Case Study · AutoScout24 · Search & Discovery

New Search

New Search reset how buyers narrow complex car inventory across app and web. I led the app filter direction and helped shape the cross-platform baseline after earlier validation exposed a weak search model. The work gave buyers a clearer way to scan, refine, and return to results across one of AutoScout24's highest-value journeys.

Role: Principal Product Designer leading the app search/filter direction across a staged app-and-web reset
Scope: iOS, Android, desktop, mWeb
Signal: Mixed earlier validation led to a stronger baseline
Proof: Research, experiments, and platform-specific rollout evidence

Weak direction reset

Resetafterevidence

Earlier app validation exposed a weak direction. We rebuilt the baseline around a scan-then-dive filter model: clearer grouping, reusable inputs, and reviewable selections before buyers returned to results.

iOS benchmark and rollout

iOSsignal

iOS gave the team its cleanest validation signal: about +6% leads from the list page, with detail-page engagement stable, before broad rollout.

Platform judgment

Adaptedbysurface

Android and web extended the same product model with narrower evidence. We adapted save behaviour and rollout scope by platform, then shipped web as a simpler baseline the team could build on.

Overview

A staged reset, not one clean launch

This initiative moved through different levels of evidence by platform, so the case study works best as one programme around what buyers were looking for, filter discovery, and return paths rather than one uniform redesign story.

A weaker app direction forced the reset. iOS became the first clear validation point, Android moved through research and validation, and web launched later as a narrower baseline.

That sequencing mattered. The goal was not to force identical behaviour across surfaces, but to establish a search model each platform could support credibly and then ship where the evidence was strong enough.

Because the full performance picture is confidential, I focus on evidence I can share: research, experimentation, and post-launch analysis showed positive movement in key commercial progression signals on the strongest surfaces while downstream engagement stayed broadly stable.

Reset at a glance

Desktop results page before the New Search reset, with a longer filter rail and weaker visual hierarchy. — Longer filter rail and weaker hierarchy on results.

Desktop results page after the New Search reset, with a clearer filter model and stronger applied-state control. — Clearer filter model and stronger control from results.

Product Judgment

What this work proves

My job was to help turn buyer needs, product constraints, and platform differences into a baseline teams could test and extend.

Buyer clarity: helped people express what they were looking for without making the filter system feel heavier.
Search systems: clarified hierarchy, taxonomy, selection management, and return-to-results behaviour.
Evidence discipline: reset the direction after weak validation and used the strongest signals to sequence rollout.
Platform judgment: aligned the product model across surfaces while adapting interaction details by platform.
Commercial awareness: connected buyer clarity to list-page progression and enquiry paths without overstating causality.

Strategic Foundation

The reset originated in a strategic vision for the buyer experience

New Search did not start as an isolated UI redesign. It came out of earlier strategic work that clarified what the buyer journey needed to do better across search and decision-making surfaces.

Before this reset moved into execution, I was part of the upstream buyer-side vision work that helped reframe the experience around guidance, selection, and trust. That work brought together research synthesis, market review, concept development, and stakeholder alignment across search, list, and detail surfaces. It gave the team a clearer strategic foundation for what a stronger buyer journey should do, and New Search became one of the clearest downstream expressions of that direction.

For this case study, I am not treating that broader work as a separate product launch or claiming sole ownership of it. I am including it because it materially shaped the problem framing, design principles, and cross-platform baseline decisions that followed.

The Problem

Why search needed a reset

This was not a cosmetic redesign. Friction in search was affecting both the buyer experience and a key commercial funnel, and the existing direction no longer had the evidence to justify incremental improvement.

Search sat close to one of the marketplace's most valuable moments: moving from browsing to enquiry. When filtering was hard to use, buyers had a harder time finding relevant cars and the path to enquiries weakened.
The reset was triggered by evidence, not preference. Earlier validation showed the existing app direction underperforming on core enquiry signals, which made another cleanup pass hard to justify.
Apps had the weakest starting point: earlier and intermediate input patterns were inconsistent, filters were hard to find, and the inherited direction still did not create a strong enough baseline.
Web had its own status quo: a denser filter model with weaker upfront hierarchy, more scrolling, and less clarity on how to scan first before going deeper.
Cross-platform discovery showed the same structural issues repeating across desktop, mobile web, iOS, and Android: filters were hard to find, hard to edit from results, and not clearly organised once applied.
Strategy work later showed that filter engagement on mobile web was shallow. That same work framed New Search as growth work tied to stronger enquiry paths.

The Change

What changed in the baseline

The high-value changes were structural: clearer filter hierarchy, stronger entry points to high-value choices, and interaction patterns the team could actually validate and roll out without hiding the inventory complexity buyers needed to express.

On Apps, the reset started with a clearer filter overview: buyers could scan the system first, then dive into one section when they needed more control.
We grouped filters around reusable patterns, made key choices easier to discover, and cleaned up inconsistent input types so the system felt easier to learn and extend.
One important navigation tradeoff came from make/model. The logical path would have sent buyers deeper into the next make/model step, but evidence showed many wanted more filters or listings faster, so we returned them to the filter overview instead.
After Apps, we moved to Web and applied the same winning principles: clearer hierarchy up front, better filter usability, and a stronger baseline adapted to Web's constraints.
The hard part was exposing richer vehicle data while keeping the experience light enough to use. The system needed to support what buyers were looking for without turning every search into configuration work.
The baseline was adapted rather than cloned: Android moved toward platform-appropriate save behaviour with extra clarity work still needed, while web launched a narrower baseline first and deferred richer enhancements until later.

Important workstream

New Make/Model Taxonomy and Filter Experience

As part of the broader New Search reset, I helped shape a new make-and-model experience for more nuanced vehicle data. The legacy interaction assumed buyers could move cleanly from make to model, but that stopped scaling once generation, variant, and engine choices started to matter.

This became a mental-model problem: how to expose a richer structure without overwhelming people. I explored search, drilldown, grouped dimensions, and clearer selection management so the system could feel more precise without feeling heavier. The work is still evolving, so I treat it as strategically important work in progress rather than a resolved win.

Mobile web make and model selection flow showing how the reset brought a high-value filtering decision earlier in the experience. — On mobile web, make and model moved earlier into a clearer selection flow.

Old problem

The old make-to-model path assumed a simple hierarchy and buried one of the clearest signals of what buyers wanted too late.

Design move

I pushed make/model earlier and explored search, drilldown, grouped dimensions, and clearer selected-state management.

Tradeoff

Navigation consistency mattered, but user behaviour showed that this filter needed a faster path back to the overview.

Why it mattered

A clearer entry point gave buyers a faster path to relevant inventory and gave the team a cleaner model to test.

Desktop results page with a focused filter modal open, showing one decision at a time without forcing the buyer to manage the full filter set at once. — On desktop results, buyers could refine one filter in context instead of managing the full filter set at once.

My Role

What I owned, influenced, and shared

My role was strongest in the app filter direction and the product decisions around the baseline. I separate direct ownership from shared team outcomes because the evidence and rollout were collaborative.

Directly drove: app filter information architecture, the main filter overview, filter-entry hierarchy, make/model placement, and completion behaviour across iOS and Android.
Influenced: cross-platform baseline decisions with product, research, engineering, analytics, and another senior designer as the direction moved into web.
Why I pushed it: make/model carried a strong signal of what buyers were looking for, while the old structure asked people to manage too much filter depth too early.
Framed the reset as both a product and people tradeoff: the team had invested in an intermediate direction, but the evidence showed we needed a cleaner baseline.
Framed risk: helped move the team away from a weak earlier direction and toward a baseline we could test, launch, and extend.
Shared outcomes: research, experimentation, rollout, and commercial results were team-owned, so I treat them as shared evidence rather than personal attribution.

Key Decisions

Five decisions changed the trajectory

This is the clearest representation of the work: when to reset, how to structure the new baseline, where to trust user behaviour, and where not to flatten real platform differences.

Decision 01 in practice

Legacy baseline, intermediate direction, stronger reset

Buyers met too much filter structure before they could clearly scan what was available. The reset brought high-value narrowing decisions forward, made selected filters easier to review, and gave the team a cleaner model to validate.

An older app search baseline with broader filter clutter, heavier visual noise, and weaker structure. — The older app baseline carried heavier noise and weaker structure.

An intermediate app search direction with layered filter cards that still carried structural friction. — An intermediate direction that still carried structural friction.

The stronger app search baseline with a clearer filter overview and simpler hierarchy. — The clearer baseline the team could validate and extend.

Reset the baseline instead of polishing a weak direction

Earlier app validation weakened confidence in the existing direction. We reset the filter model, reduced structural complexity, and handled the human side of moving away from work the team had already invested in.

Build the app template around scan-then-dive

Apps became the hard reset. We made the overview clearer, grouped filters around reusable patterns, improved discoverability, and gave each filter section enough surface area for better inputs and future testing.

Follow the desire path in make/model navigation

The logical path would have sent buyers deeper into make/model, but behaviour showed many wanted more filters or listings faster. We preserved the broader navigation model while making an intentional exception for this high-leverage filter.

Validate the new direction where the evidence was strongest

iOS became the first clean validation point. The redesigned baseline moved commercial progression signals in the right direction while maintaining downstream engagement.

Adapt the baseline by platform and ship web in stages

We adapted the same product model to each surface. Android evidence supported the direction more narrowly, while web shipped later as a simpler baseline with richer enhancements deferred.

Outcomes / Impact

A stronger search baseline across platforms

New Search gave buyers more clarity while searching and gave AutoScout24 a stronger foundation to build from. iOS validated the direction most cleanly; Android and web extended the model with narrower evidence, so the result reads as a staged product-system outcome rather than one uniform launch story.

iOS

Strongest validation

The iOS benchmark selected the new direction and later rollout evidence confirmed broad launch across AutoScout24 markets and white labels.

Android

Supporting signal

Android supported the same direction through narrower evidence. Research and validation favoured platform-appropriate save behaviour, but the available evidence is less complete than the iOS story.

Web

Phased baseline

Web launched later as a simpler baseline. Post-launch analysis indicated positive movement on key progression signals while mobile web and downstream engagement stayed broadly stable.

Broader strategy work around New Search modelled larger upside and tied the programme to stronger enquiry paths. I treat that as business-case context rather than delivered outcome. The delivered story is tighter: the team replaced a fragmented search model with a clearer baseline, validated the strongest surface first, and carried the direction across platforms with evidence calibrated to each surface.