Customer Analytics Case Study: eBay

We take a look at how eBay utilizes customer data for success


Online auction platform eBay is one of the internet’s most enduring success stories, helping to spawn a number of copycats and create a total upheaval in the way we shop. You can buy anything, from air guitars to Justin Timberlake’s uneaten french toast. Much of this success has been driven by its ability to harness the wealth of data it accumulates about customers every second and convert it into actionable insights that drive even greater sales.

With 162 million active buyers globally and over 800 million active listings, there is obviously a significant amount of transaction history there for analysis. eBay also has an advantage over other online retailers because it has even more data points throughout the customer journey, with customers able to watch items they may want to sell, bidding on items, following, collecting, saving for later, all of which serve as signals that eBay can use to better understand customer’s habits and enable better targeting. In all, eBay processes some 100 petabytes of data daily.

Speaking at last year’s Customer Analytics Innovation Summit, Zoher Karu, VP of Global Customer Optimization & Data at eBay discussed how the data is leveraged to create a complete picture of the customer, such as their attitudes (whether they’re price sensitive, for example) their behaviors (how often they visit, they whether they buy on auction or buy it now), demographics and interests, and how valuable the customers are. This information is then leveraged to personalize marketing materials so people are getting what they want, and throughout the journey to purchase to help encourage a sale.

Abandoned carts are easy to act on simply by sending email gently reminding the customer that it’s in there. The harder part is telling people things that will inspire them to buy using eBay.

Using the data, eBay can identify interesting events that indicate what customers are likely to care about and personalize search results and deals that will appear to them and provoke them to shop on eBay. In his presentation, Zoher used the example of shoes. Obviously, it’s easy to establish if someone is interested in a product based on their search history, but even if someone hasn’t looked up shoe yet, if they have browsing histories similar to other people who have bought shoes, eBay will be given to understand that that user is also likely to be interested in shows and promote to them accordingly. They can also gain understand what type of shoes you might like, brands, price preference based on metrics like household income and buying history, to direct users to the shoes that they will likely want. Obviously, presenting them with a 100 items in an email is unlikely to yield much of a result, but if they are presented with a couple of options that they are likely to buy, it is far more probable that they will purchase.

Personalization is just one way that eBay uses the data at its disposal, and analytics is used at every level and scale. eBay also examines things like sentiment analysis on social media to gauge customer appreciation, and to drive up its price listings. A/B tests are common for making sense of user responses to the site or feature changes, and policy changes. Everybody therefore has to understand data at some level as it is constantly advancing its capabilities, and they provide constant training. The online marketplace last month acquired big data analytics firm Expertmaker, which specializes in helping clients with their optimization, prediction, and personalization initiatives by combining machine learning and analytics. Most recently, Expertmaker helped eBay structure its data, to better collect insights into supply and demand.

A recent Forbes Insights global survey of 357 executives of large organizations found that 42% of executives worldwide predict data analytics will result in an improved customer experience across the board. For three in 10 enterprises, data-driven CX is already delivering a significant shift in elevating customer experiences, and only 9% noticed no changes in customer experience. The benefits are many, and following eBay’s example will not only be advisers in years to come, it will be necessary to remain viable as a business.


Image: Denys Prykhodov /

University lecture small

Read next:

How Are Higher Education Institutions Using Analytics?