Chen - Portfolio

hallochen

autoscout24

Dealer Area Redesign; Boosting revenue-driving engagement by 150%

my role

E2E Ownership - Dealers Platform

setup

Centralised Team / Kanban

Time frame

Q4 2024 - Q2 2025

AutoScout24 is Europe’s leading automotive marketplace.

Its Dealer Area is responsible for 95% of org revenue. It is where car dealers execute most of their digital daily tasks.

Problem

The dealer platform fails to deliver clear value to inventory and sales decision-makers, as fragmented workflows, scattered data, and unclear impact limit action and engagement across accounts of all sizes.

70%

Struggle sourcing vehicles due to lack of effective tools and reliable data sources

60%

Go through the same daily basic actions, without exploring impactful features

2.5%

Smart features engagement - 60% lower ROI than expected

Objective
Re-engage DMS-using accounts, by giving them valuable, strategic insights, while empowering smaller dealers with actionable insights to compete effectively. Increase overall engagement across both groups.

Assessing the Problem
We mapped out the user journeys of 8 different markets, with 9 possible users groups, each having up to 5 roles

Objective & Approach

Instead of forcing one UX on all dealers, we design around the idea of progressive relevance. I defined the core personas, then optimised taxonomy for the most universally relevant insights appear first, with deeper layers available when needed.

Cross-Functional Ideation & Testing

I collaborated with functions to understand business needs, prios and constraints. Testing ran with 9 groups of 5 testers from multiple markets.

Driving AI Innovation
With data science & engineering, I mapped AS24’s tech strengths, from listing data models to content generation & identified where AI adds value. This made a clear business case: AI could cut workload, boost listing quality, & strengthen AS24’s position.
Outcome
Enabled the first AI-assisted listing optimization prototype, reducing manual listing time by over 30% in early tests.

A special focus was put on micto-interactions using Jitter and Lottie

Release

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimizing risk.

Desktop - Multi Branch, Sales User

Overview tagged per persona

Mobile - Smart Insertion and Optimisation.

Integrating AI in My Workflow
To move fast and explore broadly, GPT became a thinking partner for concept framing, UX copy, and user flow simulation, while Lovable helped me turn prompts into interactive prototypes within hours.
Outcome
Accelerated design iteration cycles by ~40% and helped define AI-driven UX guidelines later adopted by the wider team.

case studies

AutoScout24

Express Steuer

Yilu

Lufthansa

Home

AutoScout24

ExpressSteuer

Yilu

hallochen

autoscout24

Dealer Area Redesign; Boosting revenue-driving engagement by 150%

my role

E2E Ownership - Dealers Platform

setup

Centralised Team /

Kanban

Time frame

Q4 2024 - Q2 2025

AutoScout24 is Europe’s leading automotive marketplace.

Its Dealer Area is responsible for 95% of org revenue. It is where car dealers execute most of their digital daily tasks.

Problem

The dealer platform fails to deliver clear value to inventory and sales decision-makers, as fragmented workflows, scattered data, and unclear impact limit action and engagement across accounts of all sizes.

70%

Small to Medium dealerships reported consistantkly increasing

60%

Of dealers understand the value or impact AutoScout24 of "smart features”

2.5%

Drop after the result step, & a similar drop on the following step

Objective
Re-engage DMS-using accounts, by giving them valuable, strategic insights, while empowering smaller dealers with actionable insights to compete effectively. Increase overall engagement across both groups.

Assessing the problem
We mapped out the user journeys of 8 different markets, with 9 possible users groups, each having up to 5 roles

Objective & Approach

Instead of forcing one UX on all dealers, we design around the idea of progressive relevance. I defined the core personas, then optimised taxonomy for the most universally relevant insights appear first, with deeper layers available when needed.

Cross-Functional Ideation & Concept Testing

I collaborated with functions to understand business needs, prios and constraints. Together we listed ideas up for testing. Testing ran with 9 groups of 5 testers from multiple markets.

Driving AI Innovation
With data science & engineering, I mapped AS24’s tech strengths, from listing data models to content generation, & identified where AI adds value. This made the business case clear: AI could cut workload, boost listing quality, and strengthen AS24’s position.
Outcome
Enabled the first AI-assisted listing optimization prototype, reducing manual listing time by over 30% in early tests.

A special focus was put on micto-interactions using Jitter and Lottie

New components were documented in a new UI library

Release

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimizing risk.

Desktop - Multi Branch, Sales User

Overview tagged per persona

Mobile - Smart Insertion and Optimisation.

Prioritising the most relevant tasks for mobile users.

Integrating AI in My Workflow
To move fast and explore broadly, ChatGPT became a thinking partner for concept framing, UX copy, and user flow simulation, while Lovable helped me turn prompts into interactive prototypes within hours.
Outcome
Accelerated design iteration cycles by ~40% and helped define AI-driven UX guidelines later adopted by the wider team.

case studies

AutoScout24

Express Steuer

Yilu

Lufthansa

Home

AutoScout24

ExpressSteuer

Yilu

hallochen

autoscout24

Dealer Area Redesign; Boosting revenue-driving engagement by 150%

my role

Dealers Platform — E2E Ownership

setup

Centralised Team / Kanban

Time frame

Q4 2024 - Q2 2025

AutoScout24 is Europe’s leading automotive marketplace.

Its Dealer Area is responsible for 95% of org revenue. It is where dealers execute most of their daily digital tasks.

Problem

The dealer platform fails to deliver clear value to inventory and sales decision-makers, as fragmented workflows, scattered data, and unclear impact limit action and engagement across accounts of all sizes.

70%

Small to Medium dealerships reported consistently decreasing leads..

60%

Go through the same tasks & flows, avoiding exploration of business driving features or insights.

2.5%

Engagement with new, smart features (-50% vs benchmarks), due to lack of trust in their impact.

Objective
Re-engage DMS-using accounts, by giving them valuable, strategic insights, while empowering smaller dealers with actionable insights to compete effectively. Increase overall engagement across both groups.

The main challenge
Inconsistent dealer workflows due to massive differences in dealership size and maturity. Their needs, level of expertise, workflows, and expectations are radically different.

Approach

Instead of forcing one UX on all users, we design around progressive relevance. With the research team, I defined the core personas, then optimised taxonomy for the most relevant insights appear first, with deeper layers available when needed.

Cross-Functional Ideation & Concept Testing

I collaborated with functions to understand business needs and constraints. We listed ideas up for testing with 9 groups of 5 testers from multiple markets.

Driving AI Innovation
With data science & engineering, I mapped AS24’s tech strengths, from listing data models to content generation, & identified where AI adds value. This made the business case clear: AI could cut workload, boost listing quality, and strengthen AS24’s position.
Outcome
Enabled the first AI-assisted listing optimization prototype, reducing manual listing time by over 30% in early tests.

A special focus was put on micro-interactions using Jitter and Lottie

New components were documented in a new UI library

Release

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimising risk.

Desktop - Multi Branch, Sales User

Overview tagged per persona

Mobile - Smart Insertion and Optimisation.

Prioritising the most relevant tasks for mobile users.

Integrating AI in My Workflow
To move fast and explore broadly, ChatGPT became a thinking partner for concept framing, UX copy, and user flow simulation, while Lovable helped me turn prompts into interactive prototypes within hours.
Outcome
Accelerated design iteration cycles by ~40% and helped define AI-driven UX guidelines later adopted by the wider team.

case studies

AutoScout24

Express Steuer

Yilu

Lufthansa

Objective

Re-engage DMS-using accounts, by giving them valuable, strategic insights, while empowering smaller dealers with actionable insights to compete effectively. Increase overall engagement across both groups.

Assessing the Problem

We mapped out the user journeys of 8 different markets, with 9 possible users groups, each having up to 5 roles

Cross-Functional Ideation & Testing

I collaborated with functions to understand business needs, prios and constraints. Testing ran with 9 groups of 5 testers from multiple markets.

Driving AI Innovation

With data science & engineering, I mapped AS24’s tech strengths, from listing data models to content generation & identified where AI adds value. This made a clear business case: AI could cut workload, boost listing quality, & strengthen AS24’s position.

Outcome

Enabled the first AI-assisted listing optimization prototype, reducing manual listing time by over 30% in early tests.

A special focus was put on micto-interactions using Jitter and Lottie

A special focus was put on micto-interactions using Jitter and Lottie

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimizing risk.

Integrating AI in My Workflow

To move fast and explore broadly, GPT became a thinking partner for concept framing, UX copy, and user flow simulation, while Lovable helped me turn prompts into interactive prototypes within hours.

Outcome

Accelerated design iteration cycles by ~40% and helped define AI-driven UX guidelines later adopted by the wider team.

Objective

Re-engage DMS-using accounts, by giving them valuable, strategic insights, while empowering smaller dealers with actionable insights to compete effectively. Increase overall engagement across both groups.

Assessing the problem

We mapped out the user journeys of 8 different markets, with 9 possible users groups, each having up to 5 roles

Cross-Functional Ideation & Concept Testing

I collaborated with functions to understand business needs, prios and constraints. Together we listed ideas up for testing. Testing ran with 9 groups of 5 testers from multiple markets.

Driving AI Innovation

With data science & engineering, I mapped AS24’s tech strengths, from listing data models to content generation, & identified where AI adds value. This made the business case clear: AI could cut workload, boost listing quality, and strengthen AS24’s position.

Outcome

Enabled the first AI-assisted listing optimization prototype, reducing manual listing time by over 30% in early tests.

A special focus was put on micto-interactions using Jitter and Lottie

New components were documented in a new UI library

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimizing risk.

Integrating AI in My Workflow

To move fast and explore broadly, ChatGPT became a thinking partner for concept framing, UX copy, and user flow simulation, while Lovable helped me turn prompts into interactive prototypes within hours.

Outcome

Accelerated design iteration cycles by ~40% and helped define AI-driven UX guidelines later adopted by the wider team.

Cross-Functional Ideation & Concept Testing

A special focus was put on micro-interactions using Jitter and Lottie

New components were documented in a new UI library

We launched three UIs in phases, to address the main problems. With product & engineering we defined scope, dependencies, and rollout, ensuring quality while minimising risk.