This project delves into designing an experience to better understand our users through blue sky thinking, experimentation, improving the recommendation engine and curating results for the users.
The "Discovery Process" focuses on creating a survey experience that allows users to input their needs in order to give them the best plan recommendations and increase their confidence in their decision.
The current funnel goes through from the Landing Page to the Census and straight to Quote suggestions; the Discovery Process aims to disrupt this flow by giving the users the option to add more context to their purchase before seeing the Quotes page.
UX Researcher, Solo Product Designer, Co-leading Product Thinking Brainstorming
VP of Product, 3 Engineers + QA team, Project Manager, Data Scientist, 2 designers (the other designer collaborated in the capacity to provide feedback and brainstorm solutions).
4 months for Desktop and Mobile flow
Sketch, Figma, VWO, Heap Analytics, and Hotjar
"As a user I want to see healthcare quote options that best match my
situation and needs after sharing all my personal information without scrolling through hundreds of plans."
The business needs aim to understand users and their needs better, so the recommendation engine can improve its algorithm and recommend the best options for the users, without making their experience on the site cumbersome.
To find low-lift solutions that we could test incrementally, we needed to understand the data better. Here's what we learnt:
- Users with dependents are higher intent users and want plans with better additional benefits.
- Users who compare plans are more likely to convert.
- Users are more likely to only interact with the first 10 plans, then lose interest/drop off.
Framing a hypothesis that would help us craft meaningful solutions:
- If we ask the user a series of questions, we can show them plans with the additional benefits they want.
- If we make comparing plans more intuitive, users are more likely to trust our recommendations.
- If we allow the users to progressively view plans, there will be a lower drop-off rate.
The project manager and I started crafting questions we would need answered in order to better recommend plans. In order to do this efficiently, we needed to develop a deeper understanding of how our plans and their benefits align with different user segments.
We shared all of our learnings with the bigger design/product/data science team — we wanted to be intentional with our solutions and navigating improving our technology with purpose.
In my growth-squad, we often tried to take the path of lease resistance when trying to establish learnings in a problem space. However, for this initiative, we decided to think outside the box — “What if we could reinvent the wheel?” — in this case, reinvent how people browse through health insurance plans.
I proposed a few open ended questions to the larger team & then a few more followed from the team such as...
If insurance plans were a frequent purchase, how would we go about recommendations?
If people knew everything about health insurance, what does the experience look like?
How would healthcare be purchased if it was a product in a grocery store?
What questions are we most afraid to ask our users?
What rules would the experience of buying healthcare have, if it were a game?
What does healthcare look like if the world is only O65 users?
My team left the room with a plethora of ideas and then gathered our favorites and highly voted suggestions and ideas.
We started trying to apply each idea to our given problem to see if anything would stick while also mapping out the real life limitations/scope.
We started trying to apply each idea to our given problem to see if anything would stick while also mapping out the real life limitations/scope. After reviewing, digesting, processing, and assessing in our team, I mapped out the big ideal vision.
The bigger vision for this feature would allow users to enter their ZIP code, opt in for a healthcare quiz in order to get curated healthcare plan recommendations, they could then explore each plan and calculate the cost of different healthcare scenarios (fractured ankle, fever, etc,.) while also being able to adjust limits and premiums of their the final plan.
In order to approach our problem incrementally, the VP of Product and I decided to iterate on two areas that would build a strong foundation for the bigger vision. We ran a series of tests over 5 weeks. A few key learnings were —
- Creating a new experience before the Quotes page in the funnel called "Discovery Process," with ten easy questions that explained the users needs. By the fourth iteration, the flow was optional with four questions that resulted in a 13% increase in conversion.
- Amplifying the priority of the "compare" button that led to 44% more users comparing plans and 15% lift in conversion.
The learnings from the tests we ran over five weeks helped us validate the ROI for the lift in the improvement within the overall user experience/funnel. We measured success by percentage of users who chose to go through the Discovery Process, conversions from comparison and conversions into the "Start Application" (plan details) page.
The key takeaways from the smaller tests validated our hypothesis that users do want "help", comparing plans encourages users to feel more confident in their purchase. Lastly, users who do the discovery process are more likely to buy a plan recommended to them.
The final solution asked the user if they wanted "help"; users who said "no" were directed straight to the Quotes page whereas users who said "yes" were asked 3 questions that assessed their specific needs. Once the user answered all 3 questions, they were shown a customized Quotes page that had 3 primary recommendations reflecting their needs and a comparison between the three plans.
Adding the optional Discovery Process step in between the Census page and Quotes page boosted overall conversion.
Users who opted in and completed the Discovery Process were more likely to convert till the end of the funnel. significant increase of conversions to the Quotes page.
We improved percentage of calls generated to our in-house sales center of healthcare agents within the first month that led to 400 more sales.
Comparison between recommended plans saw a significant increase which boosted the "compare" feature that is known to have a higher conversion rate.