Date
November 2025
Product
Grab App
Scope
Product Design, UX Research, Interaction Design, Predictive Experience
Industry
Mobility & Transportation
Reimagined Ride Experience
This exploration focuses on how Grab’s ride-booking experience can evolve to anticipate user habits through predictive design.
The goal was to simplify daily commuting by introducing a one-tap reorder feature, enabling users to instantly rebook their frequent rides, such as lrt ←→ office or gym ←→ home routes.
By studying behavioral patterns and context (time, location, routine), the design aims to make ride-booking faster, smarter, and more habit-forming, turning everyday travel into an effortless experience.
SCROLL
Problem Overview
The current Grab ride-booking flow requires users to re-enter pickup details every time, even for routes they take daily, such as lrt ←→ office or gym ←→ home.
While users can save frequent locations, the process still involves multiple taps and cognitive effort, making it inconvenient for habitual commuters.
This friction leads to slower booking times and missed opportunities to build consistent engagement.
The goal of this redesign is to evolve the experience from manual and repetitive to predictive and effortless, allowing users to reorder their frequent rides in just one tap → turning routine travel into a seamless, time-saving habit.
Research Insights
I conducted a heuristic evaluation of the existing Grab ride-booking flow and reviewed secondary research from publicly available e-hailing studies in Malaysia (Lim, 2022; Jais, 2020) and Southeast Asia commuter reports.
Usage is characterised by repetitive travel patterns, especially in urban centres where distance is moderate and origins/destinations are stable.
—Jais & Marzuki, 2020
These studies highlight consistent patterns in urban travel especially routine, repeated trips tied to work, school, and daily activities.
Based on these insights and observed user behavior, I formed a working assumption that a significant portion of Malaysian users frequently travel along the same routes multiple times a week, such as lrt ←→ office or gym ←→ home.
Our survey results show that frequent users of ride-hailing services tend to undertake repeated, habitual trips daily or weekly between familiar zones such as home and workplace.
While e-hailing adoption is increasing, the booking interface remains designed for flexible one-off journeys and does not sufficiently support routine structuring of travel.
—Lim & Fernandez, 2022
Despite these routines, the current experience still requires users to manually reselect their pickup point each time.
Key Observations:
Routine travel is common:
Malaysian e-hailing studies emphasise frequent and predictable commuting patterns.Flow prioritises flexibility:
Users still go through the same multi-step process even when the route is repeated.No proactive assistance:
The system does not surface likely destinations based on time, location, or ride history.Opportunity for efficiency:
A predictive one-tap reorder system could significantly shorten booking time and reduce repetitive effort for habitual riders.
Goals
The goal of this exploration was to simplify repetitive ride-booking behavior and create a smarter, context-aware experience for habitual users.
By understanding how commuters repeatedly book the same few routes, the focus was on reducing unnecessary steps, minimizing cognitive effort, and designing a flow that feels predictive rather than reactive.
Reduce Friction
Allow users to reorder a frequent ride in one tap instead of multiple steps.
Predict Intent
Leverage time, location, and ride history to surface likely destinations automatically.
Increase Efficiency
Shorten booking time from ~25–35 seconds to under 10 seconds.
Reinforce Habit
Create a familiar, rewarding experience that encourages repeat usage through speed and reliability.
Ideation
With the design goal of creating a predictive and effortless ride-booking experience, I explored several concepts to identify the most effective and scalable solution.
Each idea was evaluated based on visibility, effort reduction, and technical feasibility within Grab’s existing ecosystem.
Chosen Direction:
I combined the Home Card Shortcut, Time-Based Reminder + Location-Based Prompt approaches to create a One-Tap Reorder System that appears contextually on the homepage and via push notification.
This solution feels predictive yet remains transparent, allowing users to confirm quickly with confidence.
Home Card Shortcut - User Flow Comparison
Before:
Current Ride-Booking Flow (Existing Grab Experience)
Even though Grab provides a home shortcut card, users still have to manually complete key steps before confirming their ride.
Average time: ~25–35 seconds
Effort level: Moderate, still requires manual input and multiple decision steps
Pain point: Routine riders still repeat the same flow every day
Edge Case: When the System Shouldn’t Trigger
After:
Proposed One-Tap Ride Reorder Flow
A smarter, predictive “Reorder Ride” card surfaces the user’s most likely route based on time, location, and ride history.
Average time: ~8–10 seconds
Effort level: Low, no typing, minimal steps
Value: Fast, frictionless, and optimised for habitual commuters
Time + Location Scenario
How Time + Location Prediction Improves the Experience
On weekday mornings, I usually take the LRT to work. As soon as Im arriving at my usual station, I always open Grab and book a ride from the same pickup point to my office.
This routine happens almost every weekday. Same time window, same location, same destination.
With a predictive system that combines time and location, Grab can anticipate this pattern.
So instead of manually searching for the ride each morning, a contextual prompt appears at the perfect moment:
Because the system detects:
Time pattern: Weekday mornings between 8:00–9:00 AM
Location pattern: Im arriving at my usual LRT station
Ride pattern: I always go from this station → office
I don’t need to retype or adjust my pickup point, it’s already correct.
A single tap is enough to confirm the ride.
Edge Case: When the System Shouldn’t Trigger
Edge Case: When the System Shouldn’t Trigger
On days when I’m on sick leave and staying at home, this smart prompt doesn’t appear.
Why?
Because the system recognises that:
My location is still at home
I didn’t travel to my usual station
The usual commuting pattern isn’t happening today
By combining both signals time and location the system avoids incorrect or irrelevant reminders.
This creates a more trustworthy, context-aware, and non-intrusive experience.
Time-based prediction alone can be inaccurate. Location-based prediction alone isn’t enough. But the combination creates a highly reliable signal that aligns with real commuter behaviour, making the one-tap reorder meaningful instead of annoying.
How The Predictive Logic Works
Risks & Mitigations
Anticipated Impact & Outcome
By introducing a predictive, one-tap ride reorder experience, the redesigned flow aims to significantly improve efficiency and convenience for habitual commuters. Reducing repetitive steps not only saves time but strengthens long-term engagement by aligning with real user routines.
Faster Booking Experience
• Reduce average booking time from ~25–35 seconds to < 10 seconds for habitual commuters.
• Minimize manual input and repetitive interactions.
Higher Repeat-Ride Conversion
• Increased likelihood of booking during peak commute hours due to contextual reminders.
• Improved relevance through location- and time-based prediction.
Better User Satisfaction
• A more seamless experience builds trust and reduces friction, especially for daily commuters.
• Clear confirmation patterns reinforce confidence and help users feel in control.
Stronger Long-Term Retention
• When daily tasks become effortless, users naturally return more frequently.
• Predictive UX encourages habit formation through speed, convenience, and consistency.
Reflection
This exploration highlighted how much value can be unlocked by understanding user routines and designing for real-world behavior.
Even though the ride-booking flow in Grab is already optimized for flexibility, there is still a gap for habitual commuters who repeat the same journeys every day.
By shifting the experience from reactive (user inputs everything) to predictive (system anticipates intent), the process becomes significantly faster and more intuitive.
The key learning was how powerful contextual signals such as time, location, and ride history can be when combined to drive a seamless experience.
Designing this solution also reinforced the importance of balancing convenience with trust.
Users should feel the system is helping them save time, without removing control or causing unexpected actions.
Next Steps
If this were to move into production, the following steps would be essential:
Disclaimer: This case study was created independently as a personal exploration, and is not associated with or commissioned by Grab.
