file to a project, or assigning a task to a project member — are published in streams, a la Twitter, Facebook, and a long list of other consumer applications.
To distinguish these modern business tools from earlier generations, I use the term “work media.” They share characteristics with well-known social media tools, but they are oriented toward getting work done. And like social media, work media is fluid, with streams of information finding their way to the individual user, who may opt to follow topics, projects, and other users. These applications share the core design metaphor of streams, though they differ widely in how streams are composed, displayed, and shared.

Open tools, closed data

Business has a bias toward privacy, and so work media tools support that tendency by — almost without exception — erecting a password barrier against access to a company or group’s information. Finer-grained access controls are provided for more specific information contexts, such as projects, folders, groups, or other “spaces” managed by the work media apps. In this way, a company can restrict access to a HR group so that only a few HR staff can see the resumes and pay information managed there, for example.
This tendency, along with the relative immaturity of the burgeoning work media marketplace, is rapidly leading us to a very strange outcome: a generation of business software — work media — ostensibly based on the principles of open social media, but which are inherently closed, and which are spawning a million information silos.
But the risks and costs associated with business information stored in these applications is much higher, at least form the view of the companies migrating onto these work platforms. So once a company commits to using a specific work media platform, they may find that the information stored in their projects becomes as fixed as concrete.

Streamed, not structured, data

Let’s lump the information managed in these systems into two piles:

  • Concrete, structured, and relatively moveable information, stored in files of various sorts
  • Fluid, unstructured, and relatively unmoveable information, such as internal links, social gestures and other application specific metadata

It’s relatively easy to imagine downloading all the files stored in a Yammer account, and uploading them into an IBM Connections instance. But other sorts of information — and semantics — won’t have the same ease of movement.
Consider a hypothetical work media tool — let’s call it Work Talk. Work Talk supports both milestones and tasks, and it also allows tasks to be optionally linked to a milestone. One of its semantic rules is that a milestone cannot be complete until all linked tasks are complete.
Imagine that Work Talk supports exporting all the structured information — files, user identities, and so on — and less-well structured information, like tasks, milestones, posts, comments, and the many relationships between them. Taking that information and figuring how to import it to a tool that is architected differently would be at the least a major programming task and, at the worst, an impossibility. And the semantics of milestones and tasks might simply fail, if the new tool doesn’t implement that capability the same way, even if all that information can be exported and imported en masse.

As the market matures, standards must evolve

We’re at the start of a new era in business software, and there is an explosion of new players and new ideas about how streaming information should be structured and streamed, and how the various bits relate to each other. This is much like the early days of email, when a single corporation might have several different email systems that couldn’t communicate to each other. That problem was solved in two ways: by the emergence of well-defined standards that enabled interoperability across different implementations, as well as the consolidation of the marketplace around a small number of vendors serving large numbers of users.
It’s not too early to see some market maturation. It seems that many of the vendors in the space are making highly similar products, but differentiated around specific market needs (such as integration with specific external tools), focus on specific business functions (marketing versus software development, for example), or emphasizing the size of the company best suited for the tools. I see very little activity on the software standards side, but that usually occurs as the intersection of successful applications, as happened with email and SQL. So, there is no immediate solution in sight, and I wager that a large number of headaches are going to arise from the proliferation of work media tools, especially when vendors go out of business, or when companies outgrow the tool they selected. And there is no simple advice to give to prospective or current users of these work media tools. It is inevitable that these tools will diverge in functionality, and even if two systems are very similar that doesn’t mean that it will be possible to easily and cheaply port from one to the other.
Despite these risks, I believe there are great benefits inherent in the use of work media, and because of those, the rapid adoption of these tools will continue at an unprecedented rate. Just like the adoption of the automobile and the airplane, though, we are going to see a few crashes.
Stowe Boyd writes and speaks about social tools and their impact on media, business and society. A GigaOM Pro analyst, Boyd also writes at and is working on a new book about the rise of a socially augmented world, called Liquid City: A Liquid, Not A Solid; A City, Not A Machine.