Send email now: office@guestpostshop.com

Understanding the Three Different Types of Analytics

In a previous blog post we were arguing on the absolute importance of analytics by giving you for reasons that you should take them into consideration, no matter the line of work you are involved in. It doesn’t matter what it is that you specifically do in online marketing – from strategists to creative types, just about anyone and everyone can benefit from having access to data. That being said, data can be quite daunting, especially because there is a lot of it. Web designers and copywriters in particular will tend to feel rather overwhelmed when faced with spreadsheets upon spreadsheets of seemingly endless columns of figures. That’s where the strategy department steps in and culls the relevant facts from the larger set of metrics, which they then subsequently deliver to members of other departments. And even then, one must bear in mind that not all types of metrics were created equal. In other words, there are different types of analytical data available for interpretation, and they all serve different purposes. Here are the main three types, explained in brief below.

Descriptive data

This is the kind of data you should be pooling and looking at on a daily basis, in order to get the picture of the evolution of your website in time. One example is overall traffic, analyzed from a historic perspective. At some points in time you experience a boom, while at other traffic dwindles. In order to get behind the reasons for this, you mine for data, in search of an explanation. Other examples include the analysis of campaign performance, on a monthly, quarterly or yearly basis; similarly, analysts will also look at descriptive data for exceptional situation, such as technical downtime, in order to gauge its effects on the other key performance indicators of their website.

Predictive data

Where descriptive data is used to formulate questions and provide subsequent answers, predictive data steps in to take the process one goalpost further down the line. In other words, it will provide a probable continuation for the course of action that a given project should undertake, based on prior and current KPIs. It will also help you get as clear a picture as possible on what the performance of a site, page, or campaign will be for the coming month, quarter, or year, based on previous records. While this seems commonsensical enough a step to take, many marketers stop at analyzing historic data and then find themselves stumped when it comes to figuring out where they’re headed next. It’s true that this type of data analysis needs to be based on a very strong, well tested predictive model. Once that’s taken care of, though, it is immensely useful and even reassuring to employ.

Prescriptive data

Getting prescriptive with data is probably the most challenging, as well as the most creative type of analysis one can perform. That’s because it involves a fair amount of speculation and deduction. It’s the one step in which strategists and marketers get to peer into the future, in an attempt to figure out what should happen next, given the logical predictions and current standing. Many marketers avoid this step altogether, since they believe trends outlined by predictive data are not to be tampered with. That, of course, is not only wrong, but it’s also highly unprofitable.

Leave a Reply


*