SCOTT
Reducing Meta’s security risk by eliminating orphaned data and driving 80% faster access to critical assets, I designed SCOTT (SaaS Collaboration Ownership Transfer Tool) — a proactive offboarding workflow that safeguarded sensitive PI data, prevented costly leaks, and cut access turnaround from weeks to under two days. The solution also saved managers 6–10 hours per departing employee and reduced IT support tickets by 500–700 monthly.
Role
Product Designer
Length
4 months (while balancing other initiatives)
Users
80,000 + Meta Employees
Responsibilities
-
Product Design — Designing new internal tool from the ground up
-
Problem Definition — Redefining the core challenge and solution space
-
User Research — Identifying pain points and requirements; usability testing
-
Design Systems — Building within Meta’s internal xDS framework
Homepage
Configure File Transfer
The Problem
Orphaned data are files that have no active owner.
These files poises significant security risks...
Employee leaves
Files not transferred
Orphaned data
Locked, lost, or
legally risky
Risk of Orphaned Files
-
Locked or inaccessible files due to lack of clear ownership, making edits and permissions unmanageable.
-
Lengthy legal processes are often required to recover orphaned data, involving multiple stakeholders and high costs.
-
Sensitive information may be exposed during recovery, increasing the risk of unauthorized access.
-
Permanent loss of important work is possible when there’s no identifiable owner or point of contact to retrieve the files.
I partnered with the data analytics team and discovered that 33% of Meta-owned assets—approximately 25 million files—are orphaned data!
Project Kickoff
Starting With Ambiguity
Stakeholders wanted to build something like BetterCloud, a tool that automates data security workflows across cloud applications.
Bettercloud - The Inspiration Behind This Initiative
The initial direction focused on building a generic admin portal, but it lacked a clearly defined user problem.
At the same time, early work had begun on data risk classification systems, though it was still in its infancy with minimal user research or validation.
I advocated for a large-scale qualitative study with stakeholders to better understand the problem space and identify a high-priority challenge we could address within our constraints. The research aimed to:
Research Approach
-
Understand file storage and sharing behaviors contributing to data risk
-
Define what makes a file important, sensitive, or risky from the user’s perspective
-
Identify pain points in offboarding, file transfers, and access requests
-
Explore opportunities to improve workflows for shared drive admins and support teams
Who
13 participants, mix of teams and personas
What
1:1 60 min interviews
When
1 Month
How
Remote Sessions through VC
Narrowing Scope
Early in the research, I uncovered the concept of orphaned data—cloud assets without a clear owner—which posed a major security risk. With up to 33% of Meta-owned assets falling into this category, and because ownership could be reassigned, it offered a tangible short-term problem with significant impact. I chose to narrow my research on this area while still exploring the broader cloud security space.
Research Affinity Board
Legal & Security Risks
"Giving owners a heads up if the added email address is external will be very beneficial. Vendor sharing is much more controlled."
Sharing becomes a legal risk if not done by the owner; risk of data leakage without consent
"If a leaving employee can just check a box to send all files to manager, we would know that file sharing is legal and clean."
External sharing is risky without proper notifications and controls
Offboarding Challenges & Orphaned Data
"Whenever someone leaves, sending them a checklist or automated email saying 'Hey, make sure you do XYZ' will prevent a lot of orphaned data."
Employees don’t understand offboarding fully; lack of reminders leads to orphaned data
"We need to send leaving employees to a tool that the user can do easily—transfer content."
Current tools are slow/manual; bulk transfers are needed to prevent incomplete offboarding
Support & Operational Burden
"There are too many tasks for onboarding requests. If assets get transferred automatically during offboarding, will cut down 90–95% of support requests."
Support teams are overwhelmed; automation could significantly reduce workload
Lack of Standards & User Education
"People don’t know how it (shared drives) functions or works... there are no standards."
Employees aren’t trained on best practices for file storage/sharing; poor understanding of risk
Mapping the User Experience
Primary Users
Employee Leaving Meta
-
Needs help transferring, archiving, or flagging important content
-
Often unsure which files matter or where they should go
-
Exits may be planned or abrupt, with little time for manual handoff
Employee Requesting File Access
-
Needs access to files from a departed employee
-
Faces delays due to approval bottlenecks and permission issues
-
Needs quick, compliant access with minimal support friction
Departing Employee File Offboading
Employee Requesting Access to Orphaned Files
Defining the Problem Space
Automation
Current offboarding and support processes are largely manual and inconsistent. Automation would reduce workload, risk, and support tickets.
Security
Orphaned files are widespread. Without visibility into ownership or sensitivity, sensitive content may be shared externally or left exposed.
Awareness Gaps
Employees often don’t know how to properly offboard data or the risks involved. There's little education around data ownership and safe transfer.
Lack of Proactive Safeguards
Most remediation efforts happen after data is lost or shared. A proactive system for flagging risk and prompting action doesn’t exist.
Policy Risk
Files accessed without manager + HR approval may violate internal policy. Without proper checkpoints, this exposes teams to legal and compliance issues.
Decentralization
Ownership is hard to track across platforms and drives. Without a centralized system, audit and access reviews become time-consuming and error-prone.
Ideation
I led a design thinking workshop where participants reviewed research, mapped pain points, and co-created solutions. We prioritized ideas by effort and impact, defining a clear MVP and future opportunities.
Results
Short Term - Quick Wins
• Identify external file shares
• Flag files with risk indicators
• Let users bulk-remove access
• Email reminders before departure
• Self-service tool to reassign ownership
Mid-Term - Next Phase
• Auto-detect sensitive content
• Suggest new owners by team
• Trigger alerts when people resign
• Track file access/ownership changes
Long Term - Future Vision
• Use AI to classify files
• Auto-archive low-risk data
• Fully automate handoffs
• Standardize policy across cloud tools
The Solution
Google Drive Offboarding Tool
We focused the MVP on Google Drive—Meta’s most widely used file platform. The goal: prevent orphaned data through automation and reduce friction for both employees and support staff during exits.
Key Challenges
-
Risk Detection: Scanning for sensitive data (e.g., SSH keys) wasn't available, limiting early detection capabilities.
-
User Flexibility: The tool had to support both abrupt terminations and planned exits, with varying levels of user engagement.
Early Design Wireframes
Preliminary Direction
Our early concept for SCOTT was a self-service offboarding portal that guided leaving employees through the file transfer process. The portal allowed them to:
-
View a summary of to-dos and file transfer progress, with links to educational resources
-
Identify high-risk files and folders flagged for transfer
-
Perform single or bulk transfers using a file table
To support collaboration, the MVP wireframes originally included a receiver’s view, allowing managers or peers to accept transfers or reassign them to the right person or team. This feature was later cut to meet the project deadline.
1st Iteration
Homepage
-
Education: Clear links to offboarding resources, including guidelines on what should and shouldn’t be transferred.
-
Urgency: Time remaining and recommended files are prominently displayed, with color-coding to indicate urgency as the offboarding deadline approaches.
-
Recommended Files: Smart suggestions based on collaboration, recency, access patterns, and popularity—prioritized for research-critical assets.
-
Quick Transfer Option: One-click bulk transfer for employees who prefer speed over manual review—best used for low-sensitivity files.
Quick Transfer Modal
Provided a Quick Transfer option within the multi-step modal flow for employees who preferred not to manually review files in the detailed table view.
File Table View
-
Timeline: In this iteration, file transfers were scheduled to begin after the employee's last day, giving them time to review and prepare.
-
Assignment Workflow: The table is split into Unassigned and Assigned sections to help track transfer progress at a glance.
-
Visual Cues: Metadata such as share type, access list, and collaborators empowered users to make informed decisions without switching context.
-
Folder Navigation: A breadcrumb-style folder tree allows users to drill down into specific folders or navigate back with ease.
User Testing Results
While the sample size was small, all participants rated SCOTT higher than the existing workflow, with an average CSAT of 4.25/5.
Turning a Setback Into a Smarter MVP
Our cross-functional partners let us know that the file table view wouldn’t be ready by MVP launch, even though we had access to some individual file-level data.
This meant we had to design smarter, condition-based transfer logic—without allowing users to browse or select files directly in SCOTT.
The Solution
Since we couldn’t support a full file table view at launch, we leaned on available metadata to drive intelligent transfer logic.
We prioritized file signals like:
-
Recently edited
-
Recently accessed
-
Number of viewers
-
Externally shared
-
Internally shared
-
Private
These signals helped us surface the most relevant files without requiring users to manually browse their drive for the 2nd iteration.
2nd Iteration
Homepage Pre-Transfer
V1 Insights & Solutions
-
Unclear Homepage CTA: The main call-to-action was diluted by visual clutter and competing elements, causing key information to be missed.
-
Solution
-
Reduced to a single, primary button to start the transfer flow.
-
Streamlined page with minimal supporting text.
-
Added accordion-style dropdowns to explain each step only when needed.
-
-
-
Confusing File Counts: Users struggled to understand fractional file counts and how bucket totals related to the overall number of files left to transfer.
-
Solution
-
Introduced a bar chart to visualize the distribution and status of files in the user’s GDrive.
-
Removed individual counts for recommended file categories to reduce cognitive load.
-
-
-
Missing File Table: The MVP excluded the file table view, yet users still needed a way to review their files before transferring.
-
Solution
-
Added first-step prompt: "Review Files in Your GDrive".
-
This opened a new tab, allowing users to review and delete unwanted files before returning to SCOTT to bulk transfer everything.
-
-
Configure Transfer
Configure Transfer Step 2
V1 Insights & Solutions
-
Defining High-Priority Files: Without a file table in the MVP, users still needed a way to specify which files mattered most for transfer.
-
Solution
-
Externally shared files were auto-selected by default to reduce risk and ensure critical visibility was maintained.
-
Selecting “Some Files” revealed an additional “Define Parameters” section, allowing users to set custom transfer criteria.
-
-
-
Transfer Communication: Users wanted reassurance that their transfer succeeded before manager sign-off, in case of errors
-
Solution
-
Provided two options:
-
Transfer now to immediately verify success
-
Or schedule transfer after their last day for a more automated experience
-
-
-
Configure Transfer Step 3
V1 Insights & Solutions
-
Pre-Transfer Cleanup: Users wanted the ability to clean up Google Drive before transferring, including deleting non-recommended, private, or outdated files.
-
Solution
-
After users defined their transfer parameters, there were often leftover files without a decision. We introduced a step allowing users to take action on these remaining files—including the option to delete them directly.
-
-
Homepage Post Transfer
V1 Insights and Soltutions
-
Approval: Users wanted a final approval step from their manager to feel confident everything was properly handed off before their last day.
-
Solution
-
Once a transfer is scheduled, a manager confirmation task is automatically created and surfaced to ensure final review and sign-off.
-
-
-
Unclear File Counts: Fractional file counts confused users, and they lacked a clear way to track whether transfers were successful.
-
Solution
-
After scheduling a transfer—whether immediate or post–last day—users could monitor progress through a visual progress bar, providing clarity and confirmation of success.
-
-
-
Other Cloud Platforms: The MVP only supported Google Drive, but employees also stored important files in other platforms.
-
Solution
-
Added a banner reminder prompting users to audit their other cloud storage accounts, with quick links to commonly used platforms for easy access.
-
-
Final Designs
Homepage Pre Transfer
V2 Insights & Solutions
-
Unclear Purpose: The homepage still felt busy and didn’t clearly communicate the tool’s purpose.
-
Solution
-
Further reduced on-page text for clarity.
-
Introduced a banner with key guidance.
-
Added a system-state illustration for empty-state clarity.
-
Reintroduced a “Files Pending Decision” widget to surface unassigned files
-
-
-
-
Confusing Statuses: Progress and file statuses were unclear—terms like Unmanaged felt vague, and overlapping logic (decision vs. processing) with multiple colors made it hard to tell when tasks were complete.
-
Solution
-
Reduced status types to show only decision state and manual removals, improving clarity.
-
-
-
Inaccessible Colors: The original progress colors didn’t meet accessibility contrast standards.
-
Solution
-
Updated progress colors to blue, pink, and grey for better accessibility and contrast compliance.
-
-
Configure Transfer Introduction
V2 Insights & Solutions
-
Unclear Purpose: Users were still unsure about the tool’s purpose and continued to overlook key features.
-
Solution
-
Added an introductory step before the transfer flow to clarify the tool’s purpose and highlight important actions.
-
-
Configure Transfer: File Selection
V2 Insights & Solutions
-
Overwhelming Fields: Too many fields upfront made users feel overwhelmed and less likely to complete the flow. For configuration parameters, users were also unsure what values to input.
-
Solution
-
Split “Configure Transfer”, “Assign Location”, and “Schedule Transfer” into separate steps to reduce cognitive load.
-
Pre-filled recommended input values to guide users in targeting high-priority files.
-
-
Configure Transfer: Assign Location
V2 Insights & Solutions
-
Subfolder & Naming Flexibility: Users wanted the ability to select a subfolder within a shared drive and customize the name of the destination folder for transferred files.
-
Solution
-
Added support for selecting an optional subfolder and customizing its name during the transfer setup.
-
-
Homepage: Processing Transfer
V2 Insights & Solutions
-
Transfer Anxiety: Users felt anxious about the transfer process and wanted more clarity around system status.
-
Solution
-
Added a processing confirmation toast with the ability to cancel the request and start over, giving users more control and peace of mind.
-
-
-
Risk of Immediate Transfers: If users selected “Transfer Now” after configuring which important files to move, leftover files were sent immediately—potentially without proper review and increasing the risk of data leakage.
-
Solution
-
Automatically scheduled remaining files to transfer after the employee’s last day, allowing more time for review and reducing risk.
-
-
-
Unclear Next Steps: After completing the transfer configuration, users weren’t sure what to do next, and additional tasks (like reviewing other cloud drives) weren’t clearly communicated.
-
Solution
-
Refocused the page around a clear checklist of next steps to guide users through remaining actions and ensure nothing was missed.
-
-
Anticipated Impact & ROI
~500–700 monthly support tickets avoided
Based on IT data, file access & reassignment requests were the top inbound ticket type. Automating ownership handoff could eliminate the majority.
~40% reduction in orphaned file incidents
Stakeholders estimated at least 2 in 5 offboarded employees left behind content without reassignment.
80% faster access to business-critical files
Delays often took 2–3 weeks; tool would reduce access times to under 2 days by removing manual approval chains.
Estimated time savings of 6–10 hours per offboarded employee
Time otherwise spent by managers + IT tracking down, reassessing, and re-permissioning old content.
Outcome & Reflection
SCOTT was met with strong enthusiasm across security, IT, and legal stakeholders—especially for its potential to automate high-risk offboarding steps and reduce support load.
Although I left before development began, the handoff was smooth and the design was used to kick off implementation. Based on our research and stakeholder input, we anticipate the tool would significantly reduce manual workload, orphaned file incidents, and access delays.
The project reinforced the value of designing for edge cases, compliance constraints, and long-term systems thinking—hallmarks of enterprise UX at scale.
