Start with the why, not the screen
Every feature began with what it was for and who it helped, and competitors were read as references rather than answers. The interface was the last question, not the first.
Overview
Designing a better trading experience for the Ronin ecosystem.
From June 2024 to March 2025, I worked as a Product Designer on Mavis Market at Sky Mavis, later rebranded as Ronin Market.
This was my first experience working deeply inside a product team after coming from an agency background. What started as a focus on visual design and user flows gradually evolved into a much broader product design role — working across research, competitive analysis, UX flows, UI design, technical constraints, developer collaboration, and product iteration.
During this period, I contributed to a wide range of marketplace experiences, including:
- Add to Cart
- Sweep
- Bulk Buy
- Presale Ticket
- Global Search
- Verified Badge
- Credit-card checkout via Transak
- Collection Offer for ERC-721
- Trait Offer
- Collection Offer for ERC-1155
- Trading Contest
- Compact / List View
- Collection Detail Revamp
- Action Controls
- Search Bar behavior
More importantly, the project gave me the opportunity to evolve from designing interfaces to designing a real product within a complex ecosystem.
The context
From Mavis Market to Ronin Market
When I joined the team in June 2024, Mavis Market was focused on rapidly delivering core marketplace functionality while improving the overall trading experience.
The first phase was highly execution-oriented. We had to understand the problem, study how other marketplaces solved it, align with engineering constraints, and move quickly toward implementation.
At the beginning of 2025, Mavis Market officially rebranded as Ronin Market to better align with the broader Ronin ecosystem. The rebrand also represented a larger product shift.
Ronin was opening its marketplace ecosystem to a broader range of creators and collections. Assets were no longer limited to collections curated and verified by Sky Mavis, and trading could happen across a more open ecosystem.
This created a new challenge: how could Ronin Market become a marketplace people could not only browse, but actively choose to trade on? That question became the foundation for the second phase of my work.
The challenge
The marketplace had several friction points that directly affected discovery and trading. Through competitive research, audits, and later user interviews, we identified several recurring problems.
Marketplace actions were difficult to understand
Important actions were distributed across the interface in ways that could feel fragmented or confusing. Users needed clearer relationships between browsing, viewing an NFT, making an offer, and completing a transaction.
Search was not intuitive enough
Search behavior did not always match users' expectations when looking for collections, NFTs, or specific marketplace content.
NFT browsing was inefficient
Large cards consumed too much vertical space, limiting how many NFTs users could see at once. For a marketplace where the primary goal is to discover an asset and make a trade, excessive scrolling introduced unnecessary friction.
Key trading mechanisms were missing
Compared with established marketplaces, Ronin Market lacked important trading mechanisms such as Collection Offers and Trait Offers. These features were particularly important because they could give buyers more flexibility while creating additional opportunities for sellers.
The marketplace needed to compete in a broader ecosystem
As Ronin became more open, users could discover and trade assets across a wider ecosystem. Ronin Market therefore needed to make its own trading experience compelling enough to retain users and encourage transactions on the platform.
2024 — Learning to build a product
Starting with the why
My first phase at Sky Mavis was heavily focused on understanding how a product team operates. For every feature, we started with a simple question: why are we building this?
Before jumping into UI, we aligned on the purpose of the feature and its expected impact on the marketplace. From there, I contributed to competitive research, UI audits, user-flow exploration, wireframing, and visual design.
Because of the time constraints, the process was intentionally lean. We did not always have the opportunity to conduct formal user interviews, so we relied heavily on hands-on exploration, competitive analysis, internal discussions, testing, and iterative design.
The 2024 design process
Clarify the why
Understand the value the feature would bring to Mavis Market and how it supported the broader product.
Competitor research & UI audits
Study how other marketplaces approached similar problems. Document their solutions, strengths, weaknesses, and the reasoning behind different design decisions.
Review & compare
Compare possible approaches with the team. Challenge assumptions, discuss trade-offs, and identify opportunities rather than simply copying existing patterns.
Exploration & technical alignment
Start with low-fidelity flows and work closely with developers to understand technical constraints. Once feasibility was confirmed, move into high-fidelity design.
Team grooming & UI delivery
Present the solution to the marketplace team, gather feedback, resolve edge cases, and prepare the final UI for development.
UAT & launch
Validate the implementation, review edge cases, and work with the team to ensure the final experience matched the intended design.
A key lesson: Presale Ticket
One of my biggest learning moments during this phase came from the Presale Ticket feature. The first version lacked the confidence and polish we wanted.
Instead of treating that as a failure, we used community feedback and internal review to revisit the design. The feature went through another round of exploration and refinement, resulting in a much stronger solution.
A design iteration is not a correction to the process. It is part of the process.
I learned to treat failure as a signal for deeper understanding rather than as an endpoint.
2025 — Designing with evidence
By 2025, the team had more time and a stronger foundation. Our process expanded beyond competitive research to include user interviews. This was an important shift in how I approached product design.
Instead of asking only how other marketplaces solve this, we could now ask what our users are actually struggling with — which let us connect competitive patterns with real user needs and prioritize problems based on evidence.
Product goals in 2025
With Ronin becoming a more open ecosystem, the marketplace needed to improve both accessibility and trading performance. Our goals were centered around:
- Making Ronin Market easier to navigate and browse.
- Increasing opportunities for buyers and sellers to transact.
- Bringing marketplace functionality closer to established competitors.
- Supporting a more open creator and collection ecosystem.
- Increasing the conversion from marketplace visitors into buyers and sellers.
- Improving the overall view-to-trade conversion rate.
During Q1 2025, the improvements contributed to up to an 11% conversion of marketplace visitors into buyers and sellers. The team also targeted a 25% increase in view-to-trade conversion rate during 2025.
Research & discovery
From competitive research to user interviews
We combined two sources of information.
Competitive research — understanding how established marketplaces solved similar problems, and identifying patterns that could be adapted to Ronin Market.
User interviews — speaking directly with users to uncover pain points that could not be fully understood through competitive analysis alone.
Several important insights emerged:
- Marketplace actions felt fragmented. Users had difficulty understanding where and when different marketplace actions should happen.
- Search needed to be more intuitive. The existing experience did not always support the way users expected to discover marketplace content.
- Dense NFT discovery was important. Users wanted to see more NFTs at once rather than spending excessive time scrolling through large cards.
- Offers were an important missing mechanism. Collection Offer and Trait Offer stood out because they let buyers express intent without having to target a single NFT.
These findings gave us a stronger foundation for deciding what to solve first.
How we prioritized
After completing research, we synthesized the findings and prioritized them by user impact, product value, and technical feasibility together.
This helped the team move away from designing whatever seemed interesting and instead focus on the problems most likely to improve the marketplace experience.
Feature focus: Collection Offer & Trait Offer
One of the major areas I worked on was the introduction of Collection Offer and Trait Offer.
Collection Offer
Collection Offer allows a buyer to make an offer for any NFT within a collection at a price they choose. It changes the interaction from wanting one specific NFT to being interested in a collection and willing to buy an asset from it at a preferred price.
This creates more flexibility for buyers while opening additional opportunities for sellers.
Trait Offer
Trait Offer extends the same concept by allowing buyers to target NFTs with specific traits. Instead of making an offer against an entire collection, users can express a more precise preference.
This gives buyers greater control over what they want, and makes the offer mechanism far more relevant to collections where individual traits significantly affect value.
Designing around complex constraints
The most challenging part of these features was not the visual design. It was designing around on-chain and off-chain behavior.
Some marketplace actions required blockchain transactions, while others could happen without one. Accepting an offer could require gas and user confirmation. Managing offers involved different transaction states. Transactions could fail, offers could expire, users could cancel — and different states could occur before, during, or after a transaction.
Every one of these states needed to be considered. The challenge was to make a technically complicated system feel simple to the user.
Technical review & constraints
We worked closely with engineering to identify technical limitations before finalizing the experience. The design had to account for:
- On-chain actions and gas fees.
- Off-chain actions.
- Transaction states, including pending and failed transactions.
- Expired offers.
- Cancelled offers.
- Different user states.
- Edge cases across multiple marketplace actions.
This became one of the most valuable parts of my product design experience at Sky Mavis. I learned that a polished UI is only one part of the solution — a good product design also needs to survive the reality of the system underneath it.
Iterative flow refinement
After the initial exploration, we went through multiple rounds of flow refinement. Each iteration focused on answering three questions:
- Is the journey understandable? Can users understand what will happen before they commit to an action?
- Are edge cases handled? What happens when a transaction fails, expires, is cancelled, or requires another action?
- Does the experience remain technically feasible? Can engineering realistically support the intended interaction without introducing unnecessary complexity?
Only after these questions were resolved did we move toward final UI delivery.
Other marketplace improvements
Alongside Collection Offer and Trait Offer, I contributed to several other improvements across Ronin Market.
- Collection Offer for ERC-721 & ERC-1155 — expanding offer functionality across different token standards.
- Trading Contest — exploring marketplace mechanics that could encourage engagement and trading activity.
- Compact / List View — improving NFT browsing density so users could scan more assets with less scrolling.
- Collection Detail Revamp — restructuring the collection experience and improving how users interacted with collection-level actions.
- Action Controls — reconsidering where marketplace actions should live and how users discover them.
- Search Bar Behavior — improving search interactions and making discovery more intuitive.
Collaboration
Product design at Ronin Market was highly collaborative. I worked closely with product, engineering, other designers, marketplace stakeholders, and with users through community feedback.
The process was rarely a simple sequence of design, handoff, build. It was a continuous loop: research, explore, discuss, validate, refine, build, test, iterate.
Working directly with developers was especially important. Technical conversations often changed the design, while design exploration sometimes exposed technical problems that needed to be solved before implementation.
What changed for me
My biggest takeaway from this project was the transition from designing screens to designing systems. When I joined Sky Mavis, my strongest background was in visual and interface design. By the end of the project, I had developed a much broader understanding of product design:
- Start with the problem, not the interface.
- Understand why a feature exists before designing it.
- Use competitors as references, not answers.
- Validate assumptions with users whenever possible.
- Design for edge cases, not only the happy path.
- Collaborate with engineering early.
- Treat iteration as part of the process.
- Measure whether the solution actually changes user behavior.
The most important shift was realizing that my responsibility did not end when the UI looked good. I was designing a product that had to work for users, the business, and the technical system simultaneously.
Impact
Across the 2024–2025 period, my work contributed to the evolution of Ronin Market from a marketplace focused on core functionality into a more mature trading platform. The work helped:
- Expand the marketplace's trading capabilities.
- Introduce more flexible offer mechanisms.
- Improve NFT discovery and browsing.
- Address accumulated user feedback.
- Bring the product closer to competitive marketplace patterns.
- Create experiences better aligned with user needs.
- Support the broader Ronin ecosystem as it became more open.
- Contribute to converting up to 11% of marketplace visitors into buyers and sellers during Q1 2025.
Reflection
Working on Mavis Market and Ronin Market was an important transition point in my career. In 2024, I learned how to operate inside a real product environment. In 2025, I learned how to use research, data, technical constraints, and iteration to make better product decisions.
The project taught me that product design is not about producing the perfect screen on the first attempt. It is about continuously reducing uncertainty.
Understand the problem. Explore the possibilities. Test the assumptions. Work with the constraints. Learn from failure. And keep improving the product. That is the mindset I took away from Ronin Market.
Design process snippets
Takeaways
Iteration is the process, not a correction to it
The first Presale Ticket shipped weaker than we wanted. Treating that as a signal instead of a verdict is what produced the version that worked.
A design has to survive the system beneath it
Gas fees, pending and failed transactions, expired and cancelled offers — the on-chain reality decided the flow. A polished UI is only the part of the solution you can see.