UX Design Immersive — Project 5 Retro

Improve the Onboarding UX of Perchwell Real Estate App in Three Weeks, with a Team of Three. (with Actual Client)

Perchwell is a real estate search app for properties for sale in NYC. It was launched in iOS back in April 2016. Our team of three at General Assembly were given a creative brief to improve the onboarding tutorial process.

What makes Perchwell unique is that it lets users create market report data visualizations with a wide variety of metrics. These beautiful charts can be a great tool for real estate agents and an eye candy for real estate due diligence fanatics.

It also has a proprietary algorithm called “Perch Price”, that takes into consideration the total cost of ownership over the years including property taxes and maintenance.


We first began with research including:
 — Stakeholder interviews including the founder and real estate agents
 — User interviews
 — Contextual & Competitive analysis.

We screened for users who have bought real estate in NYC and conducted user interviews. We then wrote down notable quotes from users, categorized the quotes, and generated a persona based on this method of ‘affinity mapping’.


Out of the user interviews, the following insights were synthesized:

Nobody understood what ‘Perch Price” was.

Difficult to Navigate Iconography and Information Architecture

(which lead to a taxonomy solution later in the project)

Inconsistency in Swipe Direction and Other Gestures

Other UI Issues

To address these insights, the following changes were proposed to the client.

We proposed the client that we wanted to correct some basic issues with icon recognition and taxonomy before we moved ahead with the tutorials.


In a nutshell, Perchwell had two main types of data visualizations — Current and Historical. Underneath the main two category, each had six subcategories, and anywhere from 2–20+ sub-subcategories. Four of the six subcategories overlapped, and there were minimal sub-subcategories under Historicals, so we suggested to combine the Current with Historical, and narrow everything down to six categories. By doing that, the step to choose between the two categories could be eliminated, and steps to add more charts to the report became simpler.


Most users thought the “view” icon on the upper right hand was the virtual reality “360 view” often seen in real estate sites, especially with the arrow that indicated some sort of circular movement. In reality the “View” icon would take the user to the selection page to toggle between 3 types of views: 1. data visualization view
2. map view
3. listing search result view

So the icon was redesigned to an eye.

Another stumbling block for users, was the “side-by-side comparison” feature icon. Most users mistook it for a “send” icon. So we redesigned the comparison icon as an instantly recognizable scale symbol.

A few other icons and microcopy were readjusted for better recognition.

In the meeting to report these research results, the client was ‘surprised’ that users had difficulty understanding some icons and interactions.

I believe it is the duty of UX designers to unearth these unexpected user insights.

We prototyped some solutions in low-fidelity wireframes and put them in front of another set of users for feedback, and iterated in order to create our final high-fidelity prototype.


Here’s the final prototype in InVision.


For next steps, we suggested the client to continue user testing and iterating on the prototypes we provided. Some ideas we proposed were already something they were addressing, while other ideas will go into future versions.

Overall, it was a successful project with a very clear scope and goal from the beginning to the end. We did have to adjust course based on heuristics findings, but we were able to propose viable solutions that made our client happy in the end.