Customize and book tour packages on an interactive website
Travel
Web Design
Star Travel Group is an experienced Russian tour operator specializing in guided group tours for foreign travelers. As of 2020, 96% of its revenue came from B2B, of which over 80% was focused on the Middle East market. Star Travel Group was looking to expand its market internationally and increase its B2C revenue. Thus, it wished to redesign its website to appeal directly to international customers and sell its travel products.
Star Travel Group hosts 30,000 – 50,000 inbound tourist seach year. It hires 30-50 full-time employees and has up to 250 employees in the travel season. It makes about 10 million in revenue each year, mostly by booking hotel rooms in bulk in advance at very low prices and selling them to tourists within tour packages. The group specializes in 4-star and 5-star hotels, which are still affordable for European and North American audiences.
Our team redesigned the 2C end of the website, specifically the homepage and pages related to displaying tour packages. We also revamped the website’s branding and information architecture. At the end of our project, we delivered a high-fidelity prototype of our website redesign.
time
Apr 2020 - Jun 2020
type
MS HCDE Course Project
my roles
Lead: Design, Prototyping & Content Strategy of the Customization Function
Participate: Web Analytics, Competitive Analysis, User Flow Analysis, Persona, Card Sorting, Information Architecture, Usability Testing
Teammates
UX Design & Research: Sooa Chung, Summer Wang, Peggy Zhang, Weixi Zhang
Sponsor
Pedram Rafie on behalf of Star Travel Group
We employed various research methods to answer these questions:
We analyzed the server log and the current website flow and performed usability tests. We identified several key problems:
We reviewed popular travel sites like Airbnb, Expedia, and TripAdvisor, as well as travel agency websites specializing in Russian tours. Our main takeaways include:
To understand the needs and wants of our potential customers, we conducted 3 semi-structured interviews, as well as a survey of 105 participants, 92 of whom had booked tour packages before. We probed into participants’ feelings towards tour packages, their booking experiences, and their customization preferences. We found that participants:
“Since trips are rare times of leisure for me, it is important for me to enjoy the trip and not worry about anything else.”
“I come from a big family so when we plan for a trip it is easier if we do package tours and we don’t have to do research or negotiate on things that everyone wants to do.”
“I normally can’t find the exact tour package that includes all the things that I want to do... I have to follow the schedule... or I might lose the value that I paid for.”
We synthesized our research findings to identify our potential customers and their needs, and thus focus our redesign efforts.
We identified three personas in terms of travel planning preferences:
While Larry is our ideal customer, Penny will also buy a tour package if it meets all her preferences or is customizable. We consulted our sponsor and learned that Start Travel tours were already customizable through informal communication with tour guides. This was a potential selling point but was not mentioned on the current website. Therefore, we decided to cater to both Larry’s and Penny’s needs in our redesign, and articulate this customization feature as a website function.
Thus, our guiding question became: how can we redesign the Star Travel website to appeal to customers both with and without customization needs?
We conducted card sorting to redesign the Star Travel website’s information architecture and focused our design on the home page, the page displaying tour search results, and the individual tour pages.
We employed an agile process as we moved from low-fidelity to high-fidelity designs, performing short iterations of usability testing with 7 participants.
We performed a branding survey with our sponsor and crafted our new branding and design system to give the impression of being trustworthy, professional, personalized, inviting, and fun.
Larry and Peggy’s travel needs are distinctly different. Allowing every user to fully customize their tour packages would satisfy Peggy’s needs, but would put too much mental strain on Larry, and would also raise Star Travel’s operational cost significantly. Therefore, we defined two different types of tour packages: group tours and private tours. Customers can hover over the info icon in the search bar to view the differences, and can select either or both when they search for a tour package.
There are many elements within a tour package that can be potentially customized. As we learned from our user research, users prioritized customerizing hotels, attractions, and transportation. We decided against customizing tour length and structure, because it involves too much effort and users can simply choose another package. We also decided against customizing meals because they are relatively unimportant and easy to negotiate with tour guides on the spot.
Since hotels and transportation (optional airport pick-up and drop-off services, and transfer between cities by either plane or train) are already customizable on the current website, and the mental load involved is relatively light, we made them customizable for both group and private tours.
Customization of attractions is the standout feature of private tours. We explored several ways to achieve this in usability testing. We adopted the solution that struck the best balance between personalization and complexity: on any given day within the tour, users can remove existing attractions and add new ones from a list of available attractions with predetermined time lengths, as long as the total time including transit is within a reasonable range. The sequencing of attractions and meals will be automatically optimized by the system to reduce the mental load for users, and also to allow for some adjustments by tour guides.
Our challenge was to present adequate information in a clear, concise manner to aid users in making decisions, but also avoid overwhelming the user with too much information. Our solution was to divide the tour page into two sections. First is the overview section, including a brief description of the tour, as well as a list of services included and not included in the tour.
Second is the itinerary section, including a fixed information bar on top that updates when users scroll the itinerary or make changes to it. The items within the itinerary are displayed both in a list on the left and on a map on the right. As the user scrolls the list, the map view changes to display all items within the visible list section (e.g. all hotel options in a city or all attractions in a day).
Whenever the user clicks on a hotel or an attraction, either in the list or on the map, a card moves in to display the most relevant information, including outgoing links when necessary.
In a private tour, users can toggle on customization. This changes the UI of the list, allowing users to remove existing attractions or add new ones from a list.
Our high-fidelity prototype was well-received by usability testing participants and our sponsor, who praised the visual style and the interactive display of detailed information, especially the map feature. You can check out our prototype here:
We are very grateful to have Pedram, an experienced UX designer, as our point of contact with the sponsor. He was very supportive throughout this project and encouraged us to explore creative design solutions without worrying about the difficulties of building the website. Still, looking back, I wish we could have operated under a better understanding of Star Travel’s business models and expectations for their website, so that our redesign could have a better chance of being shipped in the future.
During our redesign, we invested most of our effort into building the customization feature, and group tours turned out to be a simplified version of private tours. This assumption is not well tested, since the usability testing participants we were able to recruit leaned more towards Peggy. We would like to amend this given more time and resources.
