In 2019, we faced a tough situation. Our company was pouring a ton of money into marketing, but we just weren't seeing the results we expected. Our lead generation efforts were falling short, with a high bounce rate and poor conversion rate compared to our visitor numbers.

As a Product Designer, I collaborated with the Product Manager and Tech Lead to develop a feature that collects visitor data to suggest real estate listings based on their profile preferences. I was responsible for:
• User research
• User testing
• Wireframes and prototypes
• Release strategies
• Interface

The ads displayed only a limited range of listings, which meant visitors were seeing only a fraction of the available options. We needed to optimize our campaigns to increase the number of real estate listings viewed per user, ultimately boosting the metric for scheduled visits.

We aimed to understand the primary user journeys, the actual volume of traffic reaching the website, and how users behaved once they were there.

• Users primarily navigated listings through the campaign carousel rather than on the website itself.

• Most users accessed only the listing page and rarely ventured to the search function.

• The average time per user session was lower than expected when compared to typical buyer behavior.

We wanted to remember the clients who click on campaigns and drop off, that we exist. To do that, the plan was:

1. Get new users to became recurrent users

2. Increase average listings viewed per user

3. Get their contact

4. Make them want to schedule a visit

We made a strategic decision to invest in a recommendation robot. Initially, we implemented a form that collects user preferences and sends 5 recommendations via WhatsApp. To promote this feature, we placed banners on both the Listing Detail page and Homepage. Additionally, brokers began filling out this form for clients during their initial visit.

The EmCasa profile was already well established. We had almost 1% of the visitors completing the form, and interacting with the recommendations. Looking through data, I discover that there was an association between people with profiles and sales. The search page was outdated. To connect the experience between the profile and the listing page, I decided to concentrate the user activities on the search page. The outcomes of this page were:

• Search using filters

• Manage your favorite listings

• Change your preferences 

• See the recommendations