More data was created in the last two years than in the previous 5000 years, this according to a report from IBM Marketing Cloud. We are now reaching 2.5 quintillion (yeah, that’s 18 zeros!) bytes of data produced in the world every day And yet not all organizations can say they are seeing the benefits of this rampant deluge of data. So, how do you know your data is useful? One of the first questions organizations should be asking is how useful all of this data is. How? Let’s start by the basics:
**Are you collecting data? **
It all depends on the industry you are in and the types of variables you believe influence your business. A simple first step would be to ask yourself if you are collecting and measuring information in a systematic way. In other words, do you make an effort to capture and store information?
There are multiple ways to collect information. You might use online forms, surveys or a CRM software to collect and manage information about your users or customers. You might also use performance measurement systems or analytics tools to collect data about your business operations or how your users or customers behave and interact with your online and offline presence such as a website or a store. What is important to know is that the collection tools you use will be determined by the types of data you believe you need to collect.
**What types of data are you collecting? **
Data can generally be boiled down to two types: static and dynamic. As the names suggest, static data implies that the state of an entity such as a user or a machine doesn’t change or rarely changes. Examples of static data would include demographic information such as a customer’s date of birth, an email address or a field such as gender.
Dynamic data, on the other hand, represents information that is in flux, describing interactions between an entity and a system. An example of dynamic data can be a click on a webpage, a payment for a purchase or a trigger of an automated task.
While static data describes a state, dynamic data allows you to both explain behavior, but also to predict it. It is important that static and dynamic data be coherent, or similar across all entities and the system with which they interact in order to make these entities comparable..
**Is your data coherent? **
Coherence simply means that the fields, events and properties that are collected across comparable entities (users and users, customers and customers, machines and machines, etc.) and their participation in the system you are tracking is similar.
For example, if I collect the physical address of 25% of my company’s customers, the email addresses for 30% of my customers and no information for 45%, then segmenting, or grouping customers into distinct clusters becomes difficult.
Similarly, if I would like to predict inventory for screws, and I collect data from a first machine each time it adds a screw to a piece of metal, and from a second machine which reports that it added a screw as well as the type of screw, then we have imperfect data across both machines making it difficult to make assumptions about all the inventory of screws used by machines and how it could be optimized.
To sum up, if you are already collecting static and dynamic data, and your data is homogeneous, you are ready to generate better insights, and even create models to make predictions related to your customers’ behavior or your product’s performance.
In our next article, we’ll talk about tools for collecting static and dynamic data, we’ll explain ways to analyze the coherence of your data and the type of insights you can generate by having coherent dynamic data. To get notified when our next article is out, please sign up to our newsletter.