S3: Institutional MLS Disposition Platform
Driving Institutional Real Estate Sales Across 70+ Markets
Through a Streamlined MLS Workflow
3,500
Transactions since launch
11,500,000
Total revenue up to date
Duration: MVP – 11 months
Revenue: Approximately 6 million in Institutional property sales (2025)
Team: Product Manager, Data Manager, Software Engineers (15), QA Engineer.
My Company
Roofstock is an online marketplace for buying and selling tenant-occupied rental homes. It connects investors to U.S. properties with data, analytics, financing, management, and due diligence, making real estate investing easier and more accessible.
My Role
I’m a Principal Product Designer at Roofstock, leading the design of tools that power institutional real estate sales at scale. I previously managed the Institutional design team but returned to an IC role to focus on product strategy and execution.
Let’s talk about the product. And me again.
Scaled Seller Services
S3 (Scaled Seller Services) is a task-based platform that streamlines the MLS disposition workflow for institutional property sales. S3 centralizes the entire sales cycle into one interface, enabling sellers, agents, and internal disposition managers to collaborate and track progress from list to close. The platform was built to support a lean, scalable operation — enabling Roofstock to manage high-volume transactions with reduced overhead and minimal manual coordination.
I’ve led S3 from its inception over three years ago through every iteration since — from discovery and user interviews to design, QA, and launch. This product evolved through continuous testing and feedback, and I remained hands-on throughout. Today, S3 is one of the company’s most profitable tools, driving institutional revenue and streamlining internal workflows at scale.
Why we did what we did
The North Star
Reduce time to disposition by centralizing all stakeholder workflows — from seller approvals to offer management — into one task-driven platform.
- Internal efficiency: Less manual work = Faster handoffs
- Seller satisfaction: Fewer delays = Better experience
- Revenue acceleration: Homes sold faster = Faster returns
- S3 wasn’t born out of blue-sky innovation — it was a direct response to operational pain. Before S3, the disposition team relied on Airtable checklists and filters to manage property workflows. What began as flexible quickly became a bottleneck — too many users, limited scalability, poor tracking, and endless workarounds. S3 was built to replace Airtable with a purpose-built platform that not only matched the flow but expanded it.
Our goals were:
- Eliminate Airtable and reduce ops overhead
- Introduce a structured task queue system that could span multiple personas
- Cover the full lifecycle of a property sale — from seller approval, to BPO, to listing, to offer management
- Create a single source of truth that removed dependence on Slack threads, spreadsheets, and email trails
- Scale without scaling headcount by automating task assignments, due dates, and notifications
- Allow team and client-level configurability to adapt the system to different operational models
- This was not just a tool to organize tasks — S3 is an end-to-end transaction engine designed to support volume, complexity, and client flexibility.
The users
Persona
Seller
- Goal: Maximize sale value with minimal operational burden
- Responsibilities: Reviewing BPOs, approving list prices, selecting make-ready work, negotiating offers
- Pain Points: Lack of visibility, unclear next steps, delays in offer or pricing decisions
- Design Needs: A streamlined task-based UI with clear approvals, timelines, and offer summaries
Disposition Manager
- Goal: Move properties through the disposition pipeline efficiently and accurately
- Responsibilities: Assigning brokers, reviewing BPOs, managing timelines, pushing listings forward, flagging delays
- Pain Points: Too many manual touchpoints, tracking across systems, high-volume pressure
- Design Needs: Centralized dashboards, real-time task queues, alerting systems, and override tool
Agent
- Goal: List and sell institutional properties efficiently
- Responsibilities: Submit BPOs, create listings, report weekly activity, upload offers
- Pain Points: Repetitive communications, unclear expectations, disconnected data entry
- Design Needs: A lightweight portal to submit property-level data, get notified of assigned tasks, and stay aligned with Roofstock ops
The steps we took
01 / Discovery & Alignment
Stakeholder and user interviews were conducted, operational tools including Airtable, Notion, spreadsheets, and Slack workflows were audited, and pain points and inefficiencies across the disposition lifecycle were mapped.
02 / Persona + Workflow Modeling
Three personas with distinct user flows were defined and validated, two disposition models (seller-driven and full-service) were mapped, and role-specific task flows were built to dynamically update with property status.
03 / Interface Design
Modular task UIs, dashboards, and submission flows were designed with a focus on visual hierarchy, clarity, and role awareness, while collaboration with engineering ensured scalable, permissioned front-ends.
04 / Testing & Iteration
Used Maze to test task completion, dashboard comprehension, and terminology, followed by scenario-based walk-throughs with brokers and internal teams, leading to iterations on labeling, flow logic, and notification triggers.
05 / Launch + Continuous Updates
The rollout was executed through phased client onboarding, with live feedback loops integrated to surface adoption gaps, and new functionality added in response to market and team demand.
Testable assumption
Hypothesis
If we provide institutional and individual property sellers with a centralized, technology-driven platform that streamlines listing, pricing, and transaction management at scale, then they will choose Roofstock S3 over traditional brokerage or fragmented processes to dispose of their assets, because it reduces time-to-market, ensures competitive pricing insights, and simplifies operational complexity, enabling them to maximize returns while minimizing effort.
The tests we ran
The Validation
To validate our assumptions and surface usability issues, we ran tests through Maze and internal pilots:
- Seller Flows – tested task completion rates for list price approvals, offer decisions, and make-ready selections
- Broker BPO Flow – observed drop-offs and frustration around multi-device form completion
- Dashboard Navigation – click maps and heatmaps used to refine layout and content order
- Terminology Testing – ran preference tests for internal vs. external language clarity
- Post-launch Feedback Loops – monitored production usage and support tickets for areas of friction
Let’s see the work
Iterative Delivery
S3 covers a lot of workflows, so rather than showing everything at once, I can walk you through any specific flow you’re interested in.
