Blog

How can you extract insight from your data ‘noise’?

One of the biggest worries many companies may have when they are investigating the potential of big data is that they will end up becoming overwhelmed by huge amounts of mostly irrelevant information that will make it harder to extract useful insight.

But it may well be the case that with so much information coming into the business, companies could be able to harness this for uses they had not previously imagined. In fact, with the range of data organisations are able to gather on customers and the wider market growing, the chances are they will be able to uncover information far beyond the project's original purpose.

It was recently noted by Infoworld that many companies are beginning to find there is useful data to be gained from what has previously been dismissed as 'noise'. Indeed, it highlighted the recent example of fitness tracker Jawbone, which uses wristband sensors to gather details on a user's activity.

The technology is designed to gather details such as the number of steps a person takes and sleep patterns to help users manage their health. But this month, it discovered another use for its sensors when an earthquake hit California.

Jawbone published a graph revealing how many wearers of its wristbands woke up when the quake hit the Napa Valley region early in the morning – with this showing the time when people awoke and the number of people disturbed related directly to their proximity to the epicentre.

Infoworld noted this is just one example of how the information gathered by businesses can be used in ways other than intended. But for enterprises, this could offer many new opportunities to learn about their customers, provided they have the right tools in place to analyse data and skilled personnel who are able to interpret the results.

"Imagine the data you're capturing data for your business. What useful noise might it contain?" the publication said. "Therein lie key opportunities for business development, loss prevention, cost reduction, and especially supply chain planning. These opportunities span business units, demand data liberalisation, and require pooling data beyond what we can plan for."

It noted that in order to succeed this, companies may need to take a different approach to how they look at their data. Infoworld observed that while conventional data miners may be very effective at pulling out useable insight, they do have a tendency to "put the blinders on" and only take a narrow view of their data.

"You need domain experts to get involved and look for stuff you don't realise is important yet," the publication said. "This requires people to be curious, ask questions, and seek serendipity." 

It can be a fine balance between concentrating on key indicators they businesses know they will be able to use to improve their performance of their business and more wide-ranging, investigative approaches to big data. Therefore it will be essential that companies have the right big data analytics tools and support to process information quickly to ensure time is not wasted on processes that may come to nothing. 

But businesses should not be afraid of failure when undertaking these activities – provided that they make sure they 'fail small' and can still learn from the results they get to boost their later efforts.