Key reasons you can’t ignore visualisation in your big data plans

It's important for businesses to remember when they are planning big data analytics deployments that actually gathering and processing information is only half the job. If firms truly want to make the most of their data, they will have to make sure their staff are able to correctly interpret the results and apply the insights to their decision-making.

But this is easier said than done, and one of the biggest issues facing the industry at the moment is a large number of people are focusing too closely on simply collecting data, without much thought as to how they gain value from this.

It was noted by author Matt Asay that one of the key themes at the recent Strata + Hadoop World conference was how to handle the ever-increasing volumes of information that are available. In a piece for ReadWrite, he observed much of the talk surrounded concepts such as data lakes and enterprise data hubs, as companies make gathering more data a top priority.

But he observed that as these volumes increase, there is no guarantee that the amount of useful details will rise along with it. Therefore, knowing what details will be useful and what can be discarded is one of the keys to success with a big data scheme.

He quoted statistician Nate Silver, who observed: "If the quantity of information is increasing by 2.5 quintillion bytes per day, the amount of useful information almost certainly isn't. Most of it is just noise, and the noise is increasing faster than the signal."

Real insight can only occur if businesses have the tools and skills to make sense of this data, which is why companies need to ensure they have good visualisation solutions in place.

However, this is something that is not happening in many businesses, which Mr Asay noted leads to several issues that can hinder effective use of data.

For instance, he said that in an environment that prioritises data above people, analysts are often unsure about which metrics they need to be focusing on. He explained: "They may know how to pick apart data to discover insights, but don't know how to communicate these through dashboards that tell a story to a particular job function."

Related to this is that metrics are not segregated based on job roles. Different parts of a business will require different data in order to gain insight, but this is not always reflected in strategies that focus primarily on increasing a firm's volume of information.

Mr Asay also emphasised the importance of training staff members in how to think about their data effectively. Without strong, easy-to-use visualisation tools, end-users will typically not be able to transform information into practical knowledge.

He concluded by observed that while it will be important to continue investing in tools such as Hadoop, NoSQL and other infrastructure solutions, businesses must not overlook the visualisation tools that will be needed to bring big data out of the IT department and into wider business units. This will be essential if the employees who will actually use the data are to understand the information they are being given.