Deal App: a map redesign that lifted interactions by 60%
I redesigned how people browse, compare and act on property listings in Deal App’s mobile map, turning a static, dead-end flow into map-first discovery. I also reworked how sellers place a listing on the map.
- My role
- Product Designer II
- Timeline
- 2 months, from August 2023
- Team
- 6
Evidence
Increase in map interactions.
Fewer misclassified listings, after the seller listing redesign.
Boost in customer satisfaction.
The 60% counts several buyer actions on the map, tracked together. The 87% followed the seller listing redesign: manual map pins and more location details. The 35% is the team’s figure from overall in-app review ratings.
The story behind it
An urgent problem, handed over in my first week.
How did it start?
In my first week as Deal App’s product designer, my manager and friend, Mohammed Medhat, pointed me to an urgent problem in the app’s property map. Clients and agents were struggling with a clunky, outdated interface that made finding a property frustrating.
The data agreed. Click-through rates, engagement metrics and Smartlook session recordings showed users bouncing almost immediately, with little interaction and visible friction. So I rethought the flow from the ground up.
A key feature had become a dead zone.
The previous flow.
Project summary
Why the map failed at its one job, and the four views that replaced it.
Discovery problem
The map was meant to help people discover properties, but it got in their way. Home seekers and agents met stiff navigation, unclear filters and no guidance, and the flow pushed them into a narrow path too early. Session after session showed the same pattern: high bounce rates, few interactions, visible friction.
Current workflow reality
The map was a static image of Saudi cities: no zooming, scrolling or panning. Tapping a city opened its districts, tapping a district jumped to a list of properties, and from there on the map was gone.
Design challenge
This was a foundational UX problem, not a UI tweak, and the pressure was real. I had to build context fast and design something that fit users’ existing mental models. I explored five map concepts, iterated quickly and aligned stakeholders on the strongest one. The harder part was earning buy-in for a navigation logic that broke with the app’s old flow.
How I approached it
With the product and business teams, I separated the different reasons people open the map. I mapped the browsing goals of clients and agents, then designed four map views: property ads by area, client-submitted requests, promoted listings and agent locations. Each view shows less at once, gives a clearer next action and lets users switch between the map and a list for comparing.
Our solution
A map-first experience replaced the static flow. Users zoom, pan, filter, compare listing types and move between map and list without losing their place, whether they’re browsing properties, reviewing requests, finding promoted listings or locating agents across Saudi Arabia.
Tools and practices
From paper sketches to a tested map in two months.
Tools
I started on paper to explore concepts quickly, moved into shared design files with the team and checked each decision against real behavior in session recordings.
- Notion
- Figma
- Pen & paper
- Zoom
- Smartlook
- Slack
Practices
Interviews and surveys gave us input, session recordings showed real behavior, and early concept tests refined usability. Prioritization kept each stage focused on what mattered most.
- Stakeholder interviews
- Surveys
- User behavior recordings
- Usability testing
- Prioritization
Visuals
The core map feature, redesigned from scratch around real user behavior.
Ads map
An interactive map of property ads to browse by price and location, with filters and an instant switch between views.
Requests map
Live property requests across the map, so agents can see demand by area and respond to opportunities nearby.
Agents map
Active real estate agents by location, so users can find professionals near the area they care about.
Ad preview
Tapping a pin opens a quick-view card with the key details, so users browse nearby listings without leaving the map. Pins update as users swipe.
Ads list
The “قائمة” (List) button shows the same ads as a scrollable list of cards, with direct actions to share, save or contact on WhatsApp.
Submit request
When users can’t find what they want, “اطلب عقارك” (Request your property) opens a short request form right on the map, turning a dead end into a request agents can answer.
Learnings
What did I learn?
Jumping into a high-impact flow during my first week pushed me to balance speed with structure.
Design around behavior. I learned to design around observed behavior, not only requested features: how users scan listings, move between map and list views and recover when they cannot find the right property.
Momentum with clarity. The project also taught me how to keep momentum without losing clarity.
What went well?
Cross-team alignment made the map redesign sharper.
Discovery modes. Product, business and design conversations helped separate different discovery modes instead of forcing every user into one map behavior.
Fast iteration. It gave us a clearer structure for property ads, requests, promoted listings and agent discovery.
What could be improved?
Test earlier. I would test the map behavior earlier with both home seekers and agents.
Document the rules. I would also document interaction rules more clearly: how clusters behave, when users switch to list view, how filters persist and how request submission should work when users do not find a matching property.
Working on something like this?
If your main flow loses people before they act, tell me where it stalls and I’ll help you test a better one.


