It’s true that Big Data is somewhat of a catchall term. So many different applications and organizations can be lumped underneath the Big Data umbrella that it can be confusing sometimes to know what exactly we’re talking about when we use the phrase.
Contrary to the belief in some quarters, it means a lot more than just “a lot of data.” At it’s simplest form, Big Data has three essential qualities: Volume, Velocity and Variety.
Let’s break down the Three V’s one by one.
This refers to the rate of the data flow — the speed at which new data is being generated. The digital age has made an extremely high rate of data accumulation possible. For example, Twitter recently announced that the 2015 Super Bowl saw rates as high as 22 thousand tweets per second. And on the day after Thanksgiving 2014, Amazon had their busiest Black Friday ever, with almost 4,000 transactions per minute. This high velocity can be a boon to companies, giving them the opportunity to react in real time; think Amazon, changing pricing on the fly depending on what products are doing well, or an airline adjusting ticket prices based on demand gauged by search traffic. These are opportunities that a company with a lower data velocity might not be able to take advantage of. On the flip side, a very real danger exists that an organization will be swamped by a data flow rate that it simply can’t process or make sense of. Successful enterprises will develop the ability to process, interpret and react dynamically to data, no matter the velocity.
As we established earlier in this post, data is proliferating at ever-increasing rates. Advances in hardware make it easier and cheaper to store massive amounts of data indefinitely. In the past, running out of storage space was a constant concern. Today, with massive storage devices and cloud services widely available and relatively inexpensive, organizations are finding new concerns. Like, when you’re able to save all your data, how do you determine what’s valuable and what’s extraneous? As volume continues to rise, advanced analytics that help sort through the massive data volumes will become increasingly important.
Feeding the constant increase in volume is the expanding variety of data sources. The rise of social media in particular has generated billions of new pieces of data in the last decade, in an increasing assortment of formats: text, photos, audio, video and so on. The variety is showing no signs of slowing down, either, with new forms of data collection constantly cropping up. Most technologists think that the next frontier is going to be the so-called “Internet of Things,” where everyday objects are connected to the cloud and generating data points. They envision a future where your car automatically informs your house that you’re on the way home, and the thermostat changes itself to a comfortable temperature, the stereo plays your favorite song and the oven heats itself up to cook your dinner. That entire interaction will be logged somewhere, and become part of the Big Data picture.
Case Study: Jamba Juice
An example of a corporation leveraging the Three V’s for success can be seen with Jamba Juice, a U.S. retail chain selling smoothies and fruit drinks. Recognizing that store traffic could change drastically day-to-day and even hour-to-hour, Jamba Juice sought out a Big Data solution to make staffing more efficient. Ideally, each store would have the perfect number of employees to handle the amount of customers. Too many employees on duty could mean wasted money on extraneous salary; too few employees could mean long lines, frustrated customers and potential lost sales.
The software solution they came up with was a sophisticated system that takes several variables into account. It combines data points such as employee availability and performance, historical sales figures, local traffic patterns, and even weather forecasts (they tend to sell more smoothies when it’s hot outside) and runs them through an algorithm to give specific, real time recommendations for staffing.
The system works for them and perfectly illustrates the application of the Three V’s.
The volume of information is huge, with sales data tracking calendar-based trends spanning the 25 year history of the company. It’s enough volume of data to reliably recognize patterns that might otherwise go undetected.
The velocity is also taken into account, with the system able to deduce patterns with enough timeliness that they can still be applied. If the lunch rush, for example, is slower than usual, the system will detect it early and can recommend sending staffers home early.
And variety is there as well, taking into account seemingly disparate data points like average temperature for that time of year, or which store employee has the highest average sales.
By properly leveraging the 3 V’s of Big Data, Jamba Juice management is able to save upwards of 5% a year on personnel costs, a savings of millions of dollars for a chain with more than 800 locations and 13,000 employees.
Looking for more practical information on one of the industry’s biggest buzzwords? Download GA’s Enterprise Guide to Big Data to learn more.