To capture a broad range of behaviors and expectations, we interviewed participants from MetaMask's most active markets, including the United States, United Kingdom, Canada, Brazil, and Nigeria, spanning novice to advanced crypto users. Research included existing MetaMask users, customers of centralized exchanges, and users of fiat-to-crypto aggregator products, providing a broader understanding of how people evaluated options, compared providers, and built trust when purchasing crypto. Despite differences in experience level, platform familiarity, and regional requirements, four recurring challenges consistently emerged: recovering from failed paths, understanding total cost, anticipating provider requirements, and trusting third-party handoffs. These insights helped define the core user needs the experience would need to address.

Alongside user interviews, I analyzed direct competitors, indirect competitors, and adjacent industries to understand how people compared options, evaluated tradeoffs, and built trust when making important decisions. Provider marketplaces revealed patterns for quote comparison and provider selection, while single-path purchase experiences highlighted opportunities to reduce friction through stronger defaults and simplified flows. Outside of crypto, ranking and comparison systems provided useful frameworks for communicating tradeoffs, surfacing recommendations, and helping users make confident choices quickly.

With a clearer understanding of user behavior, we translated research insights into a set of user needs and mapped them to specific product requirements across the purchasing journey. This process revealed that users weren't struggling with a lack of choice, they were struggling to understand tradeoffs, requirements, and outcomes when evaluating providers. By connecting user needs directly to feature opportunities, we created a shared framework for prioritization and ensured every investment addressed a validated problem while supporting future expansion across providers, payment methods, and regions.

Before exploring solutions, I partnered closely with Product and Engineering to map the end-to-end purchasing journey and identify technical constraints early. This exercise surfaced where provider variability, regional requirements, compliance considerations, and platform limitations would influence the experience, helping us distinguish true user problems from implementation constraints. Aligning on the flow upfront reduced rework, clarified MVP scope, and established a shared understanding of how users would move through the ecosystem from start to finish.

After aligning on the flow architecture and visual direction, I built an interactive prototype to validate key assumptions before investing in implementation. We tested the experience with 20 participants across multiple regions and experience levels, asking users to purchase $300 of ETH while selecting a provider and payment method. This helped evaluate how users navigated the purchasing journey, where confidence broke down, and whether the proposed framework successfully supported decision-making across different markets and levels of crypto familiarity.

Testing combined task completion metrics, participant feedback, and observational insights to understand how effectively the experience supported purchasing decisions. While all participants successfully completed the task and most reported confidence in selecting a provider, the research revealed recurring uncertainty around fee transparency, buy limits, and provider-specific requirements. These findings validated the overall direction while highlighting the areas where additional clarity and refinement would have the greatest impact on user confidence.

Launching the MVP created an opportunity to learn from real user behavior at scale. By monitoring funnel performance, conversion metrics, drop-off points, and qualitative feedback, we were able to understand how users actually navigated the purchasing journey beyond controlled testing environments. While the MVP successfully validated the overall framework, the data revealed several opportunities to improve clarity, reduce uncertainty, and strengthen confidence at key decision points. These insights helped prioritize where design investment would have the greatest impact on user outcomes and business performance.

Using insights from MVP performance and user feedback, I refined the purchasing journey around the moments that most directly influenced confidence and decision-making. Improvements focused on token selection, amount entry, payment method selection, and provider comparison areas that consistently influenced how users evaluated options and completed transactions. The resulting experience reduced friction, improved transparency, and created a more intuitive path from asset selection to provider handoff while preserving the flexibility required to support a global marketplace.

Before: Users were required to scan long, undifferentiated token lists with limited context for decision-making. Popular assets, owned assets, and trending tokens were presented within the same hierarchy, while repetitive network labels added visual noise without helping users narrow their choices. As the number of supported assets grew, the experience became increasingly difficult to scan and increased the likelihood of selecting the wrong token.

After: The final experience introduced clearer information architecture, separating popular, owned, and trending assets into distinct sections while adding network filters and search capabilities directly into the selection flow. Balance visibility, token verification cues, and prioritization of commonly purchased assets reduced scanning effort and helped users identify the correct asset more confidently and efficiently.

Before: The amount entry experience distributed important information across multiple controls, forcing users to interpret region, currency, token, payment method, and purchase amount simultaneously. The primary task entering a purchase amount—competed with surrounding interface elements for attention, while common purchase values required manual entry, adding unnecessary friction to a highly repetitive step.

After: The final design elevated amount entry as the primary focus of the screen while simplifying surrounding inputs. Region and currency selection were consolidated, common purchase amounts became accessible through quick-select chips, and unnecessary labels were removed to reduce visual noise. These changes created a clearer hierarchy that helped users progress more quickly while maintaining awareness of key purchase details.

Before: While provider quotes were technically comparable, users lacked sufficient context to understand why options were ranked the way they were or which offer best matched their needs. Token amounts, fiat values, fees, and provider trust signals competed for attention, making side-by-side comparison slower and increasing uncertainty during one of the most critical decision points in the journey.

After: The final quote comparison experience prioritized transparency and trust by clearly communicating ranking logic, strengthening provider context, and improving information hierarchy. Token and fiat values were balanced visually, provider recommendations became easier to understand, and actions were embedded directly within selected quotes to reinforce decision confidence. These changes reduced ambiguity, improved scanability, and helped users evaluate options more quickly without sacrificing flexibility.

Before: While provider quotes were technically comparable, users lacked sufficient context to understand why options were ranked the way they were or which offer best matched their needs. Token amounts, fiat values, fees, and provider trust signals competed for attention, making side-by-side comparison slower and increasing uncertainty during one of the most critical decision points in the journey.

After: The final quote comparison experience prioritized transparency and trust by clearly communicating ranking logic, strengthening provider context, and improving information hierarchy. Token and fiat values were balanced visually, provider recommendations became easier to understand, and actions were embedded directly within selected quotes to reinforce decision confidence. These changes reduced ambiguity, improved scanability, and helped users evaluate options more quickly without sacrificing flexibility.