Dark data - one of the latest trending discussions in data and analytics - is currently defined as data that a business creates and saves, but isn’t used to run the business. I’d like to offer up an addition to that definition: Dark data should also include data that a business creates but does not currently save.
The question many people are asking is: What should be done with dark data? Some say data should never be thrown away, as storage is so cheap, and that data may have a purpose in the future.
While analytics is what I do, I am an economist by training and at heart. So, here are my thoughts about Dark data from a business perspective.
So, dark data, unlike dark matter, can be brought to light and so can its potential ROI. And what’s more, a simple way of thinking about what to do with the data –- through a cost-benefit analysis –- can remove the complexity surrounding the previously mysterious dark data.
It’s also very interesting to think about how to value dark data.
Years ago, people collected data that they probably thought was not very useful. For instance, when you contact a call center, you routinely hear that your call “may be recorded for training purposes” – and it’s a training tool that has some value, I am sure. But, what about taking this a step further and capturing a caller’s voice, not for training, but for analysis, to detect emotions? While a caller is in a queue waiting for someone to answer, and murmuring to themselves, it’s possible to detect their emotional state and to use that analysis to customize treatments and interactions for them. When the service representative finally gets on the call, the outputs from the voice analysis mean they are able to serve the customer better and increase the probability of retaining them.
Items that used to be dark, like the voice patterns mentioned above, may have significant value when brought to light. So, when thinking about what potentially valuable dark data may exist in your organization, or even in another organization, it pays to think as broadly as possible. Before conducting a cost-benefit analysis about dark data, think boldly about what can and cannot be done.