E-commerce Redesign
user research | information architecture | taxonomy
Overview
Background
RepairClinic is an online store and repair help site that sells over 3,000,000+ replacement parts and maintenance products for over 200 brands. They pride themselves on educating both novice and experienced DIYers in fixing their own appliances through their online troubleshooting videos, step by step guides and repair help content. With over 1600 online videos they are one of the largest e-commerce parts distributers to deliver both product and DIY content to the public.
Originate was tasked with researching, designing, and engineering a new front-end user experience for RC’s existing e-commerce platform. We worked alongside various external vendors and partners, such as taxonomists, SEO strategists & RC’s web analytics team to help inform the product & content strategy, information architecture, taxonomy, and layout of the new site.
My role
This engagement spanned on and off for almost a year and was one of the first projects I worked on as a UX Designer. I was on a design team of three consisting of one Senior Visual Designer and one Senior UX Designer. Contracting with other vendors, I also served as the design liaison working directly with specialists in taxonomy and analytics, assisting in research activities and utilizing findings to form UX insights and recommendations for the new site. Over the course of the project my role grew and by the end I was the lead product designer delivering high-fidelity designs to our engineering team.
Some of my key responsibilities and activities for the research phase included —
Conducting stakeholder interviews, user interviews and contextual inquiries
Managing research synthesis with design team (affinity, system & concept mapping)
Developing user personas and goals, journey maps and insight reports
Taxonomy Research and IA definition (search query analysis, tree testing, card-sorting, site map)
Creating user flows, wireframes and prototypes
Presenting and updating insights to the client, product and engineering teams
the challenge
Defining the problem(s) and identifying business goals
Stakeholder interviews were key to understand processes, business goals, customer types and the history of the site’s functionality (which is 16 years old by the way). We gained valuable insight by speaking to people from customer service, marketing, IT, and the purchasing department. Pairing qualitative research with CLM trends, search queries, and a heuristic analysis, we identified some of the key problems facing Repair Clinic—
Search is the powerhouse of the site but it is homegrown,
it does not reflect a proper shopping experience.
“I search left navigation more than I use the search box. The search box doesn’t work….They call it search suggest. I call it search exact.”
“We are like the wild west. If you start (searching) wrong it can go crazy. If you start right (with the model number) it is smooth.”
Customers need help finding the right parts to repair their appliance. Business Goal: RC needed a powerful and intuitive search function and navigation system that would prevent confusion for users and drive confidence in diagnosing a problem and finding the right part.
Repair parts is not adequately linked to repair help
“Our website is silo-ed between repair help and parts for repair. The only connection is the model number... There have to be multiple ways to arrive at what part they need without the model number.”
A desire for a lot of people is to find parts by certain characteristics, but currently, “everything starts with the model number”
Customers need a site with clear and consistent navigation. An organization scheme that better links product to repair help will allow customers to browse, find products and recall content more efficiently. Business Goal: Designing an experience that connects these two flows while allowing flexibility in the search and diagnose process for different skill types would be key to improving user experience and conversion rates.
Better content that is consistent and consumer
friendly is needed for users to search on
“Two things desired by customers: 1) Manufacturer manuals for parts posted to website, and 2) Part diagrams posted to website. These things would bring them to the website and possibly keep them on the website.”
“They ‘tech’ things up rather than make them consumer friendly... Why do they call it ‘garbage disposer’ on the website? These vocabularies were created by ‘repair gurus.’”
RC’s main competitive advantage is it’s content. With an extensive library of repair help with various content types— videos, guides, schematics, manuals etc they are one of the most diverse DIY sites to merge instruction with product. However, due to some of the usability issues facing RC, they are losing customers. Business Goal: Selling parts is a niche business that will require RC to leverage user data and their library of content to differentiate themselves among the competition.
Defining success
RC defined success at the conclusion of the MVP engagement as —
Creation of a ‘best in industry’ website, customer experience, and brand identity
30+% YOY sales growth driven by increases in traffic and conversion
Creation of a foundation for further innovation to evolve RC’s business model against it’s long term strategic goals.
discovery
Leveraging data to understand the online journey
Access to CLM and search query data helped us gain insight into the online behavior of users and helped us answer two key questions we had following initial stakeholder interviews.
What brings people to RC? What’s the experience once they get there?
Below are screenshots from a presentation the product team put together. Working with the pm, it was my job to formulate UX insights that would be relevant moving forward. Being conscious of the limitations of quantitative data, we used these findings to validate some of our assumptions, narrow down on high risk areas and form our objectives for user interviews.
We also had access to organic queries captured by GA to understand how users reached repairclinic.com. We were mostly interested in what combinations of words and concern indicators were most frequent. This provided clues into users’ mental models when starting the journey to diagnose, purchase and repair.
Key Takeaways
RC’s current search experience relied heavily on users knowing their model number at the start of the DIY journey
Data showed that 86% of queries indicated a symptom about a product
Thinking in terms of concerns is the primary mental model when starting to look for a solution
Most site interactions start with global nav “Shop for Parts”, only 9% utilize site search
91% of users landing on a repair help page drop off and 68% of users landing on a shop for parts page 68% drop off
Focus
Understanding user needs & goals
From quant research we had an overview of what the online journey is like but we still needed to better understand why customers drop off and how various skill types diagnose and shop. Moving into qualitative research we wanted to answer the following questions:
User interviews paired with quant trends gave us and the client a robust understanding of the current customer journey and ideas for the proposed. We used these artifacts as tools during the product roadmap workshop to define user stories, scenarios and flows.
Hypothesis: If we allow multiple, flexible points of entry for troubleshooting, while focusing on connecting symptoms to parts on both repair help and product detail pages, users would feel more confident in finding the right solution and purchasing the right product.
information architecture
Evaluating current & proposed taxonomy
The persona types and use cases developed out of primary research were used to help guide the following concept models and system maps to form the new taxonomy and IA structure. To further validate these recommendations, we conducted online tree testing and card sorting with screened users on Optimal Workshop.
The new taxonomy of the site would separate concerns & symptoms from solution types and introduce product type, brands and part titles as attributes to part and product categories.
Defining content and functionality
We used the new taxonomy design and semantic models to identify the different content pieces and see how we could improve the relationships between them. Various maps and models were created that helped define the organization of the new site and determine the potential navigation components. We worked closely with our engineers to better understand the transition to elastic search, the options for filtering systems, and the capabilities of search suggestions with the proposed taxonomy. We explored creating a navigation system more similar to industry leading e-commerce sites while brainstorming a solution to link concerns & solution content to the overall experience.
System mapping exercise on realtime board
Proposed Site Map
design studio
Sketching & wireframing
Working with the product team we brainstormed various concepts to better link product with repair help. We focused on Unskilled and Skilled DIYers since their user goals and journeys differed most. One of the major difficulties throughout this project was getting timely access to content, technical requirements, and API documentation. This was a huge learning process that cost us a few iterations of flows and brought us back to the drawing board more than once. Below are a few detailed wireframe flows we put together to discuss content dependencies with our client.
Usability walkthroughs
Utilizing recruitment.io and validately we were able to quickly recruit, screen, and test our concepts on both skilled and unskilled DIYers. Below are examples of quick digests I put together to show our clients. To keep the internal team up to speed I set up a slack channel for all things research and google file stream to house our spreadsheets.
