Experiment: How AirAsia used machine learning to connect more travelers to flights

Download
March 2020
Experiments_banner.png

What we set out to test

Can brands increase discoverability and drive online sales by adding automation to their search campaigns?

The background

AirAsia, APAC’s biggest low-cost airline, operates flights to over 150 destinations across more than 20 countries. The brand traditionally used manual search ads to connect with prospective travelers. But as AirAsia’s network grew to exceed 5,000 routes, the brand found that manually fine-tuning its keyword strategy was sapping its resources — especially while it was developing ads in nine different languages.

The airline knew finding a better way to create relevant, timely search ads was key to establishing a larger presence in markets like India, where brand awareness was relatively low. That’s why AirAsia decided to find out whether automated solutions would help it show up and capture travelers’ attention in the right moments.

How we set the experiment up

AirAsia used Dynamic Search Ads, a format fueled by machine learning, to automatically develop tailored search ads and lead travelers to relevant landing pages based on their searches. For instance, people who searched for “ccu to del flight” were served an ad promoting affordable flights from Kolkata to New Delhi. Once they clicked on the ads, travelers were immediately directed to the airline’s booking page.

TwG_APAC_AirAsia_Nugget1.png

The brand used the Drafts and Experiments tool to see how complementing its existing manual search campaigns with these automated ads would impact performance in Bangalore and Kolkata.

  • Control group: Manual generic search ads* for four weeks
  • Test group: Manual generic search ads* and automated Dynamic Search Ads for four weeks

*Standard generic keywords like destinations and cities were used for AirAsia’s manual search campaign.

What we learned

Supplementing manual search campaigns with Dynamic Search Ads can help brands win both consideration and sales. AirAsia found that adding machine learning to its search strategy helped the brand catch travelers’ attention in the right moments with personalized ad copy.

AirAsia drove 16% more traffic to its booking site after complementing its manual search ads with Dynamic Search Ads and discovered that 50% more of its site users were immediately directed to relevant landing pages. By making it easier and faster for travelers to find the right flights, the airline was able to drive 21% more bookings and boost 11% more revenue.

TWG_APAC_ExperimentGoogle_AirAsia_Nugget1.png

After seeing the impact of automated solutions firsthand, AirAsia decided to incorporate machine learning across its search campaigns in six other markets. Moving forward, the airline plans to continue experimenting with automation to manage its inventory and make it easy for travelers to find relevant hotels and activities on its travel and lifestyle platform.

“Manually choosing and bidding on the right keywords was taking up more and more of our team’s time as the number of flights in our network grew. With Dynamic Search Ads, we can make sure we’re catching travelers’ eyes at the right place at the right time while saving our team valuable time and resources.” – Rohit Sridhara, Head of Search and Programmatic, AirAsia

This case study is part of the Experiment with Google Ads program.

Experiment: How Careem lifted conversions with automated display ads