Marble logo featuring a retro marble logo with midnight blue, orange, cream, and mustard yellow colors with midnight blue text that reads Marble.

A single hub for all your insurance needs

Within a team of four, we redesigned our client Marble’s shopping feature on their website, helping policyholders compare and choose the right policies for them and their families. I rearranged key features, improved the UX copy, and ultimately, created a mental model that makes more sense. 

A laptop with Marble’s landing page in the frame of the screen.


Marble’s unique value proposition is the insurance wallet, a digital wallet where policyholders can upload and manage all of their insurance. However, Marble’s long-term goal of being a leader in the insurtech space means that will need to add tools for a complete insurance experience.

Problem space

Enter Marble’s new shopping feature, the main focus of this sprint. Marble’s shopping feature empowers and affords their members agency during the insurance shopping process. My team and I tested and redesigned Marble’s shopping feature to make sure shoppers were able to receive quotes, but also, make meaning out of them.

My role: Generalist UX designer

In a decentralized team of four (Maryna Filipava, Stephanie Lee, Lester Li, & Peggy Lu), we collaboratively kept our team on track and ahead of deadlines and redesigned the Marble shopping experience based on our research.


Analyzed existing research conducted by teams at Marble, competitive feature analysis, competitive task flows analysis, design workshop with stakeholders, created an email survey, user interviews, affinity map, user personas, user journeys, MUSHCOWO map, feature prioritization, sketching & design studio, prototyping, and moderated & unmoderated usability testing.


Figma, Figjam, G Suite, Maze, Keynote, Trello, Zoom, & Slack


3-week sprint (May 2022)

Our challenge

Shopping for insurance

During the sprint, the tool Marble was focused on is shopping. With so much competition and other third-party companies that provide an avenue to shop for insurance, we needed to understand what the best practices were, and what others are offered that we might have missed. Additionally, as the current design stood, we needed to explore the concept of shopping, shopping carts, and quote comparisons in order to understand the mental models associated with shopping for insurance.

Because of our client's goals, we delivered solutions in two phases.

Our goals

Business goals

The shopping feature's implementation roadmap was dictated by a pressing deadline. That is, Marble would be raising funding within a few months from when we started our work. Thus, we aimed to provide multiple rounds of usability testing and iterations to their existing designs before they developed the feature.

Research & project goals

And with our insights from research, we then set out to redesign Marble's shopping feature based on what we heard from stakeholder interviews, user research, and competitive analyses.

Our user research sought to find out: (1) how Marble members would accept Marble's new shopping feature; (2) how they shop for insurance; (3) why they bought the insurances they currently have; and (4) uncover any areas of frustrations that they have may have experienced.


Phase 1

Test, iterate, and improve the existing designs provided to us by Marble

Our solution for phase 1

A screenshot of a proposed alternative for shopping for quotes.

Measuring success for phase 1

A screenshot of a proposed alternative for creating reminders.


Phase 2

Explore "shopping" based on our insights from usability testing, and business & user research

Our solution for phase 2

A screenshot of a proposed alternative for Marble's Marketplace page.

Measuring success for phase 2

A heatmap of a proposed alternative for Marble's Marketplace page showing where interviewee's clicked.

We tested our redesign as well. Because our concept of shopping changed, we were no longer able to use the same questions as we did for usability testing. As a result, we don’t have many statistical results. 

However, through Maze, we generated heat maps of where people were clicking. Each heat map shows that most testers clicked on where they were supposed to!

Here's what our client had to say about us

"Lester was quick to get to work and over the course of a four-week sprint, he and his team quickly generated a high-impact and deeply-researched set of recommendations. We've implemented these design improvements, both across UI and UX copy, and have seen an immediate improvement to our product and feature conversion."

Stuart Winchester, CEO & Founder at Marble 


End of TLDR

The 3-week sprint focused on making Marble members' experience better. We carefully considered the users' and business needs and goals & designed a solution to a problem that was uncovered during our research.


My team and I were able to provide shoppers with more accessibility, familiarity, and relatability for a more enjoyable user experience.

 Here's how we did it.

The process

First, designing the right thing Second, designing things right Iterate & repeat

Our team approached this project as a two-phased project. Because Marble's goal to launch their shopping feature was a crucial component of the business, we aimed to deliver two final designs, one they can implement easily, and another as a redesign of the shopping feature.

Phase 1

Hand drawn magnifying glass to represent "Discover".


Hand drawn paper with writing to represent "Define".


Hand drawn Lego like building blocks to represent "Develop".


Hand drawn funnel to represent "Distill".


Phase 2

Hand drawn magnifying glass to represent "Discover".


Hand drawn paper with writing to represent "Define".


Hand drawn Lego like building blocks to represent "Develop".


Hand drawn delivery person to represent "Deliver".


First, we tested their existing designs, iterated on them, and provided actionable insights and design solutions that were tested and measured to be more successful. The goal of this was to provide data-driven results and designs for their existing shopping feature. Along the way, we learned a ton of things that helped us come up with a redesign.

The second phase of the project was to provide fresh eyes and a UX perspective for their shopping feature. We did this by conducting user research, business research, and competitive analyses. And in tandem with our learnings from usability testing, we redesigned Marble's shopping feature. Here's how we did what we did...


Process for phase 1

1 existing design.

2 rounds of iterations.

3 rounds of usability testing.


Process for phase 2

Research → insights. 

Shopping carts →  quote comparison.

Usability tests

One of the first things we did as a group was review 18 interview recordings and notes that the Marble team had previously conducted. Along with a target audience provided to us by our client, we came up with a problem statement that guided us through usability testing:

How might we engage price-conscious insurance shoppers to make informed decisions on the best policy for their needs by comparing rates, coverage, and customer satisfaction? 

We had 3 rounds with a total of 35 participants of remotely moderated & unmoderated tests where we evaluated 3 features. Check out our process and results below.

Round 1: existing design

We tested the existing screens for a baseline for us to compare our changes. We wanted to see how the site was performing now and to see if and how we can improve it.

Shopping carts

A screenshot of the existing screen to create a new shopping cart.

Participants found shopping confusing and rated this task a 2.4 out of 5 on an easiness scale


A screenshot of the existing method for creating a new reminder.

Managing reminders on the shop page scored a 30% average success rate

Expiring CTA

A screenshot of the existing method for shopping for expiring insurance policies.

Shopping via the orange CTA scored a 90% average success rate

Click into the sections below for an in-depth explanation of our process.

Existing "home" screen

A screenshot of the existing Marble home screen.

The insurance wallet, Marble's unique value proposition, lives on the home page. Here, members can upload their existing policies into the wallet. Insurance cards have two states, (1) active policies and (2) policies that are close to expiring. 

Key takeaways

Existing "shop" screen

A screenshot of the existing Marble shopping screen.

Marble members come to shop in order to create a shopping cart, create and manage reminders, and view a curated marketplace of insurance companies to get quotes from.

Key takeaways

Round 2: 1st iteration

Upon review, we iterated on the existing designs. We created targeted design solutions that address issues that arose from usability tests.

Shopping carts

Still, 0 participants created a shopping cart before shopping for insurance


Creating a reminder went from 30% to 100% average success rate

Add to cart

1 out of 10 participants clicked on the add to cart button

Click into the sections below for an in-depth explanation of our process.

1st iteration "home" screen

On the home page, we added a "+ Set a new reminder" CTA since many participants were looking for a way to manage reminders on the home page.

Key takeaways

1st iteration "shop" screen

As a way to make sense of the shopping cart model, we added the function to add companies into your shopping cart. Additionally, we changed the  "+ New shopping cart" & "+ Add" new reminder to a CTA that matches with the other CTAs on the website.

We also recategorized the page by policy types. We created a page that modeled other eCommerce stores to test if shoppers would recognize and resonate more with more familiar diction.

Key takeaways

1st iteration "shop > shop home" screen

Lastly, we created an additional page with all the companies that offered a type of policy (auto, home, renter, pet, & life). We did this to model other eCommerce shops. We also added a filter and sort by function to allow shoppers to quickly find what they are looking for.

Key takeaways

Round 3: Final design for phase 1

And that brings us to the first finish line. We found data-driven results for changes that are immediately implementable upon delivery. In other words, we took actionable insights and created tangible designs and results.

Shopping carts

5 out of 10 participants created a shopping cart after changing the UX copy


Participants click on the shopping banner at a higher rate than the shop tab in the navbar.

Add to cart

Changing the shopping cart to a bookmark made no difference

Click into the sections below for an in-depth explanation of our process.

Final design for phase 1 "home" screen

We changed "+ Add new policy" to "+ Add existing policy". Additionally, we added a shopping banner, similar to other eCommerce shops, to make shopping more visible.

Key takeaways

Final design for phase 1 "shop" screen

The results from the previous usability tests led us to believe that the shopping cart mental model was not resonating with insurance shoppers. Therefore, we changed shopping carts to quote compares, and carts to bookmarks.

Key takeaways

Final design for phase 1 "shop > shop home" screen

We modified this screen to match the shop all page. This screen was not tested.


Process for phase 1

1 existing design.

2 rounds of iterations.

3 rounds of usability testing.


Process for phase 2

Research → insights. 

Shopping carts →  quote comparison.

User research & interviews

Listening to Marble members' experiences

While running usability tests, we also found it important to validate the problem space through user research. Our team conducted interviews with 6 existing marble members. We asked questions that aimed to uncover our users' pain points, frustrations, and past experiences with purchasing insurance.

Affinity map time-lapse

We then synthesized our findings through affinity mapping and distilled insights from strong groupings and themes.

Watch us shake our butts.

Key takeaways

Personas + user journey

Designing for our target audiences

After we gathered our thematic insights, we created personas that reflected our target audiences. And with one of our newly crafted personas, we mapped out a user journey that shows what a potential Marble member would experience when purchasing home insurance for a new property. We did this to see if there are any areas of opportunity for us to design a better user experience for someone like Marilyn. I combined the persona and journey in the diagram below.

Competitive analysis & task flows

Understanding how we compare

To better understand our competitive landscape, we also mapped out the following task flows with Marble's competitors. We sought to identify industry standards, what works and what doesn't, and to see if there are any features Marble could be missing that users are expecting. We completed these tasks in May 2022, flows may have changed since you read this.

Our task was to shop for home insurance. 

Our goal was to see prices. 

Our metric to measure success was the average time on task.

Key takeaways

Design studios & workshops

Generate & ideate. Evaluate & sift. Rinse & repeat.

Design workshop

As an early effort to generate ideas, we hosted a design workshop with our clients. The goal of this workshop was to uncover any themes on what shopping means to us. We asked participants, what do you consider when shopping online?

We timeboxed our brainstorming and affinity mapping. This exercise was the early stage of questioning if the confusion from usability testing was rooted in the "shopping cart" model. Additionally, we started thinking about what value means to shoppers.

Revised Problem Statement

Many customers were unsure if the time they spent researching correlated to how satisfied they felt with what they shopped for. In the shoes of our persona, Marilyn needs a way to understand the difference between what various insurance companies are offering and to purchase a valuable and affordable policy.

How might we provide shoppers a way to compare insurance quotes between different companies so they can make informed decisions on the best policy for their needs?

Design studio

We held various design studios throughout our process. For the design studio that led to the MVP of our redesign, we used a MUSHCOWO map and feature prioritization matrix. We came up with features to keep from the existing website and added new features that would address the pain points and insights we uncovered during research. 

We prioritized these new features & changes...


As a way to generate and explore a lot of ideas and concepts quickly, we ran rapid-fire timeboxed design studios. Our team sketched, ideated, iterated, and debriefed on what works and what doesn't. We focused on addressing the insights we got from our research. In particular, we focused on...

Existing vs proposed screens

Through conducting business research, user research, usability testing, business analysis, hosting design exercises, and analyzing existing research, we designed screens that addressed issues we uncovered along the way.


Shop → explore quotes 

Quote compare


The redesign of Marble's shop feature!

We sought out to provide a new perspective on what the shopping experience could be. And we did just that. We're really excited and proud to show you how our flows work. 

Shopping for a new policy

Adding a new reminder

Shopping for an expiring policy

Additional consideration...

Leveraging a browser plugin/extension

We also considered an AI-powered browser extension. The potential of this extension is massive. From auto-filling information, to automatically uploading quotes into the Marble system, browser extensions could change the way shoppers compare insurance quotes online.

What's next?

Next steps, recommendations, and reflections

Our recommendations were to optimize the existing design and to focus on the quote comparison feature. If we were contracted to continue our work, we would love to repeat this entire process - starting with understanding how users compare quotes. Additionally, we'd like to explore the browser extension to be able to automate the quote comparison feature.

When starting the project, our team was very overwhelmed with the sheer amount of information we had to catch up on. Combining the needs of the business with the needs of the users is always a challenging endeavor. Now that we've finished the 3-week sprint, I am overwhelmed with a different feeling. I am so grateful to have been part of this project and to have worked with some amazing people. 

There seems to be a cloud that surrounds insurtech at the moment. It is a relatively new space with a ton of potential. I'm excited about the Marble's future. Special thanks to Marble, GA, and most importantly, my colleagues.