After

Before

Ezi — On-Demand Home Services Platform

Ezi is an on-demand cleaning platform that connects users with trusted local cleaners. It enables quick booking through streamlined scheduling, transparent pricing, and secure online payments.

Users were dropping off before completing bookings.

The old Ezi booking form was long and clunky, especially on mobile. Users often dropped off, and since contact info was only collected at the end, they couldn’t follow up.

Make booking home services effortless.

The goal of the redesign was to simplify how customers book cleaning services—fewer steps, clearer info, and a smoother experience. The changes were all about reducing friction and making the app feel more reliable and easy to use.

My Roles

• UX Researcher

• UX/UI Designer

Deliverables

• UI redesign

• Improved booking process

• Custom design system

Team

• Product Manager

• Lead Developer

Only 7.23% of Users Completed the Booking

This chart presents a funnel analysis of user interactions with the Ezi booking form over the last 30 days, compared to the previous quarter. It tracks the conversion rate of users through various stages of the booking process, from starting the form to creating a new booking.

7.23% of users who start the booking form complete it by creating a new booking. I aimed to boost this conversion rate and encourage more users to finalize their bookings.

60%

Reported that they found it tedious to navigate the clunky booking flow.

45%

Felt that the booking process was too long.

50%

Reported that they’d like to have dynamic pricing as users book a service.

Refining the Booking Experience

The focus was on simplifying the booking process and making key information, like service options and pricing, easily accessible.

Users found it easier to compare services and appreciated seeing prices up front—it made booking feel more straightforward. Some were confused by the time slot selection, so I reorganized the layout to highlight available times and added quick tips. Each round of testing helped refine the flow before moving to high-fidelity designs.

Redesigning the Stepper to Be Mobile-friendly

The original stepper wasn’t optimized for mobile, making it hard for users to follow their progress. I redesigned it to be:


• Vertical on mobile for better readability and touch navigation

• Horizontal on desktop to use screen space efficiently

• Clear and numbered so users always know where they are in the process

This responsive layout made the booking flow easier to follow and reduced user confusion—especially on smaller screens.

Capturing User Information

To reduce drop-offs and capture leads, I added a simple contact step at the start of the booking flow.
Users now enter their name, and phone number, and their postal code upfront. This lets us:

• Re-engage users who don’t complete the form

• Send confirmations and follow-ups

• Personalize the experience from the start

It’s quick, unobtrusive, and improved our ability to connect with potential customers.

Combining 3 Steps Into One Streamlined Section

To simplify and shorten the booking flow, I merged three separate steps—chore type, cleaning type, and areas that need cleaning—into a single, cohesive section, with each chore type having its own respective service selections. This helped:


• Reduce the number of screens users had to go through

• Keep related inputs grouped together for faster decision-making

• Make the process feel shorter and more intuitive


The new layout makes it easier for users to select all core service details in one place, speeding up the booking experience.

Dynamic Pricing

As users choose services and areas to be cleaned, the total cost updates in real time on both web and mobile. This dynamic pricing approach:


• Shows the cost of each selection instantly

• Builds transparency and trust with users

• Helps users make informed decisions without surprises at checkout


The estimated total is always visible, making the experience clear, responsive, and user-friendly.

22%

Increase in completed bookings — fewer users dropped off before checkout.