ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

How predictive scores are changing big data

How predictive scores are changing big data Chris Matty

When it comes to big data, marketing departments are frequently at a loss of how to make the most of it. With all the information about data science, predictive analytics, and using big data to make better marketing decisions, it can be overwhelming. Big data is a new and dynamic industry, which means there are plenty of big opportunities for CMOs, but also potential for lots of confusion. What many don't know is that the answer to most of this confusion exists with a simple, easy to understand "predictive score."

Reuters News reports that big data will grow by 45 percent annually to reach a $25 billion industry by 2015, and a recent CMO Report states that 66 percent of all chief marketing officers are allocating more than 25 percent of their budgets to mining and understanding how to apply this data.

There are three key stages in the evolution of the big data industry. The first stage involves the introduction of infrastructure that has enabled the industry to operate; for example Hadoop, an open-source software programming framework that supports the processing of large data sets in a distributed computing environment.

The second stage involves the creation of tools that allow increased access to the new-found value that this infrastructure enables -- such as tools that support the visualization of data allowing us to see large and complex volumes of data in an organized fashion.

The third stage, which we are preparing to enter, will be where focused applications deliver true ROI value -- specific predictive analytics that deliver actionable intelligence to either increase revenues or reduce expenses. Marketing departments that focus on how to leverage data to drive change and/or optimization based on specific business case objectives will find the greatest value. Solutions that look to do the same will gain significant market share.

One of the biggest challenges with big data is that the amount of data that exists can be massive, incredibly complex and, at times, unreliable. When looking to obtain value from data, you must always ask yourself if the data you have is accurate and/or relevant to the business case and, most importantly, whether it is actionable. One CRM company notes that as much as 40 percent of the data it collects is inaccurate and not usable. In addition to your data, there are multiple external data sources that can complement your own data mining efforts and help address the questions noted above.

Ultimately, all CMOs must ask themselves these important questions: Do you really want all the headaches of dealing with all of this data, hiring hard to find data scientists, and investing in expensive platform solutions? Will the data provide value and ROI? Or will it be the intelligence that comes from the data that does that? Or, to simplify things: Do you really just want the answer to a question? If you can boil the business objective down to a question, the data -- and more importantly, predictive analytics -- can likely deliver considerable value.

More than 70 percent of marketers believe that the marketing function will significantly change over the next five years according to another CMO Report. Marketers are interested in understanding why customers want what they want, what is driving their decisions, and if it changes their buying behavior. This is how they identify and win share in new market segments.

External data can help to improve your ability to answer questions for your business. This is because data can reveal the psychology of behavior. It is better for a business to understand the psychology of consumer behavior and how data can help you predict that behavior.

Ultimately, when we talk about the psychology of consumer behavior, we are talking about who a person is -- this includes their characteristics, motivations, and desires. People with different characteristics (data attributes) behave differently, and those with similar ones tend to behave similarly. This is where predictive analytics comes in. Can data predict the types of consumer behavior that address the questions most relevant to businesses? Computer models can be built that incorporate internal and external data that predict churn, fraud, and consumer response to offers and distill all that information into a simple to understand and actionable score, similar to a credit score. It's much better than using incomplete or inaccurate data analysis, or worse, blindly guessing at how consumers will behave. Understanding the increased propensity to behave a certain way allows for action to be taken to optimize the result a business wants, or in the case of fraud, prevent a negative outcome.

The important fact in all of this is simple: CMOs, or marketing departments, do not need to understand big data or data science. An enterprise does not need to hire expensive data scientists to get answers to its toughest questions. Ultimately, a business needs to find the data that can help answer its questions and deliver ROI, without piling more work onto its marketing department.

The answer, then, to how businesses can cost effectively use big data is this: predictive analytics as a service where the insight is delivered via a predictive score. A financial institution doesn't need to hire data scientists to analyze volumes of consumer credit related data for each applicant; instead, it uses a third-party service that sends it back a simple number that answers the question, "Will this person most likely pay me back?" High score: yes. Low score: maybe not. The same can be true for enterprises looking to address specific business cases. Which of my customers are most likely to cancel? Who is trying to scam me? Which person will respond best to which offer? Who is most likely to donate? Which household is most energy or environmentally conscious? Predictive analytics as a service built into the cloud will answer these questions by delivering easy to understand scores. Contained within a predictive score are all the complexities of various predictive models and interdependencies and associations of complex data sets.

Businesses do not need to buy an expensive analytics platform and hire data scientists to run it and interpret the output for the marketing department to be able to extract value. Instead, businesses need only find solutions that deliver predictive analytics in the form of "scores," similar to credit scores. Businesses can then use these scores to find the correct consumers, to decide how it thinks consumers will respond to an offer, or to reduce churn and fraud.

Big data has many opportunities for businesses. But it's important for a business to use data in a way that is most cost and time effective for it. Nobody wants to take on more work if they don't need to. Why purchase an expensive platform, train your employees on a new piece of software, and hire data scientists to analyze the data for you, when you can simply use predictive analytics as a service? By partnering with a service that can provide you with predictive scores, you can gain an edge over your competitors and use big data to your advantage -- without the big headache.

Chris Matty is the co-founder and CEO of Versium.

On Twitter? iMedia Connection at @iMediaTweet. Follow Chris Matty at @versiumceo.

Chris Matty, co-founder and CEO at Versium, has led a number of early stage tech companies in the Data, Internet, Mobile and SaaS industries, leading strategic initiatives in business development, sales and marketing. His latest venture is a data...

View full biography


to leave comments.