Industry
Auto Finance
Client
General Motors Financial
Instant car appraisal from anywhere
Project Overview
In 2022, the company explored a new proof of concept to simplify the vehicle appraisal and selling process through a digital experience. Traditionally, people had to visit a dealership to get their car appraised—an experience often seen as time-consuming, inconvenient, and sales-driven.
The proposed solution leveraged AI-powered photo recognition to detect visible vehicle damage (such as dings or cracks) from user-submitted photos. The goal was to instantly generate a fair offer—removing the need for an in-person visit and empowering users to complete the process remotely.
Early on, there was a split among stakeholders.
Some pushed for a mobile app, citing deeper native capabilities and future scalability.
Others advocated for a web-based experience to reduce friction and increase accessibility.
To guide the decision, we conducted user interviews and a broader survey to validate user expectations around platform preferences.
Project Overview
In 2022, the company explored a new proof of concept to simplify the vehicle appraisal and selling process through a digital experience. Traditionally, people had to visit a dealership to get their car appraised—an experience often seen as time-consuming, inconvenient, and sales-driven.
The proposed solution leveraged AI-powered photo recognition to detect visible vehicle damage (such as dings or cracks) from user-submitted photos. The goal was to instantly generate a fair offer—removing the need for an in-person visit and empowering users to complete the process remotely.
Early on, there was a split among stakeholders.
Some pushed for a mobile app, citing deeper native capabilities and future scalability.
Others advocated for a web-based experience to reduce friction and increase accessibility.
To guide the decision, we conducted user interviews and a broader survey to validate user expectations around platform preferences.
Project Overview
In 2022, the company explored a new proof of concept to simplify the vehicle appraisal and selling process through a digital experience. Traditionally, people had to visit a dealership to get their car appraised—an experience often seen as time-consuming, inconvenient, and sales-driven.
The proposed solution leveraged AI-powered photo recognition to detect visible vehicle damage (such as dings or cracks) from user-submitted photos. The goal was to instantly generate a fair offer—removing the need for an in-person visit and empowering users to complete the process remotely.
Early on, there was a split among stakeholders.
Some pushed for a mobile app, citing deeper native capabilities and future scalability.
Others advocated for a web-based experience to reduce friction and increase accessibility.
To guide the decision, we conducted user interviews and a broader survey to validate user expectations around platform preferences.
Research
We used a mixed-methods research approach to gather both qualitative and quantitative insights. This included in-depth interviews with 20 participants and a survey completed by 100 participants who were either selling their vehicle or exploring the idea.
🔍 User Interviews
Participants included:
10 People who had previously gone to a dealership for an appraisal
10 People interested in selling their vehicle online or through an app
Key insights:
Most users described the dealership experience as slow, sales-driven, and uncomfortable
There was overwhelming preference for using a smartphone over a laptop for submitting photos
Participants voiced a clear aversion to downloading an app, especially for a single-use task
Trust in AI was mixed but improved with clear explanation of how the offer would be calculated
Research
We used a mixed-methods research approach to gather both qualitative and quantitative insights. This included in-depth interviews with 20 participants and a survey completed by 100 participants who were either selling their vehicle or exploring the idea.
🔍 User Interviews
Participants included:
10 People who had previously gone to a dealership for an appraisal
10 People interested in selling their vehicle online or through an app
Key insights:
Most users described the dealership experience as slow, sales-driven, and uncomfortable
There was overwhelming preference for using a smartphone over a laptop for submitting photos
Participants voiced a clear aversion to downloading an app, especially for a single-use task
Trust in AI was mixed but improved with clear explanation of how the offer would be calculated
Research
We used a mixed-methods research approach to gather both qualitative and quantitative insights. This included in-depth interviews with 20 participants and a survey completed by 100 participants who were either selling their vehicle or exploring the idea.
User Interviews
Participants included:
10 People who had previously gone to a dealership for an appraisal
10 People interested in selling their vehicle online or through an app
Key insights:
Most users described the dealership experience as slow, sales-driven, and uncomfortable
There was overwhelming preference for using a smartphone over a laptop for submitting photos
Participants voiced a clear aversion to downloading an app, especially for a single-use task
Trust in AI was mixed but improved with clear explanation of how the offer would be calculated
📊 Online Survey (100 participants)
We surveyed 100 individuals who had recently sold or were considering selling their vehicles to validate our qualitative findings at scale.
🧨 Pain Points
82% (82 out of 100) said dealership appraisals took longer than expected
67% (67 out of 100) felt pressured by salespeople
58% (58 out of 100) were unclear how their car’s value was calculated
💡 Platform & Device Preferences
76% (76 out of 100) said they preferred not having to download an app
71% (71 out of 100) said they would most likely use their smartphone for the appraisal process
84% (84 out of 100) wanted to start immediately, without needing to sign up or download anything
Online Survey (100 participants)
We surveyed 100 individuals who had recently sold or were considering selling their vehicles to validate our qualitative findings at scale.
🧨 Pain Points
82% (82 out of 100) said dealership appraisals took longer than expected
67% (67 out of 100) felt pressured by salespeople
58% (58 out of 100) were unclear how their car’s value was calculated
💡 Platform & Device Preferences
76% (76 out of 100) said they preferred not having to download an app
71% (71 out of 100) said they would most likely use their smartphone for the appraisal process
84% (84 out of 100) wanted to start immediately, without needing to sign up or download anything
Designs
While users said they’d use their phone to take photos and complete the appraisal, we learned they often started their search on a laptop or desktop to check out the company.
So we designed the marketing site for desktop first and tested it to make sure it clearly explained the service and built trust before users switched to mobile.
We ran 20 moderated usability sessions to test the desktop marketing experience—including the homepage, “How It Works,” and Help Center pages. Sessions focused on clarity, trust, and whether users felt confident enough to continue the appraisal on their phone.
Homepage
The homepage immediately positions the service as instant, remote, and stress-free with messaging like: “Instant vehicle appraisal – Score an offer without stepping out of your home. We’ll even grab it for you!”.
This directly responds to users’ top pain points—wasting time and feeling pressured at dealerships. We chose benefit-forward language to reinforce convenience and autonomy.
Throughout the site, we emphasized “Zero Friction,” “Instant Everything,” and “Human Touch”, with strategically placed CTAs like “Get Offer” and “Try it Now ↓”.
We learned users were motivated by speed, ease, and avoiding confrontation. The CTAs and supporting copy were crafted to lower emotional and cognitive effort—especially for first-time sellers.
Homepage
The homepage immediately positions the service as instant, remote, and stress-free with messaging like: “Instant vehicle appraisal – Score an offer without stepping out of your home. We’ll even grab it for you!”.
This directly responds to users’ top pain points—wasting time and feeling pressured at dealerships. We chose benefit-forward language to reinforce convenience and autonomy.
Throughout the site, we emphasized “Zero Friction,” “Instant Everything,” and “Human Touch”, with strategically placed CTAs like “Get Offer” and “Try it Now ↓”.
We learned users were motivated by speed, ease, and avoiding confrontation. The CTAs and supporting copy were crafted to lower emotional and cognitive effort—especially for first-time sellers.
Designs
While users said they’d use their phone to take photos and complete the appraisal, we learned they often started their search on a laptop or desktop to check out the company.
So we designed the marketing site for desktop first and tested it to make sure it clearly explained the service and built trust before users switched to mobile.
We ran 20 moderated usability sessions to test the desktop marketing experience—including the homepage, “How It Works,” and Help Center pages. Sessions focused on clarity, trust, and whether users felt confident enough to continue the appraisal on their phone.
How it Works
The “How It Works” page breaks down the process into 5 steps, with language like: "AI analyzes your car... our algorithm analyzes 1000+ data points..." "No negotiations, no surprises."
Users wanted clarity on how offers were calculated. By making the AI logic visible and easy to understand, we addressed transparency concerns and built trust.
Each step is broken into digestible blocks. We included time estimates (e.g., “<1 minute”, “2–3 min”) for realism and confidence.
Participants said they were unsure how long each part would take or if they were doing it right. These visual cues demystify the process and reduce hesitation.
How it Works
The “How It Works” page breaks down the process into 5 steps, with language like: "AI analyzes your car... our algorithm analyzes 1000+ data points..." "No negotiations, no surprises."
Users wanted clarity on how offers were calculated. By making the AI logic visible and easy to understand, we addressed transparency concerns and built trust.
Each step is broken into digestible blocks. We included time estimates (e.g., “<1 minute”, “2–3 min”) for realism and confidence.
Participants said they were unsure how long each part would take or if they were doing it right. These visual cues demystify the process and reduce hesitation.
How it Works
The “How It Works” page breaks the appraisal process into 5 simple steps, using clear, approachable language like:
“AI analyzes your car... No negotiations, no surprises.”
This transparency addressed a major trust gap—users wanted to understand how offers were calculated, and how long each step would take. To build confidence, we included realistic time estimates (“<1 minute”, “2–3 min”) and kept each step visually digestible and linear.
What We Tested:
Clarity of the 5-step explanation
Trust in the AI-based offer process
Comprehension of each step and expected time
Whether the flow reduced uncertainty or friction
Results:
16 out of 20 (80%) said the 5-step breakdown made the process feel easy and understandable
17 out of 20 (85%) said they trusted the AI appraisal more after reading this section
15 out of 20 (75%) said time estimates gave them more confidence to move forward
0 participants reported confusion about what to expect next
Help Center
The Help Center was designed to address common user anxieties like:
“What documents do I need to sell my car?” and “When will I get paid?”
These questions surfaced frequently during interviews. To reduce friction, we grouped FAQs into clear categories and kept answers brief, plainspoken, and reassuring—avoiding legalese or overly detailed fine print.
What We Tested:
Ability to quickly find relevant questions
Clarity and tone of answers
Whether content helped build trust
Perception of the brand’s professionalism
Results:
15 out of 20 (75%) said the Help Center made the service feel more trustworthy
16 out of 20 (80%) found the answers easy to understand and not overwhelming
14 out of 20 (70%) said they’d check this section first if they had questions
Participants called out the tone as “helpful,” “friendly,” and “clear”
Help Center
The Help Center was designed to address common user anxieties like:
“What documents do I need to sell my car?” and “When will I get paid?”
These questions surfaced frequently during interviews. To reduce friction, we grouped FAQs into clear categories and kept answers brief, plainspoken, and reassuring—avoiding legalese or overly detailed fine print.
What We Tested:
Ability to quickly find relevant questions
Clarity and tone of answers
Whether content helped build trust
Perception of the brand’s professionalism
Results:
15 out of 20 (75%) said the Help Center made the service feel more trustworthy
16 out of 20 (80%) found the answers easy to understand and not overwhelming
14 out of 20 (70%) said they’d check this section first if they had questions
Participants called out the tone as “helpful,” “friendly,” and “clear”
Help Center
The Help Center was designed to address common user anxieties like:
“What documents do I need to sell my car?” and “When will I get paid?”
These questions surfaced frequently during interviews. To reduce friction, we grouped FAQs into clear categories and kept answers brief, plainspoken, and reassuring—avoiding legalese or overly detailed fine print.
What We Tested:
Ability to quickly find relevant questions
Clarity and tone of answers
Whether content helped build trust
Perception of the brand’s professionalism
Results:
15 out of 20 (75%) said the Help Center made the service feel more trustworthy
16 out of 20 (80%) found the answers easy to understand and not overwhelming
14 out of 20 (70%) said they’d check this section first if they had questions
Participants called out the tone as “helpful,” “friendly,” and “clear”
Outcome
The research directly shaped how we approached both the content and structure of the experience. What started as a basic AI-driven idea turned into a clear, trustworthy flow that felt easy and approachable for users.
After testing the homepage, How It Works, and Help Center, 90% of participants said they’d feel confident moving forward with the appraisal. The mix of straightforward messaging, visual step-by-step breakdowns, and short, helpful answers really helped ease hesitation—especially for first-time sellers.
These results gave the team confidence to move ahead with a mobile-first web pilot instead of building a native app. It saved time, cut down on costs, and got us to launch faster. Final designs were handed off to other product designers to keep building out the experience.
In the end, we landed on a solution that made sense for both users and the business—simple, fast, and built around real needs.
Outcome
The research directly shaped how we approached both the content and structure of the experience. What started as a basic AI-driven idea turned into a clear, trustworthy flow that felt easy and approachable for users.
After testing the homepage, How It Works, and Help Center, 90% of participants said they’d feel confident moving forward with the appraisal. The mix of straightforward messaging, visual step-by-step breakdowns, and short, helpful answers really helped ease hesitation—especially for first-time sellers.
These results gave the team confidence to move ahead with a mobile-first web pilot instead of building a native app. It saved time, cut down on costs, and got us to launch faster. Final designs were handed off to other product designers to keep building out the experience.
In the end, we landed on a solution that made sense for both users and the business—simple, fast, and built around real needs.
Outcome
The research directly shaped how we approached both the content and structure of the experience. What started as a basic AI-driven idea turned into a clear, trustworthy flow that felt easy and approachable for users.
After testing the homepage, How It Works, and Help Center, 90% of participants said they’d feel confident moving forward with the appraisal. The mix of straightforward messaging, visual step-by-step breakdowns, and short, helpful answers really helped ease hesitation—especially for first-time sellers.
These results gave the team confidence to move ahead with a mobile-first web pilot instead of building a native app. It saved time, cut down on costs, and got us to launch faster. Final designs were handed off to other product designers to keep building out the experience.
In the end, we landed on a solution that made sense for both users and the business—simple, fast, and built around real needs.