Loft App

The numbers showed us that the mobile web experience was not optimized for the user context; users spent less time navigating the site and left sooner compared to desktop.

However, the majority of the traffic still came from the mobile site.

Previous research showed that users felt frustrated when using the mobile site. They said they preferred the desktop site and only accessed the mobile site when necessary and when they were not at home.

• Convenience

  Users complained about needing to select the city and fill in filters each time they accessed the website on mobile.

• Personalization

  They wanted to receive notifications about new properties, changes in property prices, and other updates.

• Performance

  Users said they usually leave the mobile site if the loading time is below expectations, particularly when loading photos.

• Market Share

  Create an additional sales/relationship channel with customers.

• Business Impact

  • •Be the primary channel to contribute to conversion by bringing in qualified leads who are more confident about purchasing a property.
  • •Increase average session time and DAU/MAU.

• Personalized Experience

  • •Build a database based on usage behavior.
  • •Personalize the experience by displaying property options that are more accurate based on browsing and search history.

I was the founding senior product designer in a multidisciplinary team, working closely with the product manager and engineering manager to define and execute our plan and strategy.

I was also responsible for mentoring and supporting other designer work, ensuring the quality and visual consistency of the interface and deliverables.

To achieve our goal of gaining market share quickly, we developed a Minimum Viable Product strategy that allowed us to be available in the app stores within the first month after the team was set up.

The main premise of the MVP was that it didn't need to be perfect; it just needed to work well enough to ensure our presence in the app stores as a first step.

During the MVP development, we gathered together to plan the next releases, focusing on the product vision and users' feedback.

I led the design efforts for each release, ensuring we delivered value to the users and followed our experience principles.

As the first step in our strategy to develop a personalized experience, I designed an onboarding flow that allowed new users to filter and visualize the most relevant properties based on their preferences.

At the time, the main app navigation was through the 'hamburger' menu. As a strategy to increase overall average session time and the number of pageviews, I analyzed the data and designed a bottom navigation that provided users with quick access to some of the most accessed sections in the app.

Users shared in interviews that they felt overwhelmed when navigating through many properties. This presented an opportunity to increase average session time and lead-to-schedule rate by displaying an updated and personalized list of property recommendations.

To continue evolving our product vision of creating a personalized experience, I designed a feedback component to evaluate if our recommendations matched users' needs. This data improved our recommendation engine and helped measure user satisfaction.

In some cases, the filters applied during the onboarding step did not return any results in the property listings.

People missed having a map to navigate and search for properties in their preferred region without needing to select a specific neighborhood.

People were able to complete the onboarding process easily.

After reviewing the insights that came from user interviews, we decided to run an A/B Test to evaluate the finding which, in the first time experience, people prefer rather to jump right to the property listing instead of filtering their options.

We found out that when people go straight to the property listing, we had a 16% increase in the Lead To Schedule (L2S) conversion rate. This discovery resulted in changing the entire onboarding flow, removing now irrelevant steps for the users.