Time Silos

Mar 30, 2011 by Jim Harris in Data Governance, Data Quality, Master Data Management

A few weeks ago, I inadvertently caused a scheduling issue with the Data Knights Tweet Jam, which had been scheduled to recur on Twitter every Wednesday from 3:00 – 4:00 PM EST using the #DataKnightsJam hashtag.

In a classic data quality blunder, I absentmindedly assumed that the time offset between Eastern Standard Time (EST) and Coordinated Universal Time (UTC) was a constant when, in fact, it is obviously a variable due to Daylight Saving Time (DST).

Additional complexities are caused by the fact that DST is not observed everywhere in the world, and even where it is observed, the time adjustments are not globally synchronized.

Following a timely tip from Daragh O Brien, I am now using the World Clock Meeting Planner to tweet an accurate starting day and time for the Data Knights Tweet Jam on a week-to-week basis.

Time Silos

In the past, I have compared a Single Version of the Time with a Single Version of the Truth, a frequently debated concept in enterprise data management – the antithesis of which are data silos, where business units, and some individual users, maintain their own version of the truth.

Data silos are often denounced as inherently bad since they complicate, and in some cases hinder, the coordination of enterprisewide business activities. However, data silos are often synchronized with a local business perspective and used to guide daily business activities in much the same way as local time zones are synchronized with a local perspective and used to guide daily local activities.

Essentially, local time zones are time silos. UTC provides an absolute time that is necessary to coordinate the scheduling of world-wide events in much the same way as an enterprise data warehouse (EDW) or master data management (MDM) hub provides a central repository that is necessary to coordinate data management with enterprisewide business activities.

However, it would be confusing to use UTC when referring to events relative to our local time zone, and it may sometimes be confusing to use EDW or MDM when referring to business activities relative to our local business unit, or relative to the perspective of some individual users.

Although I am not trying to defend the existence of data silos, I am suggesting that we consider the complexities that would be inherent in eliminating time silos (i.e., converting all local time zones to UTC) in order to better understand the organizational complexities inherent in eliminating data silos.

There are times when a “single version of the truth” is absolutely necessary and must be enforced, but there are also times when truth is relative to our local perspective and this must also be allowed.

Tags: Data Governance, Data Quality, data silos, Master Data Management, time zones