The two biggest obstacles to data quality

Data issues CSC Blogs

There’s no point in analyzing data if you can’t trust what you’re analyzing. Data of dubious quality is no better than bad data, and bad data leads to bad decisions.

Forbes contributor Bobby Owsinski, a long-time professional musician, producer and engineer, recently wrote about “the music industry’s big data problem.” The overall point of Owsinski’s column was that the digital age has done little to help musicians get paid what they’re owed, though some startups offer the hope of things changing for the better.

But it was how he described the archaic process of data collection and sharing in the music industry that caught my eye. Owsinski writes:

One of the major problems in the current world of music big data has been that although the streaming services could provide accurate info to labels and publishers, it came in a format that was incompatible with their accounting systems. That meant that all those reams of data (more than ever, thanks to the services ability to granularly collect everything) were delivered in stacks of hard copy, which then had to be manually input into the label or publisher’s system. And of course, the problem was that the person doing the inputting was often an intern or a low-on-the-totem pole employee who was not equipped to deal with some of the more complex decisions that would come up in the course of inputting, which lead to inaccurate statements for artists and songwriters. And let’s not forget the inevitable human error that goes along with manual data entry that didn’t help matters.

This explanation includes the two largest obstacles to quality data analysis:

  1. “…all those reams of data were delivered in stacks of hard copy, which then had to be manually input into the label or publisher’s system.”
  2. “…the person doing the inputting was often an intern or a low-on-the-totem pole employee who was not equipped to deal with some of the more complex decisions that would come up in the course of inputting, which lead to inaccurate statements for artists and songwriters. And let’s not forget the inevitable human error that goes along with manual data entry that didn’t help matters.”

The first problem is about siloed data. Extracting this data from its hard-paper silo comes at a cost, and an ongoing one at that. Further, failure to extract data from siloes leads to an inaccurate picture.

The second problem concerns data quality. In the case of the music industry, it’s about artists not getting paid. In the case of any other enterprise basing decisions on data analysis, it’s about the perils of dirty data.

For analytics initiatives to succeed, it’s imperative that enterprises are able to access all relevant data and ensure the data used is high quality.

RELATED LINKS

Quality data is essential to a quality customer experience

Big data is useless (and even dangerous) without this key ingredient

Too much big data for you? Focus instead on ‘smart data’

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: