Usability improvement project: mobile filters
Client:, Full-time Employee

Group 11917_2x.png

professional summary

Product Designer at

Since June 2020, I have been a fully remote product designer for in Austin, TX. Our platform is divided into two parts- The Provider side (caregivers) and the Seeker side (families). I design for the Seeker side across all platforms including desktop, mobile web, and app (iOS and Android). 

Screen Shot 2021-11-09 at 8.18.23 PM.png

Say hi to the camera!

Screen Shot 2021-11-09 at 7.48.36 PM.png
Screen Shot 2021-11-09 at 7.45.56 PM.png

Some of my components and illustrations

My Responsibilities
I work closely with my developers and project managers on a daily basis to ensure my data-driven designs come to fruition. I am responsible for:

  1. Performing competitive analysis to support my designs

  2. Organizing user tests with my researcher to validate my designs + reviewing results to inform my iterations

  3. Creating components as needed for our pattern library (icons, illustrations, and unique library components)

  4. Finalizing copy with our content design team

  5. Performing routine QA for my projects and others as creator and leader of "Design Office Hours" where I work 1:1 with any developers who wish to showcase their design progress

  6. Presenting my final designs to stakeholders


1. project summary


Mary Beth Garrido (UX/UI designer)

Emily Massa (Project manager) 

Supritha Shreyas (UX researcher)



Playbook UX


October 2021, Two weeks

Project Overview

As part of Operation Superstar (our initiative to increase our app store rating), I have been tasked with exploring usability solutions to negative reviews about our search filters.

Primary Use Case: Childcare Seekers searching for a nanny

Project Goal: Diagnose the user problems and come up with a simple and intuitive filter solution to enhance filter usability. We want everyone who is looking for a nanny to know how to use the  filters without any problems so we don't receive any negative app reviews about search filters.

Hypothesis: Nanny Seekers would benefit from a more intuitive filter experience and increased clarity of Provider availability in search results. A better usability solution to these problems will increase our app star rating.

Screen Shot 2021-11-10 at 5.15.46 PM.png

Pain Point: Filters aren’t intuitive or easy to discover 😱

We have features that sort sitters available for a specific time. Negative reviews suggest users don't know these features exist.

  • “Doesn’t let you search by your budget or schedule so it’s basically useless”      -Reviewer

  • “App doesn’t allow you to sort sitters available for specific times” -Reviewer


2. exploration

Current User Journey

I created a journey map of the current user experience in App to get a sense of common user frustrations and possible pain points throughout the flow.

Screen Shot 2021-11-09 at 3.53.30 PM.png

Current search user journey

Competitive Research

Next, I assembled an audit of common UX search patterns across competitor apps such as Sittercity and UrbanSitter.

Screen Shot 2021-11-08 at 7.23.05 PM.png
Screen Shot 2021-11-08 at 7.23.29 PM.png

Competitive analysis (Direct and indirect)


I discovered there are two popular approaches when it comes to filters- having them either upfront or secondary to your search results. For testing purposes, I went with a hybrid approach- select your vertical upfront, and then filter after seeing your results.

Screen Shot 2021-11-09 at 4.11.01 PM.png

My initial proposed solution for testing

3. user testing

Research planning

I next organized a meeting with our UX Researcher, Supritha Shreyas, to go over research goals and create a usability test game plan. Research Goal: Evaluate if the filter's interface, interaction, and journey are intuitive to users.


The result is a comparison test between the control and my proposed solution. We tested five users, all of whom have sought to hire a nanny in the past year. We also wanted to know if we should make the "Start date" input mandatory or not since similar apps (UrbanSitter) do not ask for 'Start date." We also wanted to get some user feedback to help inform our decision to keep the “show partial schedule matches” toggle on or off by default.

Screen Shot 2021-11-08 at 7.53.42 PM.png

• Together, we compiled an interview scenario script for our five users

• We focused on the usability of the current flow vs. my new proposal (below)

• We also asked about some feature specific questions to improve not just flow but content as well

Our user testing script

Control vs. Test

We performed an unmoderated comparison test between the control and my proposed solution. Five testers were tasked with walking through a scenario on two prototypes I made.

Screen Shot 2021-11-08 at 7.49.58 PM.png
Screen Shot 2021-11-09 at 3.53.30 PM.png
Screen Shot 2021-11-08 at 7.50.27 PM.png

Production (Control) vs. my proposed design (Test)

Research Conclusions

My proposal was overwhelmingly preferred by our users.  🎉

Below are some notable pain points we discovered about the control (what's in production):

"That other flow, the first one (the control), it’s choppy like disconnected... it was like they weren’t together.” -Tester

Screen Shot 2021-11-09 at 4.26.51 PM.png

Pain Point 1:
The ‘Searchbar’ filter is confusing

4/5 users experienced confusion locating the date/time input

Screen Shot 2021-11-09 at 4.27.01 PM.png

Pain Point 2:

Start date should be optional

Start date is REQUIRED if you want to filter by availability today

  • 4/5 users say “Start date should be a mandatory input”

  • However, one user said “I seldom have a start date in mind”

Screen Shot 2021-11-09 at 4.27.10 PM.png

Pain Point 3:

Partial schedule matches should be ‘On’ by default

4/5 users said they would like to see partial schedule matches in search results


4. solutions

Screen Shot 2021-11-09 at 7.22.44 PM.png
Screen Shot 2021-11-09 at 7.23.21 PM.png

Quick Wins 💪

For a short-term solution, I improved filter UI and adjusted default inputs based on user needs. This should immediately improve the user experience via:

  • Improved iconography

  • Start date is now optional

  • Partial matches are now turned on by default

Long-term Solution ✨

“This one is better. The search is just better. It seemed more intuitive... it’s far superior to the previous one." -Tester

While technically functional, the current flow isn't optimal from a usability perspective as shown by the quote above. Below is my 'dream' proposal to bring our search experience to where it needs to be. I am currently working on refining it as we hope to bring it into development this year.

Screen Shot 2021-11-08 at 7.50.27 PM.png

Test flow overview; the "Winner"


Screen Shot 2021-11-10 at 9.11.09 AM.png


Screen Shot 2021-11-10 at 9.06.33 AM.png

How many Seekers apply the 'Days and times' filter on App?
60% of Seekers who used filters on App apply the 'Date and time' filter

Why was the “10 mi” filter moved?
Only 9% of Seekers who used filters used the distance filter on App


Screen Shot 2021-11-10 at 9.18.37 AM.png


Screen Shot 2021-11-10 at 9.19.23 AM.png

“It was really easy to go through the filters. The filtering on the second option felt a lot more thorough and gave me more control over the exact criteria I was looking for.”



Screen Shot 2021-11-10 at 9.23.38 AM.png


Screen Shot 2021-11-10 at 9.23.54 AM.png

‘Reset’ clears all filters except default filters. These filters include:

  • ZIP code

  • Mi away (10 mi)

  • Who needs care (all selected)

  • Hourly rate

  • Partial matches toggle reset to ‘On’

‘Reset’ is a filter term used in apps such as Facebook and Etsy. It is more accurate than ‘Clear all,’ which is not the desired action.


Screen Shot 2021-11-10 at 9.28.35 AM.png


Screen Shot 2021-11-10 at 9.31.16 AM.png

Filters (#) indicator: This number increases for every filter changed, including default filters. This is a pattern found in UrbanSitter, Facebook Marketplace, and

“This one was so much easier to understand. I like the flow of it, it was easier to use and you could undertsand it better. It was simple, I would use it. -Tester