Help buyers move from curiosity to conviction in a messy, high-variance second-hand world.
Our goal is to reduce decision friction and build confidence so more item clicks become transactions, in the process of selecting items of interest, analyzing them and then making a choice regarding them.
3 main parts of the user journey, which Selva focuses on
Item cards in
Search Results page
Item Detail Page (IDP)
Favorited items
Because our touchpoints span multiple teams and dependencies, we split our strategy into two pillars: improving the user journey and strengthening the system that supports it.
- Pillar 1: Assessing value at a glance
- Pillar 2: Evaluating, examine with confidence
- Pillar 3: Comparing & considering, decide now or later
- Trust: seller confidence & trust signals
- Ads & UX coexistence: positive sum co existence
- E2E mapping: dependency mapping, ensure improvements connect smoothly, avoid breaking existing experiences
As we implement a series of changes across the different touchpoints we need to ensure we deliver a consistent cross-platform experience with solid mobile integration, a new web layout (consistency not parity), low-complexity experiments, and strong observability.
The Item Detail Page is one of the most critical touchpoints in the journey, with over 8M visits per month. Nearly every buyer passes through it before making a purchase, and it is often where the final decision is made.
While it has various interdependencies, it remains the touchpoint we can most directly influence: since it is a core space rather than just a component embedded elsewhere.
Evaluation instances are the key moments in the buyer journey where users process information to decide whether to move forward.
This can happen right away or later, depending on their mindset. Because users approach evaluation with different profiles and intentions, the design must flexibly support all these states to ensure decisions feel clear, trustworthy, and actionable.
The item detail page is a complex page due to the amount of information which it contains, the different categories it needs to cater for, the ads & brand campaigns it must hold and the need for it to be an "action" page where the user converts
Item Detail Page (IDP)
with all sections active
Item details variants for buyers
& sellers & per use case
We ran 1o1 user interviews focused on the item detail page to better understand what their key challenges were and ensure we were prioritising the correct touchpoints.
Quotes from users
Clarity & consistency: Info is scattered, hard to compare
Fragmentation: context lost across touchpoints, restart often, length of page
Category tension: different UIs across Motors, Fashion, Furniture, etc.
Off-platform validation: buyers leave site to check specs/price/authenticity
Trust & incomplete info: Seller credibility, total cost, photos unclear
Internal complexity: tech debt, duplicated logic, legacy favorites infra, non-modular IDP, ad clutter blocking improvements
So far (4 months in) we have prioritised and focused on 6 main areas of the item detail page, to tackle both the key user pain points and the more structural ones, ensuring we had bigger bets and also low hanging fruits to work on different levels of depth.
6 prioritised initiatives based on impact & complexity
The key leading metrics we are monitoring are:
- Sold Items
- CVR to TRX from the IDP & Item Cards
Secondary & Health metrics
- CVR to PI
- Make an Offer to TRX
- CVR from Buy & Chat
Guardrail metrics:
- Avg Item Price to Transaction
- Engagement across events
- Common paths
Previous Item Detail
Iterated Item Detail
Hypothesis: Buyers want to see the product title and key attributes immediately when they open the item detail page, to quickly understand what the item is and whether it's relevant to them.
Impact:
- Elevate key product information for faster decisions (+4,2% in web PI)
- Improve perceived quality of listings
- Reduce reliance on scrolling to evaluate value
- Reduce cognitive load in the top section
Hypothesis:
Moving the like button and introducing a counter in the UI will have buyers perceive it as high-demand, feel more confident, and be more likely to purchase.
Impact:
- Surpassed launch criteria: +6.38% increase in Favorite Rate from IDP (vs. +2% target)
- Validated hypothesis: social proof (favorite count) boosts buyer confidence and engagement
Guardrails held: No negative impact on conversion metrics, including low-favorited items
Hypothesis:
We believe that by moving the “Make an Offer” CTA from the chat area, both in web & app, users will notice it more easily, leading to increased offer interactions and a smoother, more accessible experience.
Overall impact
CVR from Click Item Card → TRX +1.11%
Total Offers +25.9% | Avg. Offers per User +2.73%
Positive outcomes
CVR from Make an Offer → TRX +16.6%
CVR from Click Buy → TRX +0.9%
Mixed effects
CVR from Click Card → Chat -1.6%
CVR from Click Card → Buy Button -6.4%
Next steps / monitoring
Assess potential negative effects on sellers receiving more offers (e.g., deal quality, negotiation effort).
Hypothesis:
If we replace multiple product detail UIs with a single, unified component that organizes all item details clearly and uses collapsible sections, buyers will understand item specifics faster, explore details more efficiently, and make more informed purchase decisions without feeling overwhelmed.
Results:
- +3.91% uplift in CVR (Click Item Card → Purchase Intention)
- In CSG categories there was a slight negative impact in the YoY purchase trend
Takeaways & Iteration:
- Clearer, more consistent IA improves user evaluation and conversion
- In categories with poor attributes population, the description becomes really important and an iteration is needed
Hypothesis:
If we show the information linked to the delivery and the possible fees linked to it up front, while removing visual clutter, buyers will be able to decide faster whether the item works for them without navigating elsewhere.
Results:
Both variants negatively impacted the primary metric (CVR Click Card → TRX)
- Variant A (price below the fold): -0.8%
- Variant B (simplified module above the fold): -1.15%
Durable Learning
Displaying all transaction fees on the Item Detail Page can harm conversion and introduce friction.
This insight will guide future design decisions and help avoid similar mistakes.
Hypothesis:
If we optimize the layout of the seller profile section to be clear, concise, and scalable, users will better understand seller credibility, trust the seller more, and engage with profile actions more consistently across the platform.
- Balance between lean and fast experiments vs solid solutions that apply to all use cases of the item detail page & cross mobile/web
- Experiments have a trade-off for engineering (increasing cognitive load, code complexity, harder maintenance, etc) and needs to be actively handled
- Iterate the experiments which either failed or seem promising
- Unify the rest of the UI of the IDP whilst aligning with the other stakeholders
- Start tackling the other 2 pillars (assessing & consider later) to spread their impact on the IDP too
