About
Resume
SIES Products
As the product designer at SIES Products, I am responsible for the end-to-end design of a high-end tea equipment Shopify store. The work is data-driven throughout — using funnel analysis, heat-map, and A/B testing to continuously diagnose problems, design solutions, and validate results.
Key results
Role & Timeline
•
Product Designer
Methods
Funnel Analysis / Heatmap Analysis / Competitive Analysis / A/B Testing / Design Audit / Visual design / Data-Driven Optimization
Challenge
SIES is a high-end tea equipment brand selling through Shopify, primarily reaching users through Instagram and Facebook advertising. After the initial store launch, data analysis revealed that purchase conversion and user retention were gradually declining. Users were arriving on the site with interest but not completing purchases.
The core challenge was: the metrics were underperforming but the exact cause was unknown. The work required progressively narrowing down where the problem lived before any design changes could be made. The core question became: How might we identify where users were dropping off and redesign the experience to convert interest into purchase?
Interest-to-purchase conversion gap
Defining the Problem
Traffic Source Analysis
Understanding who was arriving on the site was critical to designing for the right user. Analysis of traffic sources revealed that the majority of users were coming from paid social advertising, primarily Instagram and Facebook, with Instagram accounting for over 20,000 sessions compared to Facebook at approximately 3,900.
Breaking this down by device revealed that paid social traffic was overwhelmingly mobile: over 90% of Instagram and Facebook sessions happened on mobile web, compared to roughly 63% for direct and organic search. Since paid social was the largest traffic source, this meant mobile web was where most users, across the entire site, were actually arriving.
Traffic source data visualization
Funnel Analysis
The first step was mapping the full mobile web purchase funnel to identify where users were being lost. The data showed a clear picture:
The product page was the critical drop-off point. Users were reaching the product page in reasonable numbers but failing to move forward to checkout. The problem was on the product page, but the data alone could not explain why.
funnel visualization showing drop-off at each stage
Heatmap & Event Tracking Analysis
To understand what was happening on the product page, heatmap and event tracking data was analyzed. Several patterns emerged.
Heatmap screenshot showing user attention patterns on the product page
01 Key information was buried below the fold
Users spent most of their time above the fold and rarely scrolled further. Most lost interest within the first one to two scrolls, leaving little chance for the product story, usage scenarios, or social proof to be seen at all.
02 Unanswered questions had nowhere clear to surface
Product materials, capacity, and FAQ content existed, but none of it was highlighted near the top. Users were left without clear answers exactly when they needed them most.
03 Real-use context was missing where attention was highest
The page had little showing the product in actual use. Without scenario-driven visuals near the top, users couldn't picture themselves using it.
04 Social proof arrived too late to matter
Reviews and trust signals were placed at the very bottom, far past where most users had already left.
Design audit
To translate these findings into action, I audited the existing product page section by section, mapping what each part was doing against what users actually needed at that point in the page.
This audit produce a revised content hierarchy and a clear scope for the redesign: what to cut, what to move higher on the page, and what to add. It became the direct blueprint for the structure used in the next phase of design.
Design audit on the editor
Entire design audit figjam
Structure analysis
Before diving into the design, I reviewed product pages across a range of comparable brands to understand two things: what content they typically include, and in what order they place it.
Left: Screenshots of structure analysis; Top right: Recording of the competitors; Bottom right: Screenshots of the competitor
Three principles emerged from the analysis to guide all subsequent design decisions.
Surface conversion signals early
Ad-driven users with low brand familiarity need to see price, bundle options, and core selling points immediately. Every additional scroll required before reaching this information increases drop-off risk.
Let story support conversion, not replace it
Lifestyle photography and brand narrative are important for a premium product but must be positioned to support the purchase decision rather than delay it.
The Design
With all analysis complete, the product page was redesigned around three core design principles.The page was restructured into three distinct zones based on the conversion, story, and trust framework.
01 Above the fold · Conversion
Core selling points, bundle options, price differentiation, and a fixed CTA button visible at all times.
Design audit on the editor
02 Middle · Story
Lifestyle photography and short product usage videos helping users connect emotionally with the product.
Upper row: Before vs Lower row: After
Data-Driven Iteration
A/B Testing
The testing process followed a consistent loop: identify an underperforming metric, analyze the data to narrow down the cause, design a change, run an A/B test, confirm or reject the hypothesis, and iterate.
User reviews, FAQ in accordion format, awards and media features consolidated in one place.
A/B test different layout pattern for conversion rate
Final Redesign version
Final Design
User reviews, FAQ in accordion format, awards and media features consolidated in one place.
Final design
Repurchase Experience
Post-purchase discount preview and email retargeting flow designed to bring users back at the right moment.
repurchase flow screens
Impact
The data-driven redesign produced measurable improvement across all key conversion metrics.
The most significant improvement was in checkout completion, where removing unnecessary steps and adding mobile payment options directly addressed the friction identified in funnel analysis. The product page purchase rate improvement validated the hybrid page structure identified through competitive benchmarking.
Next Steps
Continue iterating on the product detail page based on ongoing heatmap and A/B testing data.
Explore personalization opportunities using purchase history data to surface relevant products to returning users.
Reflection & Learnings
The main challenge of this project was diagnosing a conversion problem without knowing where it lived. The discipline of following the data before making any design changes: mapping the funnel, analyzing heatmaps, understanding traffic sources, and benchmarking competitors; meant that every design decision was grounded in evidence rather than assumption.
The biggest lesson was that understanding who your users are before designing for them changes everything. Knowing that the majority of SIES users arrived from paid social advertising with low brand familiarity completely changed the information hierarchy of the product page. The same product, positioned differently for a different user, converts at a fundamentally different rate.
Competitive Analysis
To understand how high-converting e-commerce brands structure their product pages, I analyzed Fellow, Pureover, Millab, and 10 additional brands across four dimensions: homepage above-the-fold content, module order, product detail page information structure, and CTA placement.
Three distinct product page patterns emerged:
