What is Big Data: The Four Q’s
In 2012, the popular retail store Target famously analyzed data on consumers’ shopping habits to determine pregnancy. Target’s methodical use of data analytics and pattern identification allowed them to send advertisements and incentivize shopping for expectant mothers, effectively increasing their sales. In the past, details about the daily lives of individuals or the environment were often not cared for and thought to be useless. As technology developed more and more, we were able to harness the power of data and make high-accuracy predictions based on patterns — compared to logical reasoning in our past world. To use Big Data, there are a couple of criteria that have to be met:
Quantity:
If there isn’t sufficient data, the margin for error increases and conclusions are rather unreliable. At the core of Big Data is pattern recognition — and without enough data, it’s impossible to generate accurate predictions. With larger amounts of data, however, sample size trumps randomness and the errors decrease. On the other hand, the amount of accessible data isn’t the only important thing to look out for.
Quality:
When it comes to the quality of the data, validity, and value are what you need to look out for. To create meaningful information out of data, your data must be valid and accurate. If the data that was used does not have high accuracy, any predictions or assumptions based on the data would be useless. In addition, the data that is used must also have value — if the data is not relevant, then it would similarly be rendered useless. After all, if Target was looking for data about pregnancies, large quantities of highly accurate data on vodka and hot pockets would probably not be beneficial.
Quilt:
A quilt is a bed cover that consists of many different pieces of fabric sewn together. Like a quilt, your data must have range and variety. Going back to the Target example, although mass amounts of purchasing data on one item could show some information about the consumers, data on a larger variety of items would reveal the bigger picture and allow for better predictions. Having variability in your data also reduces the amount of errors — if all of your data was dependent on a single source, your analyses are all built upon a single point of failure.
Quickness:
Lastly, one of the biggest benefits of using Big Data is that Big Data can often predict events or outcomes in real-time compared to other methods. Because of this, one very important criterion for Big Data analytics is quickness, or how fast data can be generated and used. Being able to generate data efficiently and timely increases the value and accuracy of the information extracted from the data.
Target’s pregnancy prediction model was revolutionary in the world of retail stores as it explored a new method to gather information about consumers and increase revenue. Target’s model used a lot (quantity) of real-time information from shops (quickness), a variety of information from different sources about different products (quilt), high-accuracy and relevant purchasing information (quality) to make a highly accurate prediction on whether or not the consumer is pregnant. Through Big Data analytics, they were able to harness the power of data in our present world to their advantage and increase their revenue.
References
- Cooper, B. (2019, April 19). Target talks digital signage, using big data to deliver a personalized experience. www.retailcustomerexperience.com. https://www.retailcustomerexperience.com/articles/target-talks-digital-signage-using-big-data-to-deliver-a-personalized-experience/
