Data Points

Expand all | Collapse all

Why Old Data Can Be Bad Data

  • 1.  Why Old Data Can Be Bad Data

    Posted 06-13-2019 11:53
    Edited by James Tobias 06-13-2019 14:36
    Imagine you work for a company that just purchased some new data this month to better make business decisions. You acquire the data file and quickly upload it into your usual database server and begin exploring the attributes. The first thing you research is some known fact or location that you know well. For example, generally with spatial data people will start by exploring their house or office location. However, you realize on first glance the data is already over one year old! What would you do -- does this data still meet your company's needs or is this data now worthless?

    Data often requires updates. Although some data is purposely historical, often companies need data that is current to make the most accurate business decisions. After a certain point, stale data can become worthless data, or worse, "bad data!" If a company fails to update their data, the old data can be inaccurate, incomplete, or unavailable leading to "bad data." And if the company continues to base business decisions on that "bad data," it can cost the company heavily.

    Whenever a company is researching how to base business decisions on data, be sure to research the data currency and if updates are available. Most of the on-premise data I have worked with in my career has been updated on a certain cycle. For example, monthly, quarterly, or yearly. If the latest data is required for your company, you can often retrieve the most recent updates in an API format without having to consistently update your on-premise database and sometimes retrieve even more frequent updates than on-premise data. Be sure to research the available options and determine which update cycle is best for your company, based on your use case and budget needs. And remember, don't be that company using "bad data!"

    What else can make data "bad?"

    James Tobias
    Data Product Manager
    Pitney Bowes

  • 2.  RE: Why Old Data Can Be Bad Data

    Posted 06-13-2019 13:01
    agree completely, James...I usually like to use a produce analogy with data - produce doesn't get better with age and nor does data.  Depending on the data (produce) the decay is faster.  Geographic features are at one end of the decay curve, with other features like restaurant POIs decay very quickly due to their frequent name changes or business failures.

    Tom Gilligan
    Pitney Bowes Software, Inc.
    White River Junction, VT, USA