Profile Vision
Vision & Experimentation Methodology
Overview
I led a strategic redesign of Vinted Profile tab and Seller’s Wardrobe—driving the initiative from identifying a small but critical opportunity to shaping a research-driven, cross-domain product vision.
I aligned design, product, engineering, and data around a shared north star, balancing seller-focused business OKRs with buyer and transactional needs. Through iterative experimentation, I helped the team navigate legacy design debt, conflicting objectives, and real-world constraints to make more intentional, data-informed product decisions.
This work was also shared externally: I presented the Profile Vision and its experimentation learnings at a UX Salon Event in Vilnius, Lithuania --> watch it on youtube.
The Trigger: A Catch-All Without a Vision
The Vinted Profile wasn’t broken—but it wasn’t intentional either. As the product evolved, our Profile slowly turned into a catch-all surface, absorbing new features without a clear sense of hierarchy or purpose (see image below). What looked manageable from the inside felt increasingly overwhelming to users. During user interviews, one quote captured the problem perfectly:
“It feels like they didn’t know where to put these things and they just dropped everything there with no organisation.”
It was uncomfortable—but accurate. The Profile tab had become a catch-all surface without a clear purpose, structure, or ownership.
 |
|---|
Context & Problem Space
The Profile is the 5th tab in Vinted’s main navigation and a core weekly-used surface. From here, users access their Wardrobe and several critical buyer and seller entry points.
We identified three intersecting challenges:
1. Legacy & Ownership
The Profile hadn’t been meaningfully updated since 2018, accumulating design debt and lacking a clear owning domain.
2. User Experience
Users perceived the Profile surface as disorganized, with important actions buried or unclear—especially for sellers managing listings.
3. Business Misalignment
New Q1 2024 business priorities—including DAC7 (new EU tax reporting requirements that temporarily reduced seller activity) and a push to increase listings—introduced features on the Wardrobe surface that pushed sellers’ own items below the fold, conflicting with seller needs and listing OKRs.
 |
|---|
At the same time, we had to account for multiple user types:
-
56% of users both buy and sell
-
27% are buyers only
-
15% are sellers only
This raised a critical question: What happens if a buyer-only user lands on an empty Wardrobe? Solving for sellers alone was not enough.
Leadership Decision: Defining a Vision
I paused incremental delivery and proposed we define a Profile Vision—an inspiring but flexible north star to align teams and guide decisions.
A key leadership challenge was cross-domain collaboration. While our domain OKRs focused on sellers, Profile also contained buyer-critical and transactional entry points (e.g. favorites, orders, balance) owned by other domains. I partnered closely with those teams to understand their OKRs and ensure experiments would not negatively impact buyer funnels.
Building the Vision: Research & Alignment
I led and synthesized insights across multiple inputs:
1. Desk Research: Past qualitative insights, competitor research (C2C marketplaces such as Etsy and Kleinanzeigen, dual-role platforms like Airbnb hosting vs. traveling, and general best practices for Profile surfaces.
2. Card Sorting: Understanding user mental models (see example of digital open card sorting below)
3. Behavioral Data: Profile usage tracking of how many users visit the profile entry points at what frequency. Also via card sorting data, comparing the real usage to perceived usage (left).
4. Business OKRs: Listings, listers, revenue
5. User Interviews: wireframe concept testing between 3 potential directions:
-
A dual tab system splitting the buyer and seller profile experience
-
A dashboard like Profile with all needs in one place
-
Bringing the wardrobe content one click sooner as the profile
Key findings:
-
54% of Profile visitors go directly to their Wardrobe
-
Out of 16 entry points, most were rarely used
-
Sellers view the Wardrobe as a frequent “check-in” space
-
Buyers rely on Profile for shortcuts like favorites and orders
The Initial Vision
The initial vision proposed:
-
Prioritizing the Wardrobe as the primary destination
-
Reducing friction to listing management
-
Elevating listings above the fold
-
Preserving key buyer and transactional entry points
Experimentation & Validation
From the start, I treated experimentation as a leadership tool—to align teams, de-risk decisions, and continuously test assumptions.
Together with Data Science & Analytics (DSA), we established a shared experimentation framework and success metrics across UX, seller, buyer, and revenue outcomes.
Our approach:
-
Define hypotheses grounded in research, data, and business goals
-
Design small, focused A/B tests in close collaboration with DSA
-
Measure outcomes across UX, listings, buyer funnels, and revenue
-
Learn, iterate, or pivot based on evidence
Beyond individual experiment outcomes, the experimentation approach itself proved highly successful. It aligned design, product, engineering, and data around a shared decision-making framework, reduced opinion-driven debates, and created psychological safety to test, learn, and pivot. The methodology became a reference point for how complex, cross-domain experimentation could be done well and gained visibility across Vinted:
“Overall, your presentation will be a good reference for all DSA people of how things are supposed to be done. You’ve made massive and lasting impact.” — Director of Data & Analytics
“It’s getting attention in multiple channels—thank you for being a good example.” — Staff Experimentation Analyst
Experimentation in Practice
To bring the vision to life, we ran a series of incremental experiments, each validating a specific hypothesis.
Seller Insights and badges entry points, item list view & adding edit functionality at the Wardrobe Level, all working up to the Profile ↔ Wardrobe Swap
Hypotheses included:
-
Do features such as insights and badges encourage sellers to list more?
-
Do sellers want faster access to item management?
-
Does easier item editing improve listing quality, engagement or sell through rate?
Key Learning: Letting Go of the Vision
Despite strong qualitative feedback favoring a Wardrobe-first experience, the large-scale experiment failed to deliver expected impact. Buyer regression and net business loss led us to stop scaling.
This became a defining leadership learning moment. Industry-standard success rates show only ~⅓ of experiments scale. Failure is expected and learning is success.
We pivoted away from the swap and returned to a sidelined direction: reimagining Profile as a dashboard—a surface we fully own that serves buyers and sellers while nudging business priorities.
Decision-Making & Metrics Challenges
This project highlighted tensions between qualitative insights and quantitative outcomes—what users say vs. what they do at scale.
It also reinforced the importance of defining success metrics before launch, not retrofitting narratives afterward. For example, assumptions linking item editing to increased sales proved difficult to validate due to external variables (seasonality, price, item quality).
We also created a new UX metric: Wardrobe Stickiness (return visits). While useful, it revealed ambiguity—high engagement could signal motivation or confusion.
Leadership Challenges & Alignment
The project unfolded alongside a major company shift toward new verticals (e.g. luxury, electronics, sports). Design leadership encouraged supporting these categories, while Data leadership prioritized seller-domain validation. Navigating these tensions required negotiation, reframing success, and maintaining trust across functions. This surfaced a growth area for me: proactively aligning stakeholders earlier and more frequently as the vision evolved.
Another leadership challenge was maintaining engineering momentum while the vision was still forming. Pausing incremental delivery to define direction created periods where developers lacked concrete, immediately shippable work. While this was primarily a product planning concern, it directly affected team morale and required active leadership: communicating intent, reinforcing trust in the process, and helping engineers believe in the long-term value of the vision before outcomes were fully visible.
What I’d Do Differently
I would socialize the vision continuously—not just at creation. Earlier and broader stakeholder involvement during evolution could have reduced friction and accelerated alignment.
Additionally, leadership grew concerned as we iterated and pivoted without a consistently updated, shared visual articulation of the evolving vision. This was a missed opportunity. Rather than waiting for highly polished, “final” assets, we could have shared rough but directional visuals more frequently to maintain alignment, confidence, and buy-in throughout the journey. Continuous visibility would have reduced uncertainty and reinforced that iteration—not perfection—was the goal.
Next Steps
We continue evolving the Profile Vision by:
-
Treating Profile as a dashboard we fully own
-
Making buyer, seller, and transactional entry points intentional
-
Continuing to ship small, measure impact, and adapt
Reflection
This project reshaped how we approach legacy redesigns at Vinted:
-
From execution to vision-led strategy
-
From assumptions to validated learning
-
From fixed roadmaps to adaptive experimentation
The Profile Vision remains a living artifact—guiding decisions while staying open to change.
Watch me present (an earlier version of the Profile Vision) at the UX Salon event in Vilnius Lithuania on youtube.