Choose your language:

Hong Kong
New Zealand
United Kingdom
United States

3 underappreciated factors for achieving Big Data success

September 05, 2013

Without a doubt, Big Data analytics is one of the most important IT services to enter the mainstream in recent years. This technology allows firms in every sector to gain invaluable insight into their own processes, their competitors' behavior and their clients' demands and interests, all thanks to the the mining of unstructured and semistructured data.

This much is widely known. What is less clear is precisely how businesses can and should go about achieving the desired results from their big data analytics efforts. The technology has tremendous potential value, but there are no guarantees. Firms must be careful and strategic in their use of these resources if they want to see useful results.

While some aspects of big data analytics efforts are well-established, such as the need for advanced analytics software and a large volume of raw data, many other critical factors are frequently overlooked. With that in mind, here are three vital yet underappreciated factors for achieving Big Data analytics success.

1. Qualified personnel
This is probably the single most undervalued component of any Big Data analytics effort, and consequently the issue which leads to the greatest amount of frustration and disappointment.

The reason for this is the simple fact that big data analytics is not an entirely automated process. While automation will play a critical role in any effective solution, the reports produced will ultimately require human interpretation in order to yield any meaningful insight. The information produced will not immediately exist in an understandable, usable fashion.

This means that firms that invest in big data analytics but do not provide sufficient human talent to oversee these systems will almost certainly find their efforts prove unsatisfactory.

To highlight this issue, consider the case of J.C. Penney, as reported recently by InformationWeek's Michael Healey. Last year, Ron Johnson became the CEO of the company and initiated a controversial plan to do away with coupons in favor of "everyday value" prices. As Healey noted, Johnson cited a wide range of data to justify this decision. However, the strategy proved disastrous, and J.C. Penney's sales fell to $12.9 billion in 2012 from $17 billion in 2011. Recently, the company's new CEO, Myron Ullman, proposed a number of new plans to revitalize the company. And as Healey pointed out, he also cited data to justify these decisions. These two executives presumably had access to the same raw information and the same analytical abilities, but reached very different conclusions.

"[It's] safe to say this: Even in the age of Big Data analytics, getting the right answers is tough," wrote Healey.

This incident highlights the important role that interpretation plays in any big data analytics effort, and consequently the need for firms to invest in IT support staff when pursuing these strategies. Only well-trained, experienced Big Data experts will have the ability to interpret analytics reports effectively, leading to optimized operations.

2. Data movement solutions
Another incredibly important, often overlooked factor for achieving big data success is data migration. Without data movement solutions, firms will be unable to transfer their raw data to and from various personnel, departments and networks. And if this data is unavailable or becomes accessible too late, the value of any analytics performed upon it will be suboptimal.

One key aspect of big data is that its value is highly correlated to its freshness. As this information ages, it loses relevance and, therefore, utility. Firms need to develop the means of performing analytics as soon as possible.

Data movement solutions are essential because this data will typically come from a wide range of sources. In addition to on-premises generated data, other information will be collected from external sensors, connected networks and cloud environments. To prove useful, all of this data will need to be brought into a single, unified data warehouse.

Legacy data movement tools are not equipped to handle the huge volume of big data. That is why companies need to invest in dedicated, optimized big data movement solutions.

3. Dedicated maintenance
This may seem obvious, but it is worth reiterating. Big data analytics projects are extremely complex endeavors, more so than most IT solutions. The amount of information that must be collected, transferred, analyzed and stored is tremendous, and any failures in these areas can be devastating. Most notably, a data breach which exposes such records will undoubtedly severely tarnish a firm's reputation.

Consequently, any company pursuing big data efforts should install dedicated maintenance teams of IT support staff to ensure that these programs are monitored at all times. This way, any mistakes can be caught and rectified before they cause undue damage and without sacrificing effectiveness or productivity.

Blog Archive