Expedia

Custom data platforms to scale paid search

Hundreds of millions of keywords.

That’s the volume you’re dealing with when you’re an organisation like Expedia Group. Beyond simply managing campaigns, understanding their ROI can swing revenues and profits in either direction.

Expedia called upon DEPT®’s data team to build a tracking and analytics solution to help their team manage their paid search initiatives.

Scaling paid search

The Expedia Group (expedia.com, hotels.com, carrentals.com, vrbo.com, and more) faced challenges when scaling their paid search efforts. With millions of keywords, it took a lot of work to understand the value of certain keyword groupings, especially if there was an error in attribution during bookings.

They needed insight into which keywords, where, and when contributed a positive ROI and which keywords cost them too much money.

A three-pronged solution

DEPT®/DATA worked with Expedia’s marketing team to understand their requirements before recommending (and implementing) the following solution:

Tracking and analytics: We redesigned a tracking and analytics platform on Expedia.com/Hotels.com’s data platform, producing end-to-end reporting, mixed media modelling (MMM), multi-touch attribution, and a bidding engine based on transaction and LTV-based components.
Event-based bidding: We built an automated event-based keyword bidding solution, including bid enhancements, dynamic pricing, and stop-sell for business events like partner contract changes, inventory fluctuations, static calendars, and dynamic weather events.
Scaled search traffic engine: We built a fully automated SEM portfolio management platform focusing on travel-specific workflows like keyword expansion, landing page selection, and scheduled launches.

Better paid search results

Expedia/Hotels.com renewed its faith in managing its paid search with the new tools, which resulted in effective paid strategies with a clear ROI. With a 100 percent increase in return on ad spend and a 33 percent reduction in overall spending, these data implementation projects paid for themselves within a few months.