Rocket Mode in Delivery App
Context
Rocket Mode is a feature within the Delivery App, used by delivery drivers to manage their daily work. This feature was developed as an improvement over manual route sorting, with the aim of automating the process and making it more efficient.
Objective
Facilitate daily deliveries, especially for new users who are still not familiar with their area. From a business perspective, the goal is to reduce delivery times, increase the number of completed services, and improve the accuracy of time estimation for recipients.
My role
I co-designed all the functionality at the UX level and was responsible for the complete UI design. I worked in collaboration with development, data, and operations teams distributed between Spain and Germany. The project demanded very tight delivery times and a design compatible with older versions of Android devices.
Results and metrics
Reduction of 7-10% of deliveries outside the established time window
Increase of 15-17% in packages delivered per day
Greater efficiency in fuel consumption
Launched in selected agencies in Spain and Portugal
Working process
Research and ideation
The ideation phase began with an in-person session with teams from Germany and Spain. The German team provided a detailed overview of actual delivery driver usage of the app. They shared insights on how routes were currently organized, what criteria drivers used to schedule their workday, and what differences they observed between new and experienced drivers.
Based on this, we jointly reviewed other apps and related operational processes, both internal and external. We analyzed the entire delivery flow from the perspective of the various profiles involved—receivers, operations, delivery staff, and customer service—and identified strengths, pain points, and potential opportunities for improvement at each stage of the process.
One of the most important findings was that manually sorting the route was extremely time-consuming and generated considerable uncertainty, especially among new drivers. Furthermore, this initial process was not always well reflected in the estimated delivery times for recipients, which negatively impacted their perception of the service.
With all that analysis, we defined the key areas that the new functionality should cover:
Automate the organisation of the route to make it as efficient as possible.
Reduce delivery times and the number of undelivered packages due to lack of time.
Facilitate the onboarding of new delivery drivers, reducing the learning curve.
Improve the accuracy of the estimated arrival time to the recipient.
Prototyping
When working on an already established app, with well-defined components and flows, I did not spend excessive time creating wireframes or low-fidelity design. I went straight to high-fidelity prototyping, adapting the new proposals directly onto the real interface.
I designed a high-fidelity prototype in Figma, including key interactions to simulate the real behaviour of the functionality. What I aimed to achieve was to convey speed, immediacy, and a visual improvement that would make the system feel modern.
The prototype covered the complete flow of selecting the type of route sorting, the notification that appears when the system detects a more optimal route, and additionally, a view in the app for the recipient to track the package.
The work involved both the creation of new screens for Rocket Mode and the modification of existing components and adjustments to the architecture of the main menu. Although I had already created guidelines and usage instructions for the components, it was necessary to adapt the content to integrate this new logic without breaking the overall experience.
To share the operation and facilitate validation with other teams, I recorded a video in Loom explaining the flow and key points of the design. The prototype was tested internally, which allowed for the detection of important improvements, such as the need to adapt some interactive elements for older, low-performance devices, which are very common among users.
The prototype also served as a direct guide for the development team, acting as a visual and functional reference during the implementation.
Some visual decisions
Visually, the main challenge was to integrate a new functionality without breaking the coherence of the existing system, while at the same time not adding more complexity than necessary in a flow that needed to be fast, clear, and easy to execute. The goal was for users to be able to activate and understand "Rocket Mode" effortlessly, even in high-pressure situations or while on the move.
To achieve this, I made a series of decisions aimed at reinforcing clarity, visual focus, and speed of interpretation:
Iconography: I used the Material Design icon library to unify all the iconography of the application. In some cases, I used that library as a base to extend with a specific icon for the app.
Hierarchy in the menu: I reorganised the architecture of the main menu to give visibility to the new sorting flow, without displacing or interfering with already established functionalities.
Colors and subtle indicators: I incorporated soft micro-animations and discreet visual elements to indicate when the system had detected a better route. It was important to highlight this information without causing shocks or visual noise, as many users interact with the app while moving, on the street, or even inside a vehicle.
Consistency and performance: all new elements were built from existing components of the design system. This allowed for maintaining visual consistency and ensuring compatibility with older or low-performance Android devices, which is very common in this environment.
These decisions aimed for a balance between visibility, visual continuity, and technical efficiency, always taking into account the real context of use: high speed, less than ideal conditions, and users who need quick responses without distractions.
Challenges and learnings
One of the main challenges was to design an advanced feature like Rocket Mode in a technically limited environment. Many users use devices with outdated versions of Android and low power, which required optimising every detail of the design: reducing complex animations, simplifying interactions, and reusing system components to avoid compromising performance.
Another key point came from internal feedback: some users perceived the feature as unreliable or "too pretty" to be really useful. This made me rethink how to communicate the value of Rocket Mode, not only visually but also functionally. We incorporated clearer messages, direct explanations of its operation, and reinforced the elements that indicated that the system was genuinely optimising the route.
In operational products like this, the design must not only be usable and efficient, it must also instil confidence in high-pressure contexts with a low margin for error.
