With the leverage of carefully tailored SEM strategy, AI and automation of SEM assets, Kiwi managed to significantly increase the number of bookings during COVID-19 pandemic. The presented approach has made SEM one of the most important marketing channels for Kiwi.
Kiwi.com, founded in 2012 and located in Brno (Czech Republic), is an online travel agency and metasearch engine that specializes in offering affordable flights, hotels, and transportation options to travelers.
Kiwi wanted us to maximize the number of bookings while achieving profitability through Google Paid Search channel. The “only” challenge was that the commercial flight industry was one of the most hit during COVID-19.
Maximize number of bookings
COVID-19 pandemic impacted people’s interest to search flight related terms on Google in the same way it impacted the commercial flight industry. In order to achieve profitability with high booking volume, a key step was to scale campaigns.
How was it done?
As there are numerous possibilities for flight related search, the first key step was to distinguish between possible ways people might search. Some people will search for a specific route (e.g. “flight from Split to London”), some will search for a specific airline option (e.g. Easyjet flight), while some will search only for “cheap flights”… Basically, we strive to cover the full funnel. Since we had a limited search environment we focused on all worldwide markets; those in which users were familiar with Kiwi, but also new unexplored markets.
The second key step was automation. Imagine researching and building keywords / ad groups / ads for a huge number of routes in the world in all worldwide languages (a never ending task for tens of people). In order to optimize a process for building routes, we utilized a custom tailored automation tool which could create tens of thousands of routes in a day.
In order to optimize resources carefully and advertise routes that people are actually searching, several methodologies for route selection have been created and tested with the client.
Since we had few millions of keywords, focusing on all markets worldwide, the third key step was to maximize profitable booking volume by using Google machine learning. In order to be as precise as possible at providing profitability, instead of focusing on total booking value (which is usually done by SEM managers), we were optimizing campaigns towards net revenue.
The process described above has become the foundation for Google paid search channel and it has set up the channel to become one of the Kiwi’s top performing marketing channels, as well as a source of acquiring new users and retaining the existing ones.