The cloud hype is dying down as maturing markets and the rise of utility-based computing conspire to make this technology commonplace. Now, companies are talking about the best way to leverage cloud resources and bolster their bottom line: Big data.
For many businesses, however, the developing nature of information initiatives poses a problem. According to Tech Crunch, most businesses don’t have big data, they have bad data poor quality information from less-than-ideal sources. What’s the impact if companies can’t clean up?
Making the Wrong Moves
The ultimate goal of any big data strategy is generating actionable information to drive ROI. Using poor-quality information as the basis for this action, however, comes with two distinct risks: Making the wrong moves and forcing a company to back track, or lacking enough reliable data to fuel confident decisions.
Solving this problem means getting to the root of bad data. As noted by Business 2 Community, poor-quality data has several common origins: First is missing data fields that should be populated but are left blank. Incorrect values, data entered into the wrong field, and duplicate information also contribute to the loss of data fidelity. Looking for a fix? Start with how your data gets entered rather than focusing on results; clear data entry best practices coupled with IT oversight and process automation can reduce the risk of human error and improve decision-making.
Limiting Sightlines
Bad data can also limit visibility into your organization. This happens when internal metrics such as sales volume, customer satisfaction or even total cost of ownership (TCO) for computer hardware isn’t properly recorded or managed.
Best case? You’re getting out of date, inaccurate information that limits the effectiveness of purchasing or leasing decisions. Worst case? Using poor-quality data leaves you flying blind when it comes to staffing decisions and devising high-level strategy. Eliminating this issue requires removing the distinction between customer-facing data and internal information; both require the same degree of precision, oversight and analysis.
Accidental Analytics
Even companies with solid data can end up with poor-quality results if their analytics methods don’t measure up. As noted by Harvard University professor Gary King in a Tech Target piece, this recently happened to a group of researchers using social media to predict the U.S. unemployment rate. By tracking correlations between words like jobs, unemployment, classifieds and monthly unemployment rates they hoped to discover key predictive factors. The problem? A sudden spike in the word jobs as a result of Apple founder Steve Jobs‘ death significantly skewed their results.
Companies using the fastest, simplest method of discovering big data relationships run the same risk. Sidestepping this potential impact means finding a big data analysis solution that uses multiple methods to track, record and correlate results.
Compromising Security
Companies may also find themselves at risk of security breach or compliance challenge as a result of bad data. As noted by Wired, big data has the potential to deliver an improved customer experience and better bottom line; but in the hands of malicious actors could cause serious problems. Consider the risk of poor-quality data leaked into mainstream media or competitors; not only are companies faced with consumer backlash for ineffective data protection but possible legal and compliance challenges for recording, storing and using this information incorrectly. The easiest way around this problem? Treat big data security as an essential part of good IT security hygiene.
Big data comes with big promises. Make sure you capitalize on the potential and keep your information house in order by avoiding the critical impacts of big, bad data.
