How to Solve Driver-Cancellation in a Gig-Powered App
May 2021, Janice ChengLalamove is a sharing-economy courier service in the Asia-Pacific and Latin America. Operating a fleet of vehicles from motorbikes to moving trucks, Lalamove uses a real-time tracking mobile application to recruit gig drivers and help everyday people make spontaneous same-day deliveries.
One of the biggest challenges to sustainable business growth is to understand the sources of client cancellation. But what happens when we find cancellation initiated from the driver partner side?
By empathizing with driver partners, we can mitigate against client cancellation and help ensure a robust sharing-economy and keep parcels moving around the city.
Project Timeline
This was a solo project completed over a 7-day period following a modified d.school design process. As sole designer, I collected competitive data, conducted user interviews and delivered final prototypes and design assets.
Market Research
Google Play Reviews
In a 14-day period since the mobile update on Apr 28, 2021, the Lalamove user app received 395 reviews on Google Play. After adjusting for repetitive or illegible comments, I found that 14.6% of 82 user complaints were driver-related.
I sorted the results on an affinity map, with red tags representing 1-star reviews and green tags representing 5-star reviews. We can conclude that driver partners are major contributors to Lalamove's customer experience.
Delayed Live Map & Developer Support 39.0% — GPS tracker inaccuracy, app crashes, OTP, loading times.
Customer Service Support 20.7% — Support wait times, refunds, one-time case resolutions.
Other Categories 25.7% — Assignment time, pricing and payment options, misc feedback.
Driver Behaviour 14.6% — Driver tardiness, no-shows, and driver-initiated cancellations.
Fig 1. Affinity mapping of user comments on Google Play
I hypothesize that a better app experience for drivers will contribute to a better customer experience, measurable by higher satisfaction reviews and lower cancellation rates.
The Problem Space
Following the initial market research, I sketched out the project direction by summarizing common pain points and possible solutions for testing and design iteration.
We find interesting cancellation trends stemming from driver behaviour and complaints of poor client-driver communication. The next piece of the puzzle is to understand the how and why by empathizing with user needs within the market context.
The Pain Points
Drivers sometimes initiate cancellation
Drivers sometimes cancel upon arrival for pickup
Inaccurate map updates strains client communication
My Solutions
Empathize with driver partners to understand and mitigate cancellation
Empower customers to autonomously reschedule late driver orders
Enable notifications to override GPS inconsistencies
Competitive Research
Research Scope & Limitations
First, a quick note about geo-blocking. Geo-blocking restricts user access to certain technologies based on geographical location. Since I’m based out of North America, it was not possible for me to reach Lalamove’s direct competitors such as FlashEx, NEXT Trucking, Cabify and Travello for analysis.
Instead, I focused on top competitors in North America, like Uber and subsidiaries such as Uber Eats and Uber Connect. For those not familiar with these technologies, Uber Eats focuses on food delivery, while Uber Connects is a true competitor that focuses on real-time parcel delivery.
I also looked at comparative technologies that are popular with gig drivers, like Waze.
Uber Driver Interviews
My competitive research kicked off with interviews with an Uber Driver of 3+ years. I maintained a working relationship with this user through iterative rounds, which helped me gain important insights into gig drivers' needs and frustrations on the road.
Uber Eats • Uber Connect
To grasp limitations and best practices in the navigation industry, I mapped screen flows in Uber Eats and Uber Connect, focusing mainly on error reporting features and driver-user interactions. I also researched blogs, app reviews and tutorials written by drivers about the platforms.
Fig 2. Competitive screens show Uber Eats, Uber Pool, and Waze from left to right
Waze Road Conditions
Waze is a navigation alternative preferred by many drivers because of a reporting feature that lets users to alert others of road conditions that can’t be captured by satellite GPS, like the presence of police officers or speed cameras. This reporting feature was an important consideration in my exploration of lateral solutions to improve live-time driver-client communication on the road.
User Needs & Frustrations
Next, I sketched out a proto persona of two typical users using an empathy mapping technique that focuses on what users Feel, Think, and Want. While lean market research from the Google Play reviews helped me deduct a picture of user demographics, user interviews informed the motivations that became the basis of my user journey path.
Driver Side
Frustrated when clients blame him for unexpected road blocks (e.g., not safe to stop at pickup)
Considers cancelling a current trip to avoid another angry customer on subsequent planned trips
Wants better communication of real-time road delays to maintain good reviews as a driver
Client Side
Flustered when live location does not give accurate information about the timing of the driver's arrival
Brand dependent as a small business owner who needs both flexibility and the ability to schedule orders
Needs quick alternatives when faced with delays in order to satisfy her own customer ratings as a business owner
Our Driver, Jack
“I just want to have a good driver rating and keep getting more rides.”
Our Driver - call him Jack - is a part-time gig-economy contractor who delivers part-time with Lalamove as a second job to support his growing family. Because of this, Jack prefers to work long shifts over 1-2 days, packing in as many scheduled rides as he can over a 10+ hour shift. He loves Lalamove’s scheduled deliveries feature, and prefers to inter-sparse his working day with more profitable scheduled deliveries and smaller spontaneous delivery calls in between.
Jack’s greatest frustration is when current deliveries become late and put him at risk of missing later scheduled delivery calls. He reasons that a current customer on a late trip is already mad at him for reasons beyond his control - maybe he’s been circling around a pickup location for half an hour with nowhere to park and no easy way to inform the customer. Rather than to jeopardize the chance of getting a positive rating on the next scheduled trip, he may seriously consider sacrificing the current trip and cancelling the current delivery if he feels that the situation is already beyond repair.
In the end, Jack just wants to maintain good ratings as a driver so he can continue getting good gigs with Lalamove.
The Client, Stacey
“I need to think fast and pivot fast when things go wrong as a small business owner.”
Our Client is called Stacey. Stacey is a small business owner who’s found a perfect business solution in Lalamove’s flexible service. She’s a fan of how Lalamove can easily match her orders with local drivers to fulfill same-day or scheduled-day deliveries as her needs change.
Stacey feels frustrated that the app’s live tracking map doesn’t seem to synch up with where her assigned driver actually is. Sometimes the maps shows that the driver is 15 minutes away when he is in reality outside the building. Sometimes the driver even cancels the ride after arriving, leaving her in a bad place if she can’t fulfill a promised delivery to her own business customers. It’s happened before that the driver even gets mad at her for not being able to come down right away at a last minute notice when she’s been diligently tracking the wrong map all along.
Whatever it is, Stacey needs the ability to pivot fast and save her delivery orders when something unexpected happens.
The Iterative Process
How might we help drivers quickly update clients on unexpected delays that are not accounted for by live location?
The next phase of the design process is design iteration. The user persona exercise highlighted a closed-circuit communication dynamic between client and driver, with unexpected road conditions and delivery delays being a common source of frustration for both parties.
I knew that a solution that would simultaneously help both parties reach their goals would mean starting from the source of frustration: which is on the road, when the driver is driving to the client’s pickup point.
Adding a Snooze Button
For the driver, a solution meant the ability to update later scheduled trips when unexpected road blocks cause a delay in a current trip. That involves two things:
Updating the current trip status and ETA
Updating the scheduled trip roster in case time gets pushed back.
Since the driver would first encounter this problem while on the road, I needed to design a simple interface within the map navigation screen where the driver would have direct and easy access as soon as the problem is identified.
I landed on the idea of a traffic “snooze button,” where the driver can trigger a cascade of events to update both the client and time schedule by manually updating his ETA at 5, 10, and 20 minute intervals.
Instant Client Notification
I iterated different button interfaces by prioritizing driver road safety, and considered things like steering wheel position and notification pop-ups that would keep the driver safe while being informed on the road.
Then I moved on to designing the corresponding screens from the client side, mapping out the interactions that cascade after the driver hits “snooze.”
For the client, an ideal solution involved live location updates on delivery and pickup, as well as the ability to quickly reschedule in case of cancellation or unexpected delays.
Since map navigation synching is better left to software developers, I wanted my solution to override map inconsistencies though instant SMS-like notifications that can be integrated with follow-up actions on the main application. The follow-up actions triggered by a delay notification should give the client the ability to do two things:
Find an alternative driver if another faster option is available
Reschedule delivery to another day if no other driver option is available.
After producing block level wireframes, I received 2 rounds of user feedback before moving on to high-fidelity prototyping.
Final Comps
A Possible Solution
Driver Side
Automatically Update Next Trips
Drivers now have the ease of mind that their next trips are notified of delays.
Client Side
Quickly Reassign Driver Orders
Clients now have the option to change tracks when a delay is unacceptable.
In Conclusion
My design solution focused on empowering drivers and clients to troubleshoot autonomously when faced with setbacks.
The next step is to do usability testing on the live market to substantiate these claims.
Clients are less likely to cancel when ETA expectations are aligned with driver realities.
Clients are less likely to be upset when given choice to opt in or out of a bad situation.
By recognizing driver limitations, we can limit drivers' contributions to client cancellations both directly and indirectly.