Big data revolution – how big data shapes our online shopping habits

By Ekow Abankwa, Fourth Year, Mechanical Engineering

SciTech dissects Big data’s impact on consumption and how it’s used in medical diagnosis, research, business strategy and, more pertinently, personalisation and advertising.

Whether you're window shopping on Amazon or bargain hunting on eBay, big data algorithms have shaped the online shopping experience. From advertising to personalized feeds, much of what customers see when visiting an online store is determined by big data algorithms and insights from big data analytics. Not only do these algorithms influence our online behavior, but our behavior also influences them, as the data collected from users determines their outputs. As of this year, e-commerce constitutes 19.5% of global retail sales and is expected to surpass $6 trillion in sales. This means that any significant trends in the industry will have a massive impact on the overall economy. This article will explore the basics of big data and its effects on online retail.

But what is Big Data, and how does it differ from regular data? Simply put, the term big data refers to datasets that are too large to be managed through traditional data processing methods.

Big data’s differences from other datasets can be described through a concept often called the 3V’s: volume, variety and velocity. The volume of these datasets is on the scale of terabytes and petabytes; they contain data in various formats from various sources, and the data is generated at a high velocity, meaning that the data is stored and analysed at a high rate.

An empty shopping cart sitting on an open laptop | Pexels / Karolina Grabowska

The advanced techniques used to analyse these large datasets are known as big data analytics. These techniques aim to find patterns and correlations that would be impossible to find through traditional methods. Popular analysis methods include deep learning, which uses artificial intelligence and machine learning to find complex patterns, and predictive analytics, which uses historical data to make predictions.

Today, big data analytics have many uses, ranging from medical diagnosis, research, business strategy and, more pertinently, personalisation and advertising.

Advertisers and vendors collect various data from users, including location data, purchase history, search history and more. Often, this data is used to infer the demography and preferences of the user to create a “profile” to target users more accurately. By analysing the behaviour of users with similar profiles, they can predict future behaviour.

So, how do e-commerce companies use this technology to get ahead of their competitors? In short, companies focus on targeted advertising, personalisation and customer insights.

Targeted advertising enables companies to focus their ad campaigns on users they believe will be more receptive and even tailor the message. As advertisers shift away from using third-party cookies (Google Chrome plans to eliminate third-party cookies in 2024), the predictive capabilities of big data analytics are now used to determine which ads to display to users. In 2022, clothing retailer L.L.Bean utilized IBM's Watson Advertising Accelerator, a system that employs data analytics and AI, to boost online sales during a campaign by 48%.

Another critical strategy is personalisation. As these machine learning algorithms learn more about the user, they compare your activity with trends from existing data to create bespoke recommendations. Amazon claims that its AI recommendation engine was responsible for 35% of its sales in 2018, showing how valuable personalisation can be. Deals and discounts are also customised, so much so that many users expect as much, with Salesforce finding that 56% of customers expect offers to be personalised.

Code projected over a woman | Pexels/ This Is Engineering

Companies can also gain insight into how users feel about their products through analytics. Rather than a traditional feedback form, big data analytics offer an opportunity to better measure customer retention by measuring their engagement more objectively. Deloitte lists the main advantages of data-driven customer insights (over traditional customer research methods) as authenticity, comprehensiveness and timeliness.

So far, this article has painted big data analytics as a utopian technology, allowing companies to be perfectly attuned to their customer’s needs; unsurprisingly, this isn’t the case, and there are obvious drawbacks. The primary concern is customer privacy. Several measures exist to ensure privacy, such as de-identification or encryption, but customers are right to be wary of improper handling as data breaches have occurred in the past.

Companies such as Amazon use big data to power dynamic pricing algorithms that alter product prices based on factors such as stock, demand and competitor pricing. This can allow vendors to offer more competitive prices but can also amount to “price-gouging” or discriminatory pricing.

It’s clear that big data analytics have changed and will continue to change how we shop online. In the future, customers can expect to see an even greater level of tailoring to their shopping experience as advertisers chase greater levels of advertising efficiency. The only remaining question is the extent to which we are comfortable with them going.

Featured image: Unsplash/ https://unsplash.com/photos/a-close-up-of-a-window-with-a-building-in-the-background-fyeOxvYvIyY


To what extent are you comfortable with Big Data using your personal information to show you targeted advertisements?