Descriptive & Predictive Analytics
Mark Twain famously said, “There are three kinds of lies: lies, damned lies, and statistics.” Essentially, statistics (which we will from here on in this post refer to as ‘analytics’) are thrown around frequently without a clear basis. Without proper understanding of different types of data analytics, marketers can be easily misled with data.
The reality is that analytics convey a vast amount of important information that needs to be analyzed in order to effectively construct and optimize marketing campaigns and strategies. Marketers need to understand how to use analytics to drive better results from advertising campaigns and for their company as a whole. This post will overview descriptive and predictive analytics, and the role they play in data driven marketing.
Descriptive analytics is defined as: a summary statistic that quantitatively describes or summarizes features of a collection of information. Most people reading this blog post are probably familiar with descriptive analytics, but have never known to classify the information as such. Descriptive analytics are interwoven into most things revolving around life.
An example of descriptive analytics in marketing would be: less than 10% of your customers buy more than once, this ad campaign resulted in a 7% uplift in revenue comparatively to the last month of sales, or creative set A performed better by a factor of 3 than creative set B in our A/B test campaign. You get the point, these are analytics that describe data for what it is.
Predictive analytics are different. Predictive analytics models are analytics to predict the outcome of data in the future, or in other words, it is a branch of advanced analytics which is used to make predictions about unknown future events. Predictive analytics typically uses historical data, data models, specialized software, machine learning and artificial intelligence to make predictions about what will happen in the future based upon what information we have in the present (with some caveats, but that is for future posts).
For example, if we collect data on a sample of individuals observing their age and yearly income, we can plot a graph of an audience where the X axis is “Age” and the Y axis is “Yearly Income”. From there, we can draw a regression curve across the data set. After the data is loaded into the model, upon encountering new data (in this case new individuals), by learning the age of the individuals we can predict with “relative certainty” their yearly income. This example is an oversimplification of a complex variable (someone’s age relative to their income), but you get the point.
To summarize what you learned in this post: descriptive analytics help analyze the historical performance of your data-driven campaigns. Predictive analytics will help you predict the success of future campaigns, while also optimizing current campaigns based upon the data currently in use. Both play an extremely important role in your marketing efforts. We suggest that beginners start with reading and understanding descriptive analytics prior to crossing into the field of predictive analytics.
For more information on the role of descriptive and predictive analytics in your digital marketing strategy, please reach out to us through the Untitled contact form. Our team comprises experts with thousands of hours of training in the field, and can assist you in effectively leveraging both types of analytics to produce value for your organization.