News & Media

Cyber Monday and the Science of Data

Black Friday, one of the biggest global shopping days of the year and part of the whole Thanksgiving weekend is over – so it must be time for – Cyber Monday! The term was first coined by marketing companies in 2005 to try and persuade people to shop online, after a majority of retailers reported a substantial increase in sales on the Monday after Thanksgiving. Much of it, apparently, is due to Americans returning to work after Thanksgiving festivities and continuing their holiday gift shopping, away from the prying eyes of family members.

Cyber Monday features great bargains for electrical gadgets and homeware. Surprisingly, it is also pretty big for fashion retail too. The good thing is, unlike the sometimes violent scenes that dominate Black Friday, there’s no risk of physical injury with Cyber Monday, as because it is exclusively online, it can all be taken part in from the comfort of your own sofa!

Most major retailers, such as Amazon, Argos, Tesco and Asda will have started their online sales at midnight but as it is strictly a 24-hour event, the clock is ticking…

Online sales have increased year on year from $610 million in 2006 to a whopping $2280 million in 2015! Other countries around the world keep joining in each year too!

It’s all in the data

But have you ever wondered how online retailers can actually afford to sell their products for such low prices? Well, the answer lies in the data. For many years, retailers have used sales data historically to decide whether or not to drop prices – however, the rising popularity of data science and the ability to track every click and page view means retailers can now predict, with some accuracy what will sell best. Amazon, for example, uses complicated algorithms that mean discounted products aren’t just random but are specifically derived to drive the highest number of sales – this occurs by accurately predicting which customers are most likely to purchase which items at what price and when.

Every click you make on a major retailer’s website is tracked to see which product you were looking at, what type of device you were using, whether or not you bought the item, what other items you looked at etc etc etc. This all powers price optimisation.

Building a data model to form a pricing strategy involves a combination of macroeconomic indicators, purchase behaviour histories, popularity and profit margins. These models then need to continuously adapt to the constant incoming streams of data.

Data scientists are now in such high demand for this type of work, which can drastically boost retailers’ operating margins, they can command high salaries now too.

To find out more about becoming a Data Scientist, have a read of this career profile!