Designing the Omio Landing Page experience
I led a small team through a full product development cycle, including research, synthesis, Ideation and implementation phases. The result was a substantial increase in the number of bookings made per visit.
Problem:
After making a search on Google, a persons first experience with Omio is a landing page. These pages were primarily designed to drive users into making an Omio search. We have observed that people arriving from Google have very different needs depending on the stage of their search, or type of travel. We believe the generic nature of the experience is causing some people to abandon the page and not proceed in a booking flow.
Additionally the generic nature of the landing pages (thin content and minimal unique content above the fold) is viewed negatively by Google meaning there is potential for us to improve our rankings on Google.
Hypothesis:
We believe that personalising the landing page experience based on a users intent will help people find the content they are looking for more easily and improve the perceived value of Omio. We will know we’re right when we see an increase in the number of bookings made through a landing page flow.
Discovery research:
I began the project running a deep discovery phase. I worked with the research team to run field studies, visiting peoples homes and experiencing booking travel with them first hand. I worked with our Data science and SEO teams to analyse existing search behaviour on Omio. And I ran a competitor study to understand where we stood in the market and identify any innovation that was happening.
User Research
Key finding: People are looking for information. When performing a travel search, users will look for real data as an indicator of a website’s potential usefulness, truthfulness, and comprehensiveness.
Keyword analysis
Key finding: Through a historical analysis of keywords we were able to identify significant groups of semantic themes. In short, a lot of people made searches of similar flavours from which we could derive their intent and therefore their need.
Competitor analysis:
Key finding: There is a trend for including instant search results on landing pages. This above the fold content seems to be recognised by Google as the most relevant content and will therefore have a positive effect on SEO.
Vision
We focussed our efforts on three main intent groups we called Cheap, Fast and Soon. For example under Cheap we grouped keywords that included things like Price, Discount, Budget, Deal, Economy and Sale.
Now that we had identified these major intent buckets we designed component features for each. The idea was, to in real time, update the landing page content to match the query used. When a user clicks on an Omio SEM link we have immediate information about what exactly they’ve typed into Google. For example; if a user searches “Cheap train Berlin to Hamburg” we know they are price sensitive and should show them relevant content.
The features we designed provided immediate results filtered to match the identified intent. For the above example a user would see the five cheapest trains from Berlin to Hamburg alongside information about the best time to buy.
Implementation
We rolled out the components and A/B tested the performance. What we saw was a massive 3.27% increase in the number of bookings per visit and a positive effect on page quality score. A clear success.
An additional effect of note was a reduction in some of our upper funnel metrics. Which at first seemed counter intuitive. More people are booking tickets but less people are progressing from the landing page into a booking search flow. What we went on to learn was that by exposing results sooner in the flow, users with a low intent to buy (who might have been simply searching for a rough price and would never have purchased a ticket in that session) dropped a step sooner. We had in fact provided a quicker and easier way for these people to reach results.
We decided to extend the use of the components to organic landing pages. Here we cannot react directly to the search term someone has made in Google and therefore do not see the same changes to conversion. However these components mean we are able to bring unique content onto more than 90% of our pages. Which is a major benefit for SEO.
The release of these features are a tentative first step into personalisation of the Omio experience. They talk to only a small segment of our audience and only scratch the surface of what’s possible. They are MVP’s we will learn from and iterate on.
Their relative success has given us a glimpse of the future. It proves that tailoring the Omio experience to a user intent can provide people with a better experience and improve our core business KPI’s.