How one unclear data field on the website can cost $12m (lesson from Expedia)
How to root out hidden profits with analytics…
Online travel firm Expedia has found that data analytics can deliver a multi-million dollar kick to a company’s bottom line.
The company used SAS analytics to identify a single change to a web page that generated an overnight surge in sales, Expedia’s VP of global analytics and optimisation Joe Megibow told the SAS Premier Business Leadership Series conference in Las Vegas last week.
Expedia analysts realised the site needed to be changed after investigating why many customers who clicked the ‘Buy Now’ button on the company’s site did not complete the transaction.
“This is someone who was on our site, found the right location and hotel, put in all their billing and travel information and clicked the ‘Buy Now’ button,” Megibow said.
“As far as leading indicators of purchase intent go, this is as good as it gets and yet we weren’t taking the money.”
Analysts began examining and correlating information about these failed transactions to identify what traits they had in common.
The answer, it turns out, was quite simple: “We had an optional field on the site under ‘Name’, which was ‘Company’,” Megibow said.
“It confused some customers who filled out the ‘Company’ field with their bank name.”
After putting in their bank name, these customers then went on to enter the address of their bank, rather than their home address, in the address field.
“When it came to address verification to process the credit card, it failed because it was not the address of credit card holder,” Megibow said.
“After we realised that we just went onto the site and deleted that field – overnight there was a step function [change], resulting in $12m of profit a year, simply by deleting a field.
“We have found 50 or 60 of these kinds of things by using analytics and paying attention to the customer.”
A major strand of the statistical analysis work that is undertaken by Expedia, Megibow said, is trying to match its customers more precisely to the right hotel.
According to Megibow, for analytics to be truly useful, analysts need to be prepared to help developers build new services and change business processes on the back of the insights their work delivers.
“[Analysts] need to work out ‘What are we contributing to the business? Are we seeing things through and making sure that what we are doing is worth something?’,” he said.