Big Data Analysis: Potential & What it Means for Research

There are essentially two different ways to look at any relatively new and game-changing aspect of business tech. One example from online marketing is social media retargeting through Facebook and Twitter. A key trend related generally to online tech is the handling and analysis of “Big Data” – the massive amounts of information related to user behavior and other parameters available through your site and applications – combined with approaches to utilize the findings to grow your business.

With both of the above business movements, decision-makers must determine how strongly they want to embrace the new model and dedicate themselves to being at the forefront of its application. Competing with the desire to outpace competition is a reluctance to go “all in” with a new approach to web optimization.

This article will show how vital big data is becoming for online success, giving a general sense of the potential of strong big data analysis for any business. It will also zone in on one field, medical research, to show the specific ramifications for one area.

What big data has to offer

Management consultancy McKinsey & Company has focused extensively on big data for the last few years, with broad information on its website related to the topic. An article published in early 2013 by Stefan Biesdorf et al. argues that the question of whether to implement strategies to make as much use of big data as possible is “no longer in doubt.” In fact, taking significant forward steps with regards to big data could give you a 5 to 6% boost in both productivity and profit (according to the article “Making advanced Analytics Work for You,” from the October 2012 issue of the Harvard Business Review).

Clearly big data analysis is compelling. However, truly committing to it and strategizing how to integrate it into your business is not simple. Or is it? Biesdorf et al. suggest that succeeding with big data is primarily about taking one action with sufficient care: create a plan. Only one key component is missing for the majority of businesses that don’t achieve the goals they desire with big data, and that is simply that they did not strategize and organize their approach to maximize potential of hitting intended targets.

Here are the three basic elements of a big data analytical plan:

The data itself

First of all, you need to figure out how to bring all of your data into one aggregate, because it probably is stored across various pieces of your infrastructure. The plot thickens when you consider such elements as your social media pages. Ideally you grab data from every possible corner.

The choice of a model

You of course don’t need to use just one model, but you do need to figure out what model of analysis will work the best for a given situation. It’s not just the data that matters. The model is crucial to allow you to create the greatest business advantages at the lowest possible cost. What you don’t want is a model that leads to inconsistencies or unneeded replications. It should be streamlined and built for efficiency.

The usability factor

Obviously you do not want to deploy a big data analytical system that requires an IT degree to understand. You want a system that is intuitive both for executives and possibly any employee, with a basic interface that puts your data and analytical outputs into formats that are easy to read and understand. Additionally, the system should provide what can or should be done as a result. This piece of the puzzle is absolutely critical, because it turns the big data into actionable information for any party involved.

Big data sample applications – healthcare

Writing for the Harvard School of Public Health, Elaine Grant states that the impact of big data is massive within medical research because it allows “a new way of doing science.” In the traditional model of science, researchers have to start with a hypothesis. Big data makes it possible to perform experiments that themselves produce hypotheses. For instance, Pardis Sabeti (a professor at the HSPH), while conducting big data analysis related to natural selection, accidentally happened upon figures that he felt might lead to a cure for the West African disease Lassa fever.

Winston Hide, another HSPH professor, believes that up-and-coming researchers who are more comfortable with social media are likelier to have an “open source” attitude toward data. That different perspective could fuel a trend toward a share-and-share-alike model for scientific research, as opposed to hoarding of information until publication.

Similarly, crowdsourcing allows academics, businesspeople, and the general public to team on projects in ways never before possible. However, academic culture makes these new possibilities confusing because getting personal credit on a project and controlling the parameters of a study become fuzzier.

Big data has similar profound implications for all organizations and all businesses. Just develop a plan – remembering to consider all your different types of data, models to analyze them that makes sense for business growth, and ease-of-use.

By Moazzam Adnan of VPS & cloud hosting provider Atlantic.Net.

The post Big Data Analysis: Potential and What it Means for Research appeared first on Atlantic.Net.