Ride cancellations were reduced by 9% due to a transparent content and design strategy towards the user
November 2022 - January 2023
Freenow is a ride-hailing and multi-mobility app in 9 european countries. Users can choose where and when do they need to be picked up.
Context
Problem
+17% of users who reserve a taxi or ride ahead of time, cancel their reservation when there is more than X min to pickup* and there is still no driver allocated.
* (For confidentiality reasons, I can't display the exact details)
The overall lost core revenue due to these cancellations transforms this problem into a major opportunity for the company.
The more significant challenge is alleviating users' concerns about the service's reliability, consistency, and availability.
My role
I was the responsible product designer in a cross-functional team that consisted of a PM, User Researcher and a UX Writer. On the development side, I worked together with the EM, two iOS developers, two Android developers and two backend developers.
Throughout the project, I promoted early brainstorming and teamwork with the rest of the group.
Tools
-
Figma, Notion, Miro, UzerZoom, Usability Hub
UX artefacts
-
User journey maps, task flows, wireframes, interactive prototypes, specs
UX Research
-
Preference test and analysis, user testing, and 1 hour in-depth remote interviews
Situation
The issue was that the company was losing money due to high cancellation rates. Users who prebooked a taxi or ride, tend to panic if they don't get a confirmation that they have a driver before X minutes to pickup time.
The purpose of this project was to alleviate this fear by keeping a clear communication with the user throughout the whole waiting time.
I was given a time frame of 2 weeks to come up with a content and user experience strategy that later we would test with real users.
My strategy relied on mapping the information that users need during the waiting time until a driver accepts the request. If we didn't improve the amount of canceled rides, we were likely going to keep losing market share to rivals, like Uber.
Original user flow
Process
1. Finding the right problem
The team was asked to participate in a problem-understanding workshop led by our PM.
The goal was to present all the previous discovery and identify the most relevant problems that we saw we could solve.
Due to the need to ship something fast for the MVP, the team descoped some of the problems for future releases.
Next step was mine to take. I had to focus on solving the uncertain feeling that our users have over the functionality of prebooking with Freenow.
2. Finding the right solution
Content strategy workshop
Together with the UX Writer and UX Researcher of the team, we defined what information pieces the user should see at each stage of the user journey so that I could design accordingly.
Currently, the system only sends the order to the drivers X min before the pickup. That's why the user didn't have the driver's details before and would cancel in a panic, not knowing what happened.
We believed the user would feel more at ease by explicitly communicating when the driver information would be sent.
Design iterations: What information is the user looking for?...
...Is it the address of the pickup and destination? The service type they were going to get? The scheduled time? The driver’s details?
Once I had some directions, I brainstormed with the Principle PD. As a result of this session, we discovered that to assist the user during their wait for a driver, we could use a timeline displaying all the steps they will receive and when they should anticipate them. This way, we solve one of our primary concerns: How to explain to our users that, due to our allocation system, they do not have a driver when they place an order.
3. Solution validation: User test
We wished to verify the following during the user testing:
-
Usability of the flow
-
Comprehension and reaction: How would users react to not knowing the driver's identity until X minutes before pick-up time?
-
Findability: Do they know how to cancel a prebook? Do they know where to look for the prebook once it has been created?
-
Hierarchy of the prebook detail page: What information is more relevant?
-
Notification system: How supported do they feel during the waiting time?
The UX researcher used our customer base to make the hire. We used CRM to notify users who had already prebooked a ride with us. We invited them to a one-hour moderated remote discussion.
Unfortunately, we did not receive many responses, and we could only question three users. One of them was entirely facilitated by me.
User test: Major learnings
Constraints and trade-off
The MVP idea was submitted to two groups: the development team and fellow designers through different rounds of feedback. Both had some discrepancies in the approach of using a timeline element.
1. Development team
The team discovered that adding this feature would need them to perform additional backend development that were not prepared to accomplish at that point in the project.
I prepared a version that would take out all the dynamic elements of the page.
2. Design peer feedback
On the other hand, some of my design colleagues questioned if using a timeline element was the appropriate strategy.
They presented a different viewpoint: The only difference between an ad hoc basis booking and an advanced booking was the time between placing the order and having a driver accept it. If we keep the same narrative for both types of rides, we will bring consistency to the user.
We only had three user insights, which could be insufficient, so I agreed to consider this perspective. As a result, we had two versions of the same page.
3. Seeking further user feedback
I used the delivery delay the development team was experiencing as an opportunity to persuade the PM to consent to do more research. By doing so, we could confirm which of the two options would assist the user more during the waiting time.
1. Preference test
I built up a preference test with Usability Hub Tool. This allowed us to immediately measure the users' preferences for the two designs. The timeline version performed much better.
Timeline version - Winner!
71%
of users
Easier information architecture
79%
of users
Driver reassurance
71%
of users
Design appealigness
2. Further in-depth interviews
We also took advantage of some interviews that were taking place for other projects. I squeezed a simplified prototype with a couple of questions to gain further qualitative insights into both alternatives.
The recruitment was more successful this time, and we interviewed eight users. I moderated two of them.
The timeline version was perceived as the more appropriate. It made users feel safe and confident, and that Freenow was taking care of the prebook.
"it's nicer, I prefer this one. I feel it's talking to me" - User feedback
"If you see the progress, those things ticking must be that action is being taken"
- User feedback
Results
MVP flow
Live launch results
Overall, the MVP flow was successful in reducing cancellations by 9%.
-
When users visited the How it works page, the cancellations decreased 14%.
-
When users saw the information card at the beginning of the flow, cancellations decreased by 8%.
Learnings
1. Does the user actually value transparency?
My biggest surprise was that I encountered some resistance despite pushing for transparent communication.
I expected that the users would be pleased about alerting them that the driver's information would be shared within X minutes from the pickup time. However, it was frequently proved in the interviews that it would make people feel unsure.
As a result, we recommended running a copy a/b test to check if, in a real-world scenario, when users have the choice to book a ride, they will continue to cancel.
2. The effects of content first approach
I adopted a new approach by first discussing with the UX team on the essential messages we wanted to highlight in the flow. This had a positive effect on our collaboration:
-
The UX Writer felt much more involved in the process, able to own the storytelling and give more ideas.
-
It made me more confident in determining relevant information in each flow phase and appropriately designing a solution.