Quan Yuan

Copyright © ALL RIGHTS RESERVED. QUAN YUAN

Native Agent Finder

iOS & Android · 2025

Native Agent Finder
from zero to shipped

The first fully native agent discovery experience on Zillow — landing page, search, results, ingress entry points, and a Sponsored agent system.

Saved Homes — Find an agent ingress
Home Loans Tab — Agent Finder entry
Native Agent Finder landing page
Agent Finder search results page
4 surfaces designed end-to-end
Saved Homes ingress Home Loans Tab ingress AF Landing page Search results page
53M
Projected monthly
impressions
6×
More likely to
submit contact info
66%
Of all profile submits
from Agent Finder
0→1
Native AF — first
ever on iOS & Android

Problem Statement

Half of Zillow's users had no native way to find an agent.

Agent Finder was Zillow's highest-converting agent surface — users who arrived via Agent Finder were 6× more likely to submit contact information than average. Yet for 50%+ of Zillow's users on mobile, this experience was delivered through a slow embedded webview that missed native platform conventions, had no prominent entry points in the iOS or Android app, and couldn't support features like GPS-based location inference, native filters, or a Sponsored agent system.

"Currently, the agent directory is only available on mobile through webviews, making it inaccessible to nearly half of Zillow's traffic and resulting in a suboptimal customer experience."

Issues for Mobile Users

Three reasons the webview experience failed

1
"I had no idea Zillow had a place to search for agents. I just Googled it."
No discoverability on mobile
There was no entry point to Agent Finder on the Saved Homes tab, Home Loans tab, or anywhere else in the native app. Agent-first buyers had no path forward.
2
"It feels like I'm using a website inside the app. The back button doesn't even work right."
Webview felt broken and slow
The embedded webview missed native navigation, gesture recognition, location services, and the performance expectations of an iOS or Android app. Every interaction added friction that reinforced distrust.
3
"There's no way to filter by what I actually care about — like price range or number of sales."
No native filters, search, or premium capabilities
The mobile experience lacked native search autocomplete, location inference, and filters. Advanced capabilities like Sponsored agent placement or Showcase integration couldn't be built on a webview foundation.

The Opportunity

Agent Finder already over-indexed on conversions. We just had to make it reachable.

Before designing a single screen, the data made the business case undeniable. The gap was access — not product quality. Making this path native and discoverable for 50%+ of users on mobile wasn't a speculative bet. It was unlocking existing demand.

66%
of all profile submits from AF
Despite driving only 20% of profile views. Agent Finder users arrived with clear intent.
6×
More likely to submit
Agent Finder users vs. average user. The highest-intent path in the agent discovery funnel.
53M
Projected monthly impressions
From new native entry points on Saved Homes + Home Loans Tab across iOS and Android.
50%+
of ZG traffic on mobile
With no native Agent Finder experience to capture it. The gap between traffic share and product investment was the core opportunity.

User Research · Agent Directory Ingress IA Report

One finding shaped the entire ingress strategy: don't push users before they're ready.

The IA working group synthesized 20+ research papers, decks, and reports across Research, Touring, ZIBs, and ZPA — examining competing agent-finding experiences, how buyers actually choose agents, and what they value most. The decision to place Agent Finder entry points on Saved Homes and Home Loans Tab — and nowhere else — came directly from this research.

"Users do not want to be pushed to an agent before they feel ready. Put Agent Finder where users are already comparison-shopping or planning financing, keep the messaging lightweight, and avoid interrupting people before they are ready for agent help."

— Agent Directory Ingress IA Report synthesis
Saved Homes is the right moment. Buyers spending time there are actively comparing options — ready for agent help after some browsing. Recommendation: a scrolling upsell shown once, 1.5–3 scrolls down — visible but not disruptive.
Home Loans Tab is an early-journey signal. Users setting budget or getting pre-qualified are at a natural point to think about finding an agent. But keep the message simple — financing already carries high cognitive load.
Users prefer trusted recommendations over cold search. They start independently before involving an agent — and are more willing to share information when the benefit is clear. This directly shaped the CTA copy on both ingress placements.
Suppress for users with an existing agent relationship. Showing the AF CTA to a MyAgent user is redundant and erodes trust. Both entry points suppress for these users — a direct research recommendation.
Saved Homes — ingress placement
Saved Homes — Find an agent ingress
6th slot for 5+ saved homes, last slot for 2–5. Shown once. Research-validated as 1.5–3 scrolls — visible but not disruptive to browsing.
Copy tied to context. "An agent can help you tour the homes you've saved" — connects agent value directly to what the user is doing on this surface.
Suppressed for MyAgent users. If a user already has an agent relationship, this module doesn't appear — respecting the existing relationship.
Projected reach: 40.2M monthly impressions across iOS (26.5M) and Android (13.7M).

Design Solutions

Five principles that shaped every design decision

1
Make it truly mobile-native, not a webview port. Native navigation, Constellation components, GPS-based location inference, platform-appropriate gestures. The product hypothesis: nativizing improves engagement, submits, and satisfaction — not just aesthetics.
2
Drive discovery through high-intent ingress, not broad placements. Entry points only where intent was already high — Saved Homes and Home Loans Tab. Explicitly avoided broader top-of-funnel placements that would disrupt monetized lead flows.
3
One unified search bar across landing page and SRP. Early designs had two distinct search patterns. Unified to one Constellation-aligned bar — reducing visual complexity and cutting engineering duplication.
4
Sponsored agents — labeled, never hidden. ASA agents on Zillow Pro can purchase elevated placement. Every Sponsored card is clearly labeled. Impression-based rotation ensures fairness. Transparency principle held through Legal and ASA review.
5
Design for measurable iteration, not just parity. Success tracked through AF Searches, Profile Card Clicks, Profile Submits, and PLC (Propertyless Connection) Clicks/Submits — with A/B testing and entry-source instrumentation built in from day one.

Design Iterations

Key decisions that improved the native experience

Early design
Hero image at top of landing page

Early designs opened with a hero image above the search bar. The visual was appealing but pushed the search bar and featured agents below the fold on smaller screens.
After feedback
Search bar + featured agents first

Removed the hero image entirely — utility over marketing. Featured agents now populate immediately from inferred location. Title refined to be more eye-catching.
Early design
Two separate search bar patterns

Landing page and SRP each had distinct search bar implementations — different visual patterns, different interaction models. Made the experience feel inconsistent.
After alignment
One unified Constellation search bar

Aligned both surfaces to one reusable native search pattern, confirmed with the Constellation iOS team. Reduced visual complexity and cut engineering effort.
Early design
Missing agent photos → broken image

When an agent photo wasn't available, the card showed a broken placeholder — common in production data and especially visible at launch when profile completeness was variable.
After feedback
Initials-based avatar fallback

Graceful initials-based avatar when photos are missing. Company logo intentionally hidden on load failure. Every card looks polished regardless of data completeness.

Delivery Timeline

Q1 '25
All P0 native AF designs completed and handed off. Landing page, search, SRP, agent cards, filters, error/empty states.
Q2 '25
Sponsored agent treatment designed and delivered. Impression-based rotation, transparent labeling, premium card differentiation. Supports Zillow Pro ASA H2 launch.
Q2 '25
Showcase agent highlighting designed. Seller filter and Showcase attribution on agent cards. Supports $30M Showcase marketing campaign.
Q3 '25
Saved Homes ingress designs delivered to engineering. 40.2M projected monthly impressions. HLT ingress designs in parallel.
Q3 '25
Native filter updated with Showcase filter. Ongoing dev support through implementation sprints on iOS and Android.

Interactive Prototype

Try the native Agent Finder — click through the real flow

This is the working Figma prototype for the native Agent Finder experience. Click through the screens to explore the landing page, search, agent results, and profile flow exactly as designed for iOS.

Native Agent Finder — Interactive Prototype
iOS · Figma

Mockup Sample

Every flow, across both platforms and both modes

The full set of native Agent Finder flows — landing page, search by location, search by name, results, empty/error states, and adding a second search parameter mid-flow — designed natively for iOS and Android, in both light and dark mode. Toggle below to compare.

Native Agent Finder screen library — iOS, Light mode
Click to zoom
iOS · Light mode Scroll horizontally to see every flow

Outcomes

Impact across the native Agent Finder project

53M
Monthly impressions projected
From Saved Homes + Home Loans Tab native entry points across iOS and Android — a new discovery flywheel for agents.
0→1
Native AF delivered
Full feature parity with web on iOS and Android. No native Agent Finder experience existed before this project.
$30M
Showcase campaign supported
Showcase agent highlighting in AF connected user experience directly to Zillow's 2025 Showcase marketing investment.

Recognition

"Big thanks for improving and streamlining the design for the native Agent Finder! Your attention to detail, focus on consistency, and strong understanding of native design principles have transformed something complex into a simple, refined experience."

— Johnny Taboada (Constellation design partner)

"It was such a pleasure working with you on the Native Agent Finder. From shaping a clear product vision to organizing the journey from start to finish, your leadership has been instrumental."

— Colin Bertrand (Senior PM)

"Led end-to-end design for native Agent Finder, enabling Zillow to transition from webview to fully native on iOS and Android. Impact: projected 53 million monthly impressions, creating a scalable discovery flywheel for agents."

— Sahithya Baskaran (Product design manager)

"Quan demonstrated 'Own It' by proactively leading the AF redesign with clear communication and timely deliverables, and 'Better Together' by leading frequent design reviews and ensuring stakeholder transparency."

— Erik Lovelly (Senior principal PM)

Reflection

Designing native-first means thinking in platform affordances, not just screen states. The biggest wins — pinned search bar, GPS-based location inference, native modal for PLC — came from asking what the platform made uniquely possible rather than how to port the web experience.

The ingress strategy taught me that placement is a design decision, not just a product one. Where you put an entry point determines who sees it, what mental state they're in, and whether they're ready for the action you're offering. The IA research recommendation to avoid interruptive top-of-funnel placements wasn't conservative — it was a user trust decision that made the experience more effective.

"Led end-to-end design for native Agent Finder experience, enabling Zillow to transition from webview to fully native functionality on iOS and Android — aligning with Zillow's enterprise-wide push for platform-native experiences, improving app performance, discoverability, and long-term scalability."

— Sahithya Baskaran (Senior Product Design Manager)