Background.

Quicken Loans' lead generation forms don't perform as well as the business would like them to with an average completion rate of 28%. An OKR the entire company is working on is increasing our market share and an increase in leads is critical to meet this goal.

During this project, I was the sole Designer on a team of 3 engineers, a product manager, and a quality assurance tester. My task was to design a conversational landing page that has a higher conversion rate than the current lead generation forms. I was in charge of the UI/UX, conversation design, and content design for the entire experience.

Initial Research.

Because this was a product starting from scratch there was no historical performance data I could anchor my research to. I started internally by meeting folks in the performance marketing team to learn more about how they designed their lead forms. My goal was to uncover any insights and best practices they had when it comes to capturing leads.

I uncovered the following initial insights:

• Not every data point collected is critical. The average lead form has 19 questions and only 5 of these questions are necessary to create a lead in our backend systems. 7 of these questions collect data that is nice to have (according to our sales team).

• We ask for any personal information towards the end of the lead form. Prior testing has been done here to confirm that a user is more likely to provide us with this personal info when they’re already invested in the experience.

Initial Design.

We didn’t have a UX budget allocated for this project and only had 8 weeks before our A/B test launch date. Our product was to be tested with Quicken Loans’ highest performing lead form which had been in production for over 2 years. Because of the absence of a budget, I had to get scrappy with how I designed.

I held co-creation sessions with internal team members such as customer service representatives and home loan agents. These individuals spend most of their working day talking to users and have a finger on the pulse on what resonates with users.

Continuing Research.

We lost the initial A/B test by 12% and went back to the drawing board to continue improving our product. I still didn’t have a UX budget but I now had 2 new tools.

First, I had data from our A/B test and could see how users interacted with our product. Analyzing this data I uncovered the following:

• Only 50% of clients who land on the experience interact with it. Once a user responds to the first question there’s an 85% chance they’ll finish the experience and submit a lead.

• Our second biggest drop-off rate (after beginning the experience) was when we asked clients for the zip code of the area they were hoping to buy a house in.

• A user interacts with an average of 5 pages on Quickenloans.com before landing on our experience.

• 45% of users indicated that they were looked to buy a home in 2 - 6+ months.

• Users who began by clicking on a faq on the sidebar were less likely to submit a lead.

The obvious question I had was why are users not beginning the experience? This wasn’t something that quantitative data could tell me so I managed to convince my director to allow us to run a user test. We used a tool called Usertesing.com and got a free pilot license with 1 test. The stakes were incredibly high as I knew that this test not only had to uncover insights that would allow us to better design version 2 of our product but also convince stakeholders that investing in UX was pivotal.

The video above highlights one of my favorite discoveries of all time. Over 50% of users we tested with didn’t realize that the starting button was a button. It blended in and their attention was rather drawn to the sidebar. This made sense as our quantitative data had uncovered that users who clicked on the sidebar rarely completed the experience. Once users realized how to start the experience, they had minimal difficulty completing it.

Design Part 2.

Quantitative data uncovered that users interacted with an average of 5 pages on Quickenloans.com before landing on our experience. Our A/B test was launching from Quickenloans’ Mortgage Calculator page and every person that landed on our experience had to click the same CTA. I changed the design of our buttons in our experience to be consistent with this CTA as every user who landed on our experience had already proven they know how to interact with it.

45% of our clients indicated that they were at least 2 months away from buying a house. When talking to users I realized that people often don’t know the exact zip code they want to buy a house in (or don’t remember it off the top of their head). We changed the copy to ask for City/ State instead. In addition, we made certain questions conditional based on how users answered previous questions. For example, if a user told us they were at least 2 months out from buying a house, we would skip asking them intimidating questions such as the amount they’ve saved for their down payment and what the exact purchase price of their house is.

Results.

Our conversational landing page won the A/B test with a 69.58% lift. Our completion rate (lead submittal) is 47% which is amazing in the lead generation world. This page is currently live in production and can be found by clicking here. My favorite part however is that the success of this project convinced stakeholders such as my director to invest in UX. At the start of this project, I was the sole user experience professional on my team and had no budget to conduct research or test hypotheses. Now, I’m one of 5 UX designers and my team has an allocated UX budget.

Reflection

This project anchored my belief in using quantitative data to identify problem areas, qualitative data to diagnose the problems, and design to ultimately fix those problems. I learned a ton about stakeholder management, UX buy-in, and even got to flex my baking skills by making these dope cookies for our product launch.