Decoding Rental Decisions: The UX Research That Reshaped Strategy
📖 Background
This project was created while working at a company that designs robotic furniture for modular apartments. Initially focused on robotic systems, the company was exploring ways to expand its focus from robotic furniture to complete living space solutions. These apartments are smaller than average, but the modular design makes them feel more spacious and adaptable.
Because this was a fast-paced startup, advocating for research was a challenge, especially when focus was often on short term "wins". However, I knew that in order to succeed, we needed to deeply understand our users. To align stakeholders, I posed a key question: “Are we so focused on short-term momentum that we're making irreversible mistakes now—ones that could ultimately prevent us from achieving our long-term vision?”
With this in mind, we took a closer look under the hood.
Because this was a fast-paced startup, advocating for research was a challenge, especially when focus was often on short term "wins". However, I knew that in order to succeed, we needed to deeply understand our users. To align stakeholders, I posed a key question: “Are we so focused on short-term momentum that we're making irreversible mistakes now—ones that could ultimately prevent us from achieving our long-term vision?”
With this in mind, we took a closer look under the hood.
🤔 Understanding the Core Challenge
Apartment renters sought a valuable living experience, while developers wanted a product that would sell. However, the company was making assumptions about the key value of their apartments without concrete facts. This led me to investigate cause vs. correlation in why people were renting from the list of apartments that the company offered. To explore this, I broke down the possible factors influencing rental decisions and systematically tested them:
Key Variables examples:
1️⃣ Potential contributing factors (Independent Variables):
- Lower rent compared to competitors
- Modern building with amenities
- Robotic/modular furniture increasing functional space
2️⃣ Effect (Dependent Variable):
- Renters signing leases
3️⃣ Possible Correlations (Non-Causal Relationships):
- Renters in a certain age group may be more likely to rent
- Cities with higher rent may have higher demand for these units
Key Variables examples:
1️⃣ Potential contributing factors (Independent Variables):
- Lower rent compared to competitors
- Modern building with amenities
- Robotic/modular furniture increasing functional space
2️⃣ Effect (Dependent Variable):
- Renters signing leases
3️⃣ Possible Correlations (Non-Causal Relationships):
- Renters in a certain age group may be more likely to rent
- Cities with higher rent may have higher demand for these units
📋 Data Collection Approach
To determine cause vs. correlation, I lead the charge on qualitative research, while quantitative methods were distributed to other team members.
The objective was to collect and analyze qualitative data to evaluate the assumptions made by the team prior to conducting the research. Though causation experiments were out of scope, my focus was on trying to establish possible correlations between different variables (semi-furnished, amenities, price, location, apartment size, etc.) to better define the needs of our users.
1️⃣ 1:1 Interviews
The objective was to collect and analyze qualitative data to evaluate the assumptions made by the team prior to conducting the research. Though causation experiments were out of scope, my focus was on trying to establish possible correlations between different variables (semi-furnished, amenities, price, location, apartment size, etc.) to better define the needs of our users.
1️⃣ 1:1 Interviews
Screen shot taken from interview with current resident
2️⃣ Behavioral Data & Experiments:
- A/B Testing on Listings
- Virtual reality apartment tours
- A/B Testing on Listings
- Virtual reality apartment tours
Image from A/B listings test
Still image from virtual apartment tour
👨💻 Examples of Correlation
- If lowering rent increased signings across all variations, price was a correlation.
- If renters repeatedly talked about amenities, then amenities could be a correlation.
- If renters repeatedly talked about amenities, then amenities could be a correlation.
🕵️♂️ The Research Execution: Scientist, Flight Attendant, Sportscaster
For this study, I leaned on a concept from UX designer Jared Spool, who described how the best usability test moderators balance three personas: the scientist, the flight attendant, and the sportscaster.
👨🔬 The Scientist: Guiding the data collection
I structured the research by defining clear objectives and methodologies:
Target Audience:
1️⃣ Current residents
2️⃣ Non-residents fitting our consumer criteria. For this group I created a screener and recruited twenty participants through userinterviews.com
Methodology:
Remote, moderated, "think out loud" environments. All sessions were video and audio recorded as I guided tasks and questions for two test groups.
1️⃣ Current Residents
- Conducted 1:1 tests/interviews while residents walked through their apartments, sharing their experiences.
2️⃣ Non Residents:
- I built a mid-fidelity apartment listing site for A/B testing.
- Shared virtual walkthrough of apartments (similar to Sims or first-person video games).
* Virtual walkthroughs were designed and built in collaboration with an external design studio and the rest of the team.
Test Environment & Scenarios:
- Developed a structured test outline and script with guided questions and onboarding introductions.
- Created realistic user tasks for participants to carry out
- Defined success metrics and analysis methods.
👨🔬 The Scientist: Guiding the data collection
I structured the research by defining clear objectives and methodologies:
Target Audience:
1️⃣ Current residents
2️⃣ Non-residents fitting our consumer criteria. For this group I created a screener and recruited twenty participants through userinterviews.com
Methodology:
Remote, moderated, "think out loud" environments. All sessions were video and audio recorded as I guided tasks and questions for two test groups.
1️⃣ Current Residents
- Conducted 1:1 tests/interviews while residents walked through their apartments, sharing their experiences.
2️⃣ Non Residents:
- I built a mid-fidelity apartment listing site for A/B testing.
- Shared virtual walkthrough of apartments (similar to Sims or first-person video games).
* Virtual walkthroughs were designed and built in collaboration with an external design studio and the rest of the team.
Test Environment & Scenarios:
- Developed a structured test outline and script with guided questions and onboarding introductions.
- Created realistic user tasks for participants to carry out
- Defined success metrics and analysis methods.
✈️ The Flight Attendant: Ensuring Comfort
User testing can be stressful—meeting someone for the first time, being recorded, performing tasks, and answering questions. My role was to ensure participants felt comfortable and safe. I used:
Ice breakers to make the session feel relaxed
Encouragement and reassurance to reduce anxiety
Attentive monitoring to pivot when necessary and support participants through difficult tasks
✈️ The Flight Attendant: Ensuring Comfort
User testing can be stressful—meeting someone for the first time, being recorded, performing tasks, and answering questions. My role was to ensure participants felt comfortable and safe. I used:
Ice breakers to make the session feel relaxed
Encouragement and reassurance to reduce anxiety
Attentive monitoring to pivot when necessary and support participants through difficult tasks
🎤 The Sportscaster: Narrating the Action
During testing, I actively facilitated the sessions to:
- Encourage participants to think out loud
- Narrate user actions to help shy participants find their rhythm
- Reel in off-topic discussions when necessary
- Ask probing questions: "Why do you feel that way?" "Would you still choose this option if X were different?"
Knowing the company's internal hurdles, I was able to highlight input about difficult company decisions, ensuring insights were directly applicable to strategic pivots.
🎤 The Sportscaster: Narrating the Action
During testing, I actively facilitated the sessions to:
- Encourage participants to think out loud
- Narrate user actions to help shy participants find their rhythm
- Reel in off-topic discussions when necessary
- Ask probing questions: "Why do you feel that way?" "Would you still choose this option if X were different?"
Knowing the company's internal hurdles, I was able to highlight input about difficult company decisions, ensuring insights were directly applicable to strategic pivots.
🥹 Results & Impact
By the third test, we had already disproved major company hypotheses. This research:
1️⃣ Quickly revealed insights that challenged initial assumptions.
2️⃣ Identified key pillars that had previously been overlooked.
3️⃣ Provided clear insights into what truly influenced rental decisions—allowing the company to refine its product offering.
The findings became a fundamental catalyst for our playbook, aligning research insights with the startup’s fast-paced timeline, while providing a much needed introspective breath of fresh air.
1️⃣ Quickly revealed insights that challenged initial assumptions.
2️⃣ Identified key pillars that had previously been overlooked.
3️⃣ Provided clear insights into what truly influenced rental decisions—allowing the company to refine its product offering.
The findings became a fundamental catalyst for our playbook, aligning research insights with the startup’s fast-paced timeline, while providing a much needed introspective breath of fresh air.
