As the saying goes, “You can’t manage what you don’t measure.” Attributed to both Peter Drucker and W. Edwards Deming, this sums up the theory behind big data analytics: to run your company successfully, you need to pull insights from the data your company accumulates. But what does big data actually mean?

First incorporated into the Oxford English Dictionary in 2013, big data is all the rage with marketers and data analyst firms. You might be asking, “Is big data just a fancy name for analytics?” And while they are similar, in that big data seeks to pull intelligence from data and apply it to a business advantage, there are three key differences that crop up in the commonly repeated definition: volume, variety, and velocity.

1. Volume

Big data is comprised of big volume. It is estimated that there will be at least 110.3 exabytes of data created per month in 2016, compared to roughly 75 exabytes per month in 2012. More data crosses the internet every second than was stored on the entire internet just 20 years ago.

The benefits gained from the ability to work with such large amounts of information are obvious: the forecasting ability of a company is more effective when run against 3,000 sets of data rather than 30. As the database grows, the applications and architecture needed to support the data should be reexamined and evaluated. Often the same data can be re-evaluated from other angles, with new intelligence offering new insights.

2. Variety

Another important feature of big data is diversity within the types of data collected. Big data takes the form of messages, updates, and images posted to social networks; GPS signals from phones; a raw feed from sensors, and more. Although the large range of data sources gives an equally impressive range of conclusions, this also means the data rarely falls into neat, sortable genres, ready for processing or integration into an application.

Concurrently, the main sources of this data are themselves relatively new: Facebook launched in 2004, Twitter in 2006. The same holds for smartphones and other mobile devices supplying this data: they are so ubiquitous, it is hard to remember that the iPhone was only launched in 2007. Although the databases frequently used to store this information is, of recent years, ill-suited and outdated, the steadily declining costs associated with computing (memory, storage, processing, etc.) mean that data-intensive approaches are becoming more economical.

3. Velocity

The speed at which data is collected and analyzed is frequently more important than the volume. Real-time or nearly real-time information can make the difference when fighting for the edge over a competitor. The internet and mobile era means that the ways we deliver and consume products and services is increasingly instrumented, generating a flow back to the provider. Those who are able to quickly utilize that information, online or by accessing customer’s mobile phones with geolocated data, gain competitive advantage – especially in ecommerce.

It is vital to note that equally important to the amount of data accessed are the insights drawn from the set. For example, in ecommerce sites, the four main uses of big data analytics are:


Increase efficiency by alerting you to merchandising efforts that are not effective, and products that are not selling. For example, you may have a shirt in five colors, but only two colors are being purchased.


Increase conversions by tracking down and analyzing successful sales transactions.

Increased purchases:

Increase the amount of purchased items by suggesting similar and complimentary products to ones your customer is purchasing, in real time.

Inventory management:

Manage your inventory more efficiently by easily discovering and eliminating items that are not selling, and increasing inventory of items that are flying off the shelves.


In the end, the far-reaching nature of big data analytics is both messy and specific. While the tools you need are there, remember that without a question or a problem to solve, you will figure out nothing – and wind up with a whole lot of random information. Pick a real business problem first, then use analytics to guide your implementation. Only with direction can you narrow down all of that big data!