As a result of effectively mitigating key user frictions when buying an air ticket online, we increased the visitor-to-sale conversion rate by 7%.
Trimydream.com is a popular air ticket aggregator. It drives revenue from helping users find and book the best deals for air tickets. It has ~1 million monthly sessions (according to SimilarWeb). Our task was to increase the visitor-to-sale conversion rate.
We broke down the analysis into two parts:
1. Analysis of conversion funnel to identify stages that generate significant traffic leaks;
2. Identifying frictions that users experience when buying an air ticket online and that act as conversion barriers.
We found that there was one natural drop off point in the ticket purchasing flow when shopping online – the flight search result listing. This drop off on the flight search result listing is explained by the natural absence of appealing flight options like flight time, price or connection length. But we also found that there was a significant proportion of users who found an acceptable flight option, but refused to buy and left the conversion funnel.
In order to address this finding, we conducted 10 user tests with recruited testers and carried out an extensive interview with the client’s support team.
The results of the research have shown that once a user finds the acceptable flight option, he then goes to other sites or to the website of the airline operating that flight to compare the ticket price. In many cases, the user never returned to the client’s site, even if the price on the airline’s website was not lower. This created a significant challenge for us, as users’ search for the best price was commoditizing the product, meaning that we could only compete on price or brand equity.
We also found that users were hesitating to complete the purchase because they believed that there could be more flight and tickets options available. So they were continuing their search for more options on other sites.
We decided to introduce a new block on the flight search result listing that would mitigate the user frictions and hesitations found in the research. The copy in this block had three assertions:
To validate this hypothesis we conducted an experiment that measured the impact of the newly introduced block with benefits on the visitor-to-sale conversion rate. The experiment had the following characteristics:
Experiment type: A/B test
Traffic split: 50/50
Key metric: visitor-to-sale conversion rate
The alternative variation with benefits has shown a 7% increase in visitor-to-sale conversion rate with a 95% statistical significance of the result.