This piece is the first installment of a multi-part series called “Getting your Analytics Program off the Ground”

It seems like everyone is touting analytics these days, whether it is stories about its ability to predict market trends, coalesce seemingly disparate information into knowledge, or streamline business processes. On the other hand, horror stories – such as a store uncovering personal information about a teen before her parents, based on its analytics on her recent purchase history – also permeate social media.

So what is the definition of analytics? According to Wikipedia, the short answer is that analytics is the practices and processes employed to “facilitate discovery, interpretation, and communication of meaningful patterns in data.” These patterns are then analyzed to provide insight into business data, decisions, or processes. The long answer is a good deal more complicated, but one we have endeavored to outline here. Analytics is an umbrella practice, in that it has applications in many fields and can be employed for many purposes. The information patterns arising from the use of analytics can increase revenue, enable decision making, improve customer or public engagement, enhance operational efficiency, disclose compliance issues, reduce risk, and even predict problems before they occur. However, it is not a panacea, nor is there a ‘one-size fits all’ approach that will be effective. The unifying feature of analytics is that it is difficult to initiate and implement, and nigh impossible without the assistance of experts, who often only understand one field of analytics.

There are many hurdles that complicate the process of discovery, interpretation, and communication of the data. The most immediate issue is the quantity of data that needs to be analyzed. In the world of big data, the data sets are so large and complex that traditional tools are inadequate for sorting through them. To make matters worse, much of this data is unstructured. Defined by being text heavy or otherwise defying apparent classification, it is understandable how unstructured data can further complicate the difficult process of applying analytics to big data.

Another hurdle to producing relevant analytics work is simply time. Statistics like buying behavior change substantially over time, and other information can similarly become irrelevant. Thus, data has to be processed quickly and efficiently, or it can deteriorate in value. Information must remain relevant for the analytics process to be accurate, but is often stored for longer than this period. As obsolete information piles up, legal risks develop for those without clear retention schedules. The process of gathering information can cause privacy and ethics issues, since it can involve monitoring data usage patterns from individuals to predict future behavior.

One of the most obvious hurdles can be the most difficult to grasp and encompasses much of what we’ve discussed so far. Simply put: what is relevant? Incorrect data sets can obviously lead to bad outcomes and inaccurate conclusions. Less obvious is how to exclude incorrect or irrelevant data sets. As an example, consider the 2016 election, when President Donald Trump subverted expectations and emerged victorious, baffling political scientists across the country. Relevant data was overlooked, incorrect conclusions were drawn, and election analysts are left to question their suddenly irrelevant data collection strategies.

The overwhelming stockpiles that comprise big data can dissuade the organizations that need analytics the most. Employing analytics in areas rich with recorded information can offer efficiency and information to improve processes throughout an organization. The use of analytics can improve public health outcomes, prioritize information governance initiatives, lower acquisition costs, and reduce the cost of a data breach. But where do you start? Follow Cadence Group as we explore realistic and actionable approaches to get analytics programs off the ground. View our success stories with work in data analytics to create highly successful analytics program. Follow us as we post more about data analytics in the coming future for actionable strategies and information.  For more information on how we apply that knowledge and know-how through a customized level of service see our Practices Page.