Wednesday, 24 January 2007

Social Data Analysis... But what about compliance?

OK, when I heard about Swivel the other day I thought it was good, but Many Eyes from IBM's Alphaworks is even more impressive because of the charts available, including network maps and treemaps - something even Excel can't do (as far as I'm aware).

Both Many Eyes and Swivel also hint at the possibility of providing private access to graphs and data (Swivel talks about a fee-based "professional" version), but I can equally see potential for a "social data analysis" tool like this inside the firewall with options to restrict data to groups or specific users. Of course sharing the data as widely as possible is what's its all about.

Its also interesting to consider with these social data analysis tools that not only are they making data analysis "social", but I can also see tools like begin to challenge our concept of a basic information unit like a "document".

Think about this - typically to create an ad hoc graph you might paste the data into a spreadsheet and use the tools in the spreadsheet to create the graphic. This activity is something people must be doing everyday all around the world and the net output of course is a spreadsheet file containing the data and the graph that gets stored somewhere. But in a Web 2.0 mashed up world there is no concept of a document with social data analysis tools. In fact it would defeat the purpose.

Unfortunately for an enterprise thinking of using social data analysis tools I can see that this may create a compliance and records management problem, unless the current generation of electronic document and records management systems evolve to automatically capture both permanent structured data and transient unstructured in context together. Now, that's going to be interesting to watch.


  1. James,

    The vision you outline for the future of data is an exciting one. What got us started on Swivel was the belief that our best insights into data usually happened in 'public' in a social way.

    And that instead of knowing what we were looking for ahead of time, many of the best insights often came while going rapid fire (well as fast as possible in Excel at the time) from one hypothesis to the next.

    That's why when you see a data set like this in Swivel you'll see that it's not about pre-configuring the visualization you think you want, but instead rapidly cruising through all sorts of variations.

    My hunch is the gap between this kind of hypothesis building and compliance issues is pretty large.

    But your point about the evolution of "electronic document and records management automatically capture both permanent structured data and transient unstructured in context together" could give ideas like Swivel or ManyEyes a lot more gravity..and who knows, it could evolve into a sort of...operating system for business? (ok, that might be a stretch ;~)

    Inspiring post. Thanks for mentioning Swivel.

    Brian Mulloy
    CEO & Cofounder

  2. BTW, I'm Brian Mulloy, CEO & Cofounder of Swivel. Nice to meet you :~)


Note: only a member of this blog may post a comment.