Cloudy days for data, Part 1

Several years ago I was in a product data management role for a major US manufacturer, assessing our information management landscape and helping my boss develop a road map for bringing the 160-year-old company’s engineering systems and processes into the modern age.  What I discovered shocked me although I really should not have been surprised: the vast majority of our mission-critical business data was sequestered in spreadsheets and shared virally via emails.

This sort of working environment tends to spring up as a consequence of two conditions:

1. Often the information management system(s) are lacking in necessary features, disconnected, difficult to utilize, poorly represented or even non-existant;
2. People want to hold on to their stuff

Anecdotally, I found the latter to be the greater evil.  When information managers try to improve the first condition, they encounter resistance due to the second.  After all, information is power, and the gut feeling is that if we relenquish any control over it we lose apparent value.  So the people who could benefit most from fixing broken sharing systems often hurt themselves by actually becoming part of the problem.

There’s a common fear among the rank-and-file that processes are so people-dependent that if any aspect is automated, the result will be a headcount reduction.  This can be very true, but does not mean the converse is absolute; I can attest first-hand to the reality that subject matter experts are released from employment regardless of their intrinsic value or the sober fact that no one else in the immediate organization can backfill the lost role.  The decisions to eliminate a position, regardless of its criticality, are typically made from lofty seats far removed from the playing field.

 The desire to control information ipso facto creates a culture of file-centric thinking.  Knowledge managers of any sort (and I use that term very loosely– we are all knowledge managers) highly covet their spreadsheets, their word processing documents, their presentations.  We even assume ownership of commercial media, such as music and graphic files, that’s actually owned by other content providers.  Intangibles possess, for us, an abstract tangibility.

Information management systems tend to be at odds with the very concept of a file.  In these environs, well-managed-yet-flexible data sets are paramount, coupled by normalized tables and key field relationships and massaged into myriad read-only presentations.  But too often I’ve seen a critical piece missing: the middle of the puzzle where report servers and other important pieces to content delivery lie.  After all, most of the people involved in the data sharing process are consumers, not creators, and all they want is a snapshot of certain formatted data.  They don’t really need to lug around the baggage of files.  But without the necessary elements of infrastructure to support their actual needs, they are begrudgingly forced to become unwilling (and unqualified) content managers and subsequently introduce chaos and wasteful overhead (i.e., file storage needs) into the system.

This is not only a conundrum for system developers and administrators trying to herd disparate domains and data sets into a cohesive whole, but for today’s highly-mobile workforce in general.  This growing body of on-the-go employees demands constant access to their information, in business ecosystems often not up to the task. Typically this means syncing up to the source of their files and downloading copies to the handheld device of choice, whether that be Blackberry, iPhone, E90, other smartphone… or in more limited usage, internet tablets.

This is where cloud computing is supposed to save us from a data access drought.  But the cloud cover is currently spotty, at least in the US, and there’s still a singular, significant fear to overcome: the loss of the file.

In part two of this series, I’ll tie in blog articles on the subject by Michael Dominic K. and Henri Bergius (note: only by inference), and explore the current shaky landscape of data access and management as well as what I think the future will hold.