Redefining hotel content for a digital audience

Thomas Cook's package content had always been brochure led, and was not optimised for online customers.

The tone of voice was "salesy" and it was delivered in long form blocks of text. It was no surprise then that the most common customer enquiries on the holiday package page, either via live chat or the call centre, were content related. Most of the time this information was there, just undiscoverable. We wanted to ensure that user needs were being met by shaping a new content architecture.

Building an Understanding

Initially we focused on outlining our primary design goals; create a digital customer led content structure for multiple markets, improve discoverability to reduced contact centre queries and improve on page metrics to drive conversion. This allowed us to outline a process for the project, and measure our success throughout.

Our first step was to run some card sorting, tree testing and multi-market comparison to outline and validate a base structure. We recruited 50 participants for an open card sort and compared the outcome to the proposed structure that the content team had been working on. After tweaks were made a second tree test was conducted with 120 participants. 50% saw the old structure and the other half the new.

Speed of completion and directness significantly improved with the new structure. After making some further additions based on the findings a comparative test was conducted. Participants were shown the old and new content in context on a hotel page. Only the content was changed, no new design elements were introduced. The preference was for the new content, with the detractors stating that the longer page was their only issue; a problem that a new design could solve.

Finally we wanted to validate the structure across all key markets; Belgium, Netherlands and Germany. We worked with the content team to source translations, and the customer experience teams in each market supported with the analysis. Through this testing we were able to highlight market specific improvements that we needed to make, particularly in Germany.

Forging a Design

It was important to us to see how our current page design was seen compared to our competition at the time; TUI, Expedia & Virgin Holidays. We used the same hotel across all sites and asked users to navigate the page and share their thoughts on the presentation and content of each. From this research it was clear we were well behind.

We also compared the current content with the new structure on our existing page design. 80% of participants preferred the new content, even though there were no design changes.

With the goal of improving the discoverability and scalability of the hotel page design, we started creating variations for mobile. Whilst a tile design was more visually appealing in reviews, an accordion option was the chosen path due to the variance in number of sections, the content within, and unnecessary modal interactions required with the other design.

As this could not be A/B tested, I wanted to get some qualitative feedback before launching on a limited number of hotels. I created an HTML prototype to ensure the interactions were robust and representative of the final product. It was also unbranded to try to avoid any bias. I ran a test with 10 participants, asking them about the design in general and the trustworthyness of the content. All participants liked the design and interactions and 90% trusted the content itself.

Going Live

We launched initially with 20 hotels, but steadily increased that number to over 300 as more updated content was produced. We saw great improvements in key metrics when compared to the previous design and content. I also retested against our competitors and saw very positive swings towards our page vs TUI, Virgin and Expedia, with our page becoming the preferred choice in all comparisons.

This new content structure also allowed us to improve other areas of the experience. New filters were added, allowing customers to narrow down by hotel facilities or sporting activities. We also started looking at segmentation personalisation and showing families key selling points in the search results themselves. The wishlist feature received a boost as we were able to use this structured data to help customers compare holidays that they liked. Users with wishlists had a 5x higher conversion rate than those without so it was massively important to add as much value here as possible.

This project was used as template for evaluating our destination content, using the same research and design process, helping us understand what sort of content users wanted to see throughout the journey.