If you've been in marketing for awhile, you're as used to working with data as you are combing your hair or brushing your teeth. So the advent of Big Data probably has you confused. How, exactly, is big data different from just having lots of data? When do the traditional tools and techniques become overloaded, causing you to need to take on better tools for data storage, cleansing, and analytics?
Depending on whom you ask, there are between three and seven factors that separate 'big data' from 'lots of data'. We'll discuss all of them here, though it's important to note that big data is primarily characterized by the first three. It differs in terms of volume, velocity, and variety. The others, however, serve to highlight how working with big data is different from traditional marketing analytics.
Volume: There's Lots and Lots and LOTS of Big Data
Marketing data of old could be measured in gigabytes. Big data is measured in zettabytes or yottabytes. In comparison:
A gigabyte = 1,000,000,000 bytes
A terabyte 1,000,000,000,000 bytes
A petabyte = 1,000,000,000,000,000 bytes
An exabyte 1,000,000,000,000,000,000 bytes
A zettabyte = 1,000,000,000,000,000,000,000 bytes
A yottabyte = 1,000,000,000,000,000,000,000,000 bytes
If you're suddenly feeling a little insignificant, you aren't alone. Nobody really comprehends that amount of data. But today's computers typically have storage capacities expressed in gigabytes or terabytes. Once you start bumping up against data sets too big to fit on a single machine (usually when you reach the high terabyte to low petabyte volume), you need to switch to big data tools.
Velocity: Big Data is Growing at Breakneck Speeds
The term 'big data' is misleading. The amount of data isn't the only factor that makes it different. Big data is growing at remarkable rates. In fact, most organizations are dealing with data sets that double every couple of years. When you've got gigabytes of data, that's not a big deal. But when you start working with petabytes and exabytes of data, that amount of scalability is impossible to achieve without specialized tools for managing big data.
Variety: Big Data Isn't Homogeneous
Here's where the biggest obstacles come into play. Big data is widely varied in terms of what kind of data it is and what sources it comes from. By nature, it comes from many disparate sources, including legacy systems, proprietary software systems, mobile devices, IoT devices, websites, social media, and others. Most of it is unstructured, meaning it doesn't fit well into the traditional databases. It also comes in a number of different formats, many of which are proprietary or manufacturer-specific, meaning it can't be read by just any old tools.
Variability: Big Data is Highly Varied
This isn't the same thing as variety. The variability of big data is like the variability of your favorite vineyard's wine. Variety is when your favorite vineyard makes four different kinds of wine: Merlot, Cabernet Sauvignon, Cabernet Franc, and Shiraz. Variability is when the Merlot ranges from bright red to deep burgundy in color. Data isn't just varied in the sense that you have social data, website data, and data from your legacy software system. It's variable, meaning the data changes continually. Those changes are how you track trends and phenomenon within the industry.
Veracity: Big Data Has to be Accurate
Though many consider this to be a feature of big data, it's not that different than cleansing traditional marketing databases, except in terms of scope. If you don't do it, the data yields inaccurate results. But big data requires specialized tools and processes for assuring veracity because the old tools just can't handle the volume, variety, and variability of big data.
Visualization: Big Data Has to Be Explained to the Masses
Big data analytics is essentially useless to the non-mathematician when it comes out in its raw format. It takes a professional at data visualization to put the data into presentable form so that decision makers like marketers, executives, and customers can make informed decisions based on the analytical findings.
Value: Big Data Has to Deliver an ROI
In theory, the value of the results of big data analytics is greater than the sum of its parts. For instance, with traditional marketing data, you had information like, "X percentage of customers prefer to buy through a website" or "X many customers prefer to pay by credit card". Since it combines a wide variety of different data sets, it can yield much deeper insight. For example, it can be predictive: "the trend next quarter will shift away from on-premises software to cloud-based software" or "by 2020, most businesses will use IT outsourcing vendors". These results make big data more valuable than analytics of old.
ReachForce helps marketers increase revenue contribution by solving some of their toughest data management problems. We understand the challenges of results-driven marketers and provide solutions to make initiatives like marketing automation, personalization and predictive marketing better. Whether you have an acute pain to solve today or prefer to grow your capabilities over time, ReachForce can unify, clean and enrich prospect and customer lifecycle data in your business, and do it at your own pace.