Case Study : Skyscanner redesign
For the third challenge of Ironhack’s prework, I had to jump into the shoes of a world traveler to redesign a flight comparison website. This exercise allowed me to understand the importance of findability, and I will detail here what happens when important features stay mysteriously hidden like a forgotten easter egg.
After a brief benchmark, I chose Skyscanner, which I found easy to use because it was the most intuitive, logical and aesthetic. I quickly reached the results I was looking for and the prices found were better than on other websites.
The challenge was to propose a redesign after conducting user interviews and observing how these users navigated on Skyscanner. I had three days to complete it.
Who were the users?
I interviewed three of my friends.
Pierre is a translator and a big travel and adventure enthusiast.
Laura is a publisher and has been a digital nomad for a while.
Emmanuel is a musician and he has often had the opportunity to travel.
Here is the user type I chose, among those proposed by Ironhack :
« World trotter, Backpacker — 18–38 y/o
You’ve decided to finally go visit that wonder that has been sitting in your dreams for a long time now. Yo don’t have a long time to plan but also you don’t need it. You’ll be traveling in 6 months and are open to almost any possibility but have a budget constraint. You are price-cautious and prefer experiences where you have a chance to meet people and make acquaintances to enjoy the wonder together. You are not picky and you can accommodate the most affordable, adventurous, genuine experience. »
And the wonder of the world I have chosen is the Taj Mahal, because I am fascinated by India.
What I found out from the interviews
What is amusing, when you do the exercise of interviewing your friends, is that you discover, thanks to the details of their choices, personality traits that you did not suspect. You get surprises almost every time.
Emmanuel likes cleanliness and security even though he likes adventure and meeting new people. He was quite surprised to find hotels at 6 euros a night and that made him suspicious. Pierre didn’t necessarily want to book his hotel online because he likes to do real research for that, to find himself in surprising places, so like Laura, he prefers Airbnb or other alternative solutions. Moreover, youth hostels have the disadvantage, which is a big one for digital nomads, that they do not offer a great security for their belongings and therefore their computer. Sleeping on the floor would not even be a problem, but having your computer stolen while you sleep is not an option! The friends interviewed do not always have a driving licence and would be more interested in taking the train and seeing the landscape rather than renting a car, as proposed on the website.
India is a destination that raised safety concerns for them: Laura had been advised by a friend never to go to mixed dormitories in India, and Pierre expressed his concern about arriving in the middle of the night in New Delhi because he heard that interpersonal relationships can be a bit rough in big Indian cities.
According to my findings, Skyscanner offers almost all of the common and necessary features of a flight comparison website, but these features are not always easily found and some were just considered not available by the users, when they actually were right there on the screen. When they were not found, the experience was so affected that users complained that they did not have the functionality they were looking for.
Those features were so hard to find! For example, Laura complained about the lack of a “notification” button allowing her to be notified of future promotions on the flights for her destination. The functionality existed but was hidden below the proposed flight window and in a window identical to it, so she didn’t see it.
A major obstacle that caused bad mood among the three users was the ubiquitous modal windows interrupting the navigation flow: information about Covid 19 when they didn’t ask for it, the acceptance of cookies that darkens the whole screen until you answer (that one might actually be mandatory but the impatience it provokes at that moment could be taken into account for the rest of the navigation), the proposal to log in that attracts attention when you try to think about your flight, the suggestions for hotel reservations that are recurrent throughout the user’s journey and cause redundancies on the screen, the proposal to create a price alert that, in the form of a modal window, is invasive and arrives too early. It is therefore not even read because it is untimely. These numerous suggestions for hotel reservations are unwelcome, especially for world trotters who like to improvise or organize their stay in a different way, because booking one’s trip, hotel and car in advance is not necessarily a practice adopted by world trotters.
The lack of clarity about the cancellation policy has also been very problematic, as Skyscanner, instead of informing you, sends you to the airline’s website to find out what its policy is on the subject, and that causes a concern that in the eventuality of having to cancel, Skyscanner and the airline may both abdicate responsibility when it comes to making a refund. In addition, the link to find out about the airline’s cancellation policy sends you to the airline’s home page and not to the “cancellation policy” section, so finding out this information is very tedious and faced with these uncertainties, you may be tempted not to even travel.
Another obstacle was the redirection to external websites. Emmanuel found that this website was only a comparator and did not allow him to plan his trip, even though « plan your trip » were the words of the value proposition. Pierre noticed that with each redirection, the flight becomes a little more expensive than advertised because of the cancellation guarantees. The weakness of this value proposition is also due to the fact that the website shows prices from 4 days earlier, as Pierre noted, so they might not be accurate anymore. This means that the result of what is offered is not very reliable, which is a pity for a website whose only interest is to find the cheapest offers. Laura, for her part, noted the usefulness of flight comparators, which allow a great saving of time compared to the practice of going from company to company to find the best price.
However, while Laura confidently thought she had found the best possible price, this was not the case at all. The website did not suggest that she could get a better price by changing her travel dates or airport. Not noticing the discreet filters offered to her, she was going to pay more than twice as much for a trip that would have been significantly cheaper if she had travelled on other dates and to an airport almost as close to the Taj Mahal.
Some important information was missing, such as the duration of the stopover, and a filter to specify the time of arrival, in addition to the existing one of departure date, because arriving in New Delhi at 1 am is not reassuring and it is important to feel safe.
The most unpleasant frictions were therefore: the lack of clarity of the filters and the waste of time this lack engenders, the invasive and too numerous modal windows, the redirection to other websites, the uncertainties and contradictory messages concerning cancellation during the Covid 19 period and the non-completion of the value proposition.
I proposed a “tailor-made” path, with filters more adapted to the user’s needs and more visible, to then offer a “favorites” selection that takes into account the important criteria for the person. The filters are present and modifiable all along, and the “saved as favorite” items are also consultable and modifiable all along. The « favorites » section is inspired by Airbnb’s favorites’ lists.
The Original screens
My lo-fi wireframes
My wireframes, commented
I benchmarked simultaneously with the protocol writing, interviews and prototyping.
Right after the first interview, I summarized by memory the pain points and needs, and I completed this summary as new interviews came along. It was only after the interviews were over that I immersed myself in the details, and my first summaries served as a guideline and allowed me to naturally prioritize the most important issues.
Simultaneously too and from the start, I took notes on the solution I wanted to offer, in a very short and clear statement.
At each problem encountered by the user, I thought of details of the prototype that could solve it and wrote them down as a to-do list, to be used when designing the prototype.
During the brainstorming phase, I started by writing down exactly what would be in each screen before I even drew the lo-fi prototype and this helped me realize that I was about to mix two screens into one (« my recorded trips » and « favorites ») and I was able to untangle this by writing about it. Without this step I would surely have been stuck when I was drawing. It helped me realize that I wasn’t going to offer one new feature, but two. I finally gave up the second one, « my recorded trips », for lack of time.
I received a very good advice from a friend, which helped me a lot during this exercise: an efficient benchmark is an open benchmark. Airbnb inspired me a lot even though it’s not a flight booking website.
What I learned
I’ve learned that memory is a great ally for synthesis. The step of noting down what I retained before transcribing the interviews in detail allowed me to prioritize the information and see what stands out as most striking.
Although I explained to my friend Laura that she couldn’t go wrong and that anything she did would be of interest to me, she apologized for not being able to get the task done fast enough and for perhaps not doing things right. I’ve noticed before that my friends are often sorry for not interacting effectively with websites. No matter how much I tell them that it’s just a matter of observing how the website helps them and how it slows them down, they always think they are responsible for what goes wrong. This shows how much a tool’s flaws go unnoticed, and how much we would all benefit if the consequences of usability problems were known to everyone.
In my next project, I will force myself to use my wireframing kit and look at examples of mid-fi wireframes to find out what details to omit. I have wasted a lot of time making wireframes that are too detailed. I find it difficult not to copy the screen to the nearest pixel because it is a fun activity, but in professional life you have other priorities and you need to save time!