What is a Predictive Score?

Blending enterprise data and lifedata for predictive analytics

If you do a quick search of current marketing trends, you are likely to see the word ‘predictive’ among the top trending terms.

Predictive marketing has become a focus for many companies recently for several reasons. First, the amount of business and consumer data being created every day is increasing at an exponential rate, and companies know that being able to find and understand patterns within that data enables them to predict future outcomes. Second, digital technology has evolved to a point where a fair amount of data science and data analytics can be automated, allowing companies to derive insights and intelligence from their data much more cost-effectively.

Versium has combined these two things—the enormous amount of data being produced each day and new digital technology advancements that enable automation—to offer a predictive marketing solution. This solution can be used by any size company, and is now available on a self-serve basis at Datafinder.com.

Versium’s predictive scores show the probability any given individual will take a specific action. What that action is can vary: you can create a predictive score to show how likely someone is to transact, to commit fraud, to donate to a cause, to cancel membership, to become a brand advocate, and more. Companies can also create their own custom predictive scores to answer almost any business question imaginable (e.g. What color of my product is this person most likely to purchase?). While the math and data science that goes into predictive scoring is somewhat complex, the scores themselves are very easy to interpret. This is because a predictive score can be built to take into consideration many different factors at once but gives back a single output—the propensity of someone to act in a certain way.

To show how this works, let’s look at ‘green’ scores. At its most basic, a green score answers the question of how environmentally conscious someone is. However, without also considering other factors that pertain to your specific business objectives, the value of knowing whether or not someone is environmentally conscious is somewhat limited.

For example, if you are selling solar panels and are only looking at whether or not someone is environmentally conscious, you may miss the fact that some people who are environmentally conscious live in areas where solar panels are too inefficient to make sense. However, by making geography a factor in your predictive score that tells how likely someone is to buy solar panels, your predictive accuracy is greatly improved.

Predictive scores can be built similarly—with various degrees of complexity—to answer almost any business question you have. You can then use these scores in a variety of ways, including improving targeting and segmentation, increasing content engagement, preventing attrition, reducing list costs, improving the lifetime value of your customers, and many more.

Related Posts