Improving the performance of Farfetch's top fashion eCommerce brands

Summary and outcomes

I joined Farfetch as the first Product Designer to lead a new initiative aimed at improving the performance of B2C clients operating on the Farfetch e-commerce platform. This included working with brands such as Off-White, Emilio Pucci, Roberto Cavalli, Ami, Stadium Goods, and Browns Fashion. I led efforts to establish fundamental design practices and processes focused on optimizing e-commerce performance indicators and enhancing the overall user experience.

We collaborated with over 12 different fashion brands to improve the results of their online retail stores. In one of our first projects, we worked with Pucci, achieving an 87.5% increase in conversion rate, a 57% reduction in bounce rate, and an 8.69% revenue growth from 2019 to 2020.

About Farfetch

Farfetch is a leading global marketplace in the luxury fashion industry. The Farfetch Marketplace connects customers from over 190 countries with more than 1,400 of the world’s top brands, boutiques, and department stores, offering a unique shopping experience and access to the largest selection of luxury fashion on a global marketplace.

In addition to its marketplace, Farfetch launched Farfetch Platform Solutions (FPS) in 2015 — a suite of commerce solutions and retail technology for luxury brands and retailers. FPS enables brands and retailers to reach new customers, expand globally, and offer an elevated shopping experience both online and in-store, serving luxury brands such as Harrods, Off-White, Chanel, Stadium Goods, Palm Angels, Pucci, and Roberto Cavalli.

My role and responsibilities

At Farfetch, I led efforts to establish key design practices and processes focused on optimizing e-commerce performance indicators and enhancing the user experience for boutiques on the Farfetch platform. I worked closely with 3 Product Managers, a Data Analyst, and a UX Researcher, providing deep insights into customer behavior, friction points, their impact on the business, and proposing solutions to improve key performance metrics.

My key responsibilities included:

Leading design strategy and establishing practices by applying UX Research, Usability Testing, Data Analysis, Benchmarking, and Sketching to address friction points and meet customer needs, turning them into high-value business solutions.
Cross-referencing data and insights from multiple sources to design solutions based on complexity and business impact.
Collaborating closely with the development team to implement and monitor the success of implemented features.
Providing design consultancy and sharing knowledge with internal teams and stakeholders across the company.
Establishing processes, design practices, and ceremonies for onboarding new brands onto the Farfetch e-commerce platform.
Proposing new features for the Farfetch platform based on learnings and insights gathered from clients.

Methods and tools

For each project at Farfetch, a variety of methods and tools were used to generate insights, which were then cross-referenced to identify opportunities for improvement. The Optimization team worked closely with the business, engineering, marketing, and commercial teams to plan and prioritize solutions that would enhance the performance of our partners. Below are the main tools we used in our analysis:

UserTesting.com: a platform for gathering quick customer feedback through remote user testing. We used this tool to recruit fashion shoppers, identify pain points in their purchasing journeys, and test new features.
Hotjar: a product experience insights platform that provides behavioral analytics and feedback data to better understand customer behavior. We primarily used heatmaps to identify areas of focus and recordings to analyze navigation patterns.
Google Analytics: our main analytics tool for measuring website performance. Google Analytics helped us track key performance indicators and events, giving us insights into how users interacted with the websites.
A/B Testing: a methodology for comparing two versions of a webpage to determine which performs better. Farfetch has an extensive knowledge base of A/B tests conducted across its eCommerce brands, providing a valuable source of insights.
Knowledge Database: Farfetch maintains a robust knowledge database containing consolidated learnings from all teams across the company. We leveraged this resource to optimize our projects and also contributed significantly by adding our insights from working with various brands.
Predictive Eye Tracking: this tool provided insights into users' visual attention, identifying which elements on a web page attracted the most attention. It helped evaluate if the website had a clear visual hierarchy and ensured that users' attention flowed through essential page elements. AI-generated heatmaps also revealed any distractions from call-to-action buttons (CTAs).

Case study Emilio Pucci

Emilio Pucci is one of the most influential brands in fashion history. Known for its lightweight fabrics and iconic prints, the brand became synonymous with carefree elegance, attracting high-profile clients such as Jacqueline Kennedy and Marilyn Monroe. In 2000, Pucci was acquired by LVMH (Moët Hennessy Louis Vuitton), a luxury conglomerate that manages some of the world's most prestigious brands, including Louis Vuitton, Dior, Fendi, and Off-White.

We worked with Pucci for nearly a year, focusing on improving the key performance indicators of their e-commerce website. Our team conducted a comprehensive, 360-degree analysis of the store, mapping user pain points, usage patterns, and performance metrics to propose targeted business solutions.

Below are the key insights collected during our analysis of the Pucci website:

Product Listing Page (PLP): issues and pain points

Product Listing Pages (PLPs) display products based on selected categories or applied search filters. A well-designed PLP strikes a balance between showcasing products and providing a user-friendly interface that allows customers to easily narrow down their search. On Pucci's website, several issues with the filters were identified:

Filters are hard to find: Most users had trouble locating the filters, as they were hidden under a small dropdown labeled “Filter.” This made it difficult for them to refine their search.
Filter options don't update as users click: During user testing, we found that users struggled to narrow down product results because the filter options weren't dynamically updated as they made selections. For example, when a user filtered by an attribute like color, the other filter options (such as size or material) didn't adjust to show only the available choices based on that color selection. Dynamic Filters would solve this issue by ensuring that once a user selects a filter, the remaining options automatically update to reflect only the available choices. This approach helps users see relevant filtering options as they navigate, making it easier to refine their search and find what they're looking for.
Filters missing key product attributes: Users noted the absence of important product attributes in the filters that are essential for decision-making. During testing, they mentioned attributes like clothing style, sleeve type, and length as essential information to support their purchase decisions.

Product Listing Page (PLP): improvements

Make Filters Dynamic: dynamic filters ensure that once a user selects a filter attribute, the remaining options automatically update to reflect only the available choices. This approach helps users see relevant filtering options as they browse, making it easier to refine their search and find exactly what they're looking for. Additionally, dynamic filters significantly reduce the chances of showing empty results, as users can now clearly see the remaining available options as they apply different filters.
Improve filter visibility and usability: to implement dynamic filters, the filter display structure had to be redesigned. We separated different filter attributes into individual sections, allowing the system to update other attributes in the background when a filter is applied. This not only made the filtering process smoother and more intuitive but also improved visibility, as all filter attributes are now clearly displayed on the page for users to select from.
Add essential product attributes to filters: we made all key product attributes available for filtering. These attributes vary depending on the product type, such as different filters for shoes versus dresses, ensuring users can easily find products based on the most relevant characteristics.

Achievements of the improved Product Listing Pages:

After analysing and implementing the improvements, we achieved the following results:

Increased Click-Through Rate (CTR) by 9.68%
Decreased bounce rate by 41.45%
Decreased percentage of Exits by 15.88%

Product Detail Page (PDP): issues and pain points

A Product Detail Page (PDP) must provide all the essential information for customers to make an informed purchasing decision. It should give users everything they need to know about a product, eliminating the need to visit a physical store. This includes details like the product's appearance, materials, fit, delivery costs, and return policies. A well-designed PDP builds confidence, increasing conversion rates and reducing returns. Below are the main findings from the analysis of Pucci's e-commerce site:

Size Guide is Difficult to find: user testing revealed that customers struggled to locate the size guide. This resource is crucial in fashion eCommerce as it helps users determine the right size and fit. A clear and comprehensive size guide should explain how to measure correctly and select the right size, boosting purchase confidence and reducing returns.
Sizing system is not specified: many users reported that the sizing unit was unclear, leading to hesitation in completing the purchase. Sizing systems differ between countries and product types (e.g., Italian vs. American sizes). Clearly stating the sizing system helps users choose the correct size with confidence.
Insufficient information about how the piece fits the body: users noted a lack of details regarding how the product would fit their body. Beyond the sizing system, additional information such as whether the fit is tight or loose, body measurements, and fabric details help customers understand how the item will look and feel on different body types.

Product Detail Page (PDP): improvements

Specify sizing system and improve size guide visibility: we ensured that the sizing system for each product was clearly displayed on the website. The size guide link was made more prominent with improved affordance, making it easier for users to locate and access.
Provide relevant fit information: fit information is one of the most critical aspects of a product detail page. We ensured that essential details — such as whether the fit is tight or loose, body measurements, and fabric specifics — were clearly available to help customers understand how the item will fit and feel on different body types. To achieve this, we not only made the information available in the database but also revised how products were photographed and documented before being added to the website. Every product was reviewed to ensure all necessary data was included.
Product detail tab opened by default: based on insights from A/B testing conducted with other Farfetch brands, we made the product detail tab open by default when the page loads. This change increased “Add to Bag” rates and boosted user interaction with all three product detail tabs.
Button "Add to Bag" always enabled: another valuable insight from the Farfetch knowledge base was to keep the “Add to Bag” button enabled, even when the size has not yet been selected. This adjustment increased the likelihood of users adding items to their shopping bag.

Achievements of the improved Product Detail Pages: