Project Overview

Problem:

Digital travel technologies often fail when exposed to the unpredictable, high-stress conditions of busy city streets. This project evaluates how users navigate e-commerce bike-hiring apps (like Forest and Lime) in the wild, identifying cognitive friction and system mismatches.

Solution:

Formulated data-driven design guidelines focusing on Industry Flow Alignment and Multimodal Signifiers to bridge user knowledge gaps and eliminate systemic freeze states.

My Role:

Lead UX Researcher & Interaction Designer.

Context:

Understanding User Interaction Module (Master's in Human-Computer Interaction & Design (HCID) at City St. George’s University).

Methodologies:

Naturalistic Field Observation, Contextual Inquiry (Master-Apprentice Model) and Framework Matrix Analysis.

Tools Used:

Excel, Word Document, Video Recorder and Handwritten notes.

Theories Used:

Mental Models and Situated Actions

Key Insights and Breakdown:

• Participants frequently suffered from structural mismatches when their prior experience with competing apps (e.g., Lime, Santander) collided with the Forest app interface. When a hurdle appeared, it triggered an immediate comparison of interaction costs,Participants frequently suffered from structural mismatches when their prior experience with competing apps (e.g., Lime, Santander) collided with the Forest app interface. When a hurdle appeared, it triggered an immediate comparison of interaction costs, highlighting high system viscosity.

• When the interface froze or failed to display the map clearly, users faced an immense gulf of evaluation—the app left them completely blind to its actual status. To bypass this systemic block, participants developed an improvised workaround: repeatedly closing and reopening the app. They treated the reset as a necessary behavioural habit to clear environmental confusion.

Outcome:

Data-Driven Design Guidelines

Guideline 1: Design for Industry Flow (Conceptual Scaffolding)

The Issue:

Extraneous business elements (like food teaser ads popping up during a bike rental) interrupt immediate tactical actions, forcing users to rely on slow cognitive recall rather than instant recognition.

The UI Solution:

Align the product's system image with standard industry mental models. Prioritise immediate rental mechanics on the home screen. Use conceptual scaffolding to acknowledge the user's probable background with other micro-mobility services, transforming structural mismatches into gentle, guided learning opportunities.

Guideline 2: Implement Multimodal Signifiers

The Issue:

A lack of feedback during data loading or screen freezes leaves users feeling disconnected, causing panic-induced app restarts.

The UI Solution:

Provide continuous, clear system state context. Integrate high-contrast visual cues alongside haptic/vibrational feedback.

Critical Reflection & Evaluation

Success:

Using a dual-theoretical framework approach provided a deep visibility into both internal thought processes and external environmental realities of users hiring a bike. Focusing on situated action delivered the high ecological validity that standard laboratory testing lacks, exposing how real-world noise and chaos alter user performance.

Challenges:

I took an active role as an observer probing the participants and asking for explanations during the tasks. This may have introduced additional mental processing and extra workload in completing the tasks, think aloud and to formulate answers. While this did not alter task success rates, it significantly prolongs task completion times.