When Good Data Goes Bad

Now might be the time to admit it: You have a data problem. You’re surrounded by it. It’s flooded your desk, overwhelmed your hard drive and is piling up like floor-to-ceiling garbage in your consciousness. And you’re afraid to throw any of it away. “What if I need it one day? I invested in it, so it has to be worth keeping, right?” You’ve ended up in this cluttered mess, surrounded by information when all you really wanted is to be surrounded by answers.

 

You’re a data hoarder. But you are not alone and help is out there.

 

For the last couple of years, Big Data has been the fixation of every marketer struggling to understand its audience better, and for understandable reasons. We know the world of content is fractured. People consume wide, varied streams, from traditional TV to search engines, social networks and mobile apps.

via www.adweek.com

I ran across this article today on Ad Week and I couldn’t help but agree–in a world where content is being created exponentially, how do you even begin to comprehend all of the different data points measuring that content? First, I agree with Ms. Klau that we do need to start cleaning up our act, and really understand what data points we really need and what questions they answer. Collecting data just to collect data is useless. Maybe, it gets shoved into a multitude of Excel spreadsheets and PowerPoint graphs for future reference. Or maybe it just dies a small death in a data file somewhere. So understanding the objectives first is important before simply embarking on a research project.

But what Ms. Klau does not touch upon in her commentary is the next step in cleaning up our act: after you know why you need that data point, make sure that the research is being done in the most objective or unbiased manner. Are the metrics being collected in the right manner? Are populations being left out? When was the data collected? How was the question asked?

This is just a start of the questions marketers need to ask before making decisions based on the data they have available to them–don’t just accept “garbage in”, because you really will get “garbage out”.