🚗 Improving Car Sharing Operations
To make a centralized car sharing business viable, the operations behind preparing and maintaining vehicles for new customers is essential. At Avail, I launched this investigative research study to understand how the Operation's team processes may run more efficiently. Ultimately, findings from this project improved physical operations site designs, staffing schedules, standard operating procedures for both the Operations and Customer Service teams, and enterprise tooling experiences.
👩🏻💼 My Role
UX Research Lead
🥅 Project Goal
Reduce the time-on-task of vehicle intake transactions.
💭 Methods
-
Ethnography
-
Contextual Inquiry
-
Survey
🗓️ Duration
8 weeks (2022)
📈 Impact
-
Reduced time-on-task by 30+ minutes for four key tasks
-
Increased System Usability Scales (SUS) score by 14% (from a 59.3 to 67.6)
-
Sculpted 2021 roadmap for enterprise tools
Context
This project was for Avail, a peer-to-peer car sharing marketplace owned by Allstate.
At centralized locations nationwide, Avail Field Operations (FieldOps) teams facilitate transactions between car owners and borrowers. Their tasks include handling all customer interactions, cleaning & inspecting cars, and handling vehicle maintenance.
The Problem
FieldOps employees took 8 hours longer to complete each vehicle intake transaction than what a cost-sustainable labor model required. (A vehicle intake transaction includes cleaning, inspecting, and preparing a vehicle for sharing.)
Some business leaders hypothesized that poor usability of the digital enterprise tool caused this slowdown. However, this hypothesis lacked concrete data. Therefore, I collaborated with the Director of Operations and Operations Product Manager to propose a research study to identify the root problems before proceeding with design and development. This effort led to setting specific goals and key questions for the study.
Research Goals:
-
Understand the FieldOps team's processes, experiences, and digital product usage.
-
Identify opportunity areas to reduce vehicle intake transaction times.
Key Research Questions:
-
What are the daily tasks of a FieldOps employee, and how do they vary by role?
-
How is the vehicle intake process conducted?
-
How does the digital enterprise tool support or hinder the vehicle intake process?
-
What changes can improve operational efficiency?
My Approach
Desk Research
1 week
Stakeholder Interviews
0.5 week
User Survey
1 week
Ethnography
2 weeks
Artifact Creation
2 weeks
Share Insights
2 weeks
1. Desk Research
Before developing a rigorous research plan, I gathered available data to evaluate the problem space, eliminate redundant work by compiling existing knowns and unknowns, and guide next steps for research. My methods included:
Before developing a rigorous research plan, I compiled available data by:
-
Reviewing product analytics to understand vehicle transaction time and its variation by location and customer volume.
-
Reading FieldOps training documents to learn about employee training, procedures, and technology integrations.
-
Watching session recordings to observe how FieldOps employees use the enterprise tool and identify potential friction points.
-
Reviewing customer volume trends to compare weekly customer volume by location with product analytics.
-
Conducting enterprise product walkthroughs to understand tooling functions first-hand.
-
Clustering past research insights to identify previous trends in employee feedback and confirm existing knowns & unknowns.
2. Stakeholder Interviews
While conducting desk research, I also held one-on-one discussions with stakeholders in Operations, Customer Support, and Product teams. In doing so, I confirmed team priorities and achieve alignment across functional areas on goals. Additionally, these insights informed the research direction by identifying hypotheses to validate, complaints to understand, and visions to target.
💡Key Learnings
-
There was a discrepancy between enterprise tooling product flows and physical operation processes.
-
The FieldOps team handles numerous asynchronous tasks daily.
-
Tasks are distributed across four digital tools, not limited to the primary enterprise tool.
3. User Survey
Next, I distributed a survey to the FieldOps team. This was a low-cost method to understand how locations, roles, goals, and environments influence product interactions, frustrations, and desires. My survey aimed to build empathy and understanding for the FieldOps team, center the research on end-users, and guide the decision on a Field Study location.
💡Key Learnings
-
Physical lots vary significantly by location.
-
Connectivity can be spotty, impacting digital tooling load time.
-
Job descriptions between Managers and Associates differ, yet their day-to-day has overlap.
-
Device type usage (iPads vs. iPhones) differs by user role & location.
4. Ethnography
I planned a 3-day site-visit to the LAX operations site. Because operations is dependent on interactions with customers & digital tools, physical lot setups, state of vehicles upon return, and more, it was essential to understand what factors impact vehicle intake and how these factors interact.
I observed the FieldOps team and environment without interruption to accurately capture how tasks are naturally performed, as people's actions in real-life situations often differ from what they report in interviews. This method also aimed to provide insights into the physical environment and how all service components interact, offering a comprehensive understanding crucial for optimizing processes.
Output
-
Rough time-on-task numbers
-
Tally counts of how many times certain behaviors occurred
-
Detailed documentation using the "People, Objects, Environments, Messages, & Services" (POEMS) framework
User Observation
I asked questions to FieldOps team members as they worked through their jobs. This helped gather real-time thought processes during tasks, allowing me to ask probing questions to learn about variations in operational processes and gain clarity in my observations.
Output
-
Robust understanding of work processes, mental models, and common behaviors
-
Task analysis
Contextual Inquiry
💡Key Learnings
-
Vehicle intake was constantly being interrupted by (1) customers arriving early & late, (2) phone calls requesting a shuttle, (3) higher priority tasks requiring attention at different locations.
-
If left completely uninterrupted, the full vehicle intake process took ~1 hour.
-
The physical spread of the LAX lot made it impossible for the 1–3 FieldOps employees to tend to customers while simultaneously completing all tasks efficiently.
5. Artifact Creation
I documented findings tailored for various stakeholders to distribute insights.
I created fictional representations of the FieldOps employee roles. These helped engineers, product team members, and other non-FieldOps employees develop empathy for FieldOps team members by increasing their understanding the constraints, mindsets, and environments in which FieldOps employees work when interacting with the products being developed for them.
Output
Two personas of FieldOps employees:
-
Managers
-
Lot Leads/Associates
After creating, personas were linked in all Jira tickets to shape user stories.
FieldOps Personas
Process Mapping
I mapped out flow charts of how all owner, borrower, and Field Staff employee user journeys overlap with technology to deliver the car sharing service. This artifact provided a visual representation of how everything functions together as a system to enable holistic, cross-functional thinking.
Output
Flow mapping of user journeys & product interactions for:
-
Car owners
-
Car borrowers
-
FieldOps employees
I consolidated documentation of all learnings and observations during the research process to allow cross-functional teams to read and digest the information at their convenience.
Output
26-page document detailing Field Study learnings.
Research Report & Slide Deck
6. Share Insights
After allowing cross-functional stakeholders time to gather their thoughts after reading the research report, I facilitated multiple workshop discussions to help them internalize the findings, ask questions, and determine actionable next steps.
Key Findings
While the enterprise tool had plenty of room for improvement, it was not the root cause for why vehicle intake took so long. The FieldOps team was constantly task-switching, which resulted in lengthy times to complete vehicle intake.
The high volume of task switching was due to:
-
Limited staff capacity during times with high customer volume
-
Unexpected & undesired customer behaviors upon arrival and drop-off
-
Inefficient physical lot setups
The primary digital tooling problem (that did not impact vehicle intake) was the lack of inventory transparency & status process flows.
Key Opportunity Areas
-
Re-evaluate staffing model to account for busier and slower times
-
Reduce physical spread of Ops work by prioritizing team roles & responsibilities
-
Reduce # of touch-points required to solve a case by automating cases and providing greater customer transparency
-
Shape customer behavior by setting expectations and sending push notifications
-
Reduce manual tracking by building process flows & inventory transparency to reduce # of FieldOps tasks
-
Make small usability improvements to speed up repetitive tasks
📈 Impact
Immediate
-
Sculpted 2021 product roadmap for the enterprise tooling suite
-
Improved operational site designs by creating physical layout guidelines for the launch team opening new locations
-
Revamped operational staffing hours at busy markets to reflect staffing needs based on the volume of customers and necessary vehicle intakes
Long-Term
-
Increased System Usability Scales for the enterprise tool by 14% (from a 59.3 to 67.6) in 3 quarters
-
Reduced time-on-task by 30+ minutes for 4 key operational tasks:
-
Conducting vehicle intake
-
Tracking late return & stolen vehicles
-
Handling unfit vehicles & facilitating maintenance
-
Auditing inventory
-
-
Streamlined standard operating procedures for late returns and unfit vehicles
💭 Reflections
👍 What went well
-
Thorough approach to understanding problem space and guiding research direction
-
Became subject matter expert on FieldOps processes and enterprise tooling
-
First documentation of FieldOps processes & technical flows in company history was frequently referenced as a source of truth
👎 Missed opportunities
-
Process map was not turned into a formal service blueprint
-
Didn't observe a full week of FieldOps processes due to budget constraints