The Five C’s of Big Data Management for Retail Analytics

Insights AnalyticsBlogDataTechnology
Eric Green
By Eric Green
Sep 04 2014

Many people talk about the opportunity that’s created by Big Data analytics, but before we view it as potential, first we need to recognize it as a fact. The multiplication of touch-points along the Internet path to purchase, the influx of affordable sensors, and the deep penetration of mobile technology have all exponentially increased both the sources and amount of data available to companies.

At one time Big Data was just a new feature on the Business Intelligence landscape, but now it’s become the lens that business leaders use to study the whole field, from consumer behavior to supply chain efficiency to marketing and ROI. The question is not whether to look through the lens, but how to focus it.

“The only thing you have to change is everything.”

When it comes to making changes, this might not be the first thing you want to hear. But as scary as it sounds, no other principle or axiom is going to better prepare you to transition your company to Big Data-powered systems.

Anyone who’s made a serious effort to alter a habit or improve a behavior knows that change often creates a domino effect: as one pattern shifts, many others move with it. Similarly, as vendors and retailers seek to enter the new paradigm of Big Data analytics, it quickly becomes apparent that many other adjustments flow out of this decision. Those changes fall into five broad categories.

1) Consolidation

One of the greatest impediments to finding value in Big Data is fragmentation. Often customer information is scattered haphazardly throughout different databases, from transaction logs and loyalty programs to social networking and third-party sources.

Retailers and vendors need the right kind of data architecture to deploy effective retail analytics, but companies can’t construct the necessary framework for data usability until they’ve first laid a solid foundation. Consolidating and normalizing databases is the first step toward building profitability.

2) Cooperation

Consolidating data streams necessitates cooperation between the various data owners. Deriving value from Big Data analytics depends on the transparent flow of information — from business to business, from department to department, and from business to customer.

Retailers and vendors need to work together to integrate their data or they will be stuck with truncated information, which in today’s intelligence context is tantamount to disqualification. But ultimately, it’s consumers that are driving this demand for collaborative synchronization of data between retailers and vendors, because consumers are learning to expect targeted marketing, sharpened assortments, and localized promotions, all the way down to the store level and even the individual.

3) Correlation

Until the advent of data tools that could realistically collect ALL the information streaming from inputs, as well as the invention of storage hardware to contain such vast quantities of information, if business analysts wanted to know something, they had to develop a theory about it and implement focused research. It was creative work, but severely limited, and rooted in conjecture rather than evidence.

Business Intelligence works differently now. Extrapolating and predicting results from small databases such as surveys no longer applies. By overlapping vast sets of real information, analysts can see connections that are wholly unpredictable through guesswork. Think of Target identifying pregnant mothers in their first trimester based on purchasing patterns. In order to get these kinds of results, companies need to shift their creative efforts from imagining causation to envisioning correlations.

4) Culture

Learning to discover insights by correlating data speaks to a deeper cultural shift that organizations must adopt to improve decision-making. The skill business leaders need is in recognizing the implications of the objective information, not validating their own conjectures. Moving from theoretical to evidence-based processes is one of the most radical changes a company can make.

“Retailers need to shed some of the art and embrace some of the science,” asserted the IBM Institute for Business Value in a recent report on analytics. “The top one-fifth of Leaders base all their business decisions (that is, both strategic and operational) on information provided by analytics.” That’s good, but it’s still a small fraction: for the most part, decision-makers are relying on other skills — instinct, experience, and estimation — rather than hard evidence. Companies that access the wealth of value in Big Data are able to moving away from guesswork at a systemic level.

5) Capability

None of the previous methods for managing Big Data are possible without the right tools for interpretation and analysis. The information is useless if businesses can’t process and visualize it. The point is not to apply more pressure to people in the system, but to provide automated resources that will free up workers to translate real information into genuine insights.

Thankfully, companies no longer need to purchase dedicated hardware and hire a staff to run it. That’s why on-demand database services represent a rapidly growing sector of the IT industry. As the barrier to entry basically evaporates, Cloud-enabled database platforms have become the primary way that vendors and retailers obtain and use their analytics.

Download our Big Data Analytics vs Old School Reporting infographic to learn how today’s retail analytics tools provide accessibility, efficiency and actionability.

Share

Other Posts