Improving Search Results
Overview
The search function in any e-commerce website will determine whether the user is able to find the exact category or item they are looking for. Navigation is absolutely crucial when it comes to directing the user to the cart - and eventually checkout. This Case Study will showcase the effect of having an improved user flow through the search bar.
Problem Statement
The Lolë brand has been around for many years. Throughout those years, the library of products (some that have completely different product makeup, and some just slight variations) has increased dramatically.
In a situation when you have thousands of product SKUs and barcodes through many product lines, colours and sizes, it is essential to allow the customer to navigate the site in a way that is simplified. The user needs to be empowered.
Before this release, search autofill and recommendations were not enabled - limiting our users from narrowing the search to what they are looking for.
In a situation when you have thousands of product SKUs and barcodes through many product lines, colours and sizes, it is essential to allow the customer to navigate the site in a way that is simplified. The user needs to be empowered.
Before this release, search autofill and recommendations were not enabled - limiting our users from narrowing the search to what they are looking for.
The Goal
The Goal is to make suggestions and recommendations on categories and products based on user input. In this case, Auto-complete and recommendations are the early solutions from the search bar.
Users & Audience
In terms of User Search, the user demographic can be spliced in to two groups (regardless of conversion) - our repeat users interact with the search bar at a rate of about 42%, new users will use the search bar at a much lower rate of 14%. This could be attributed to how our product names are built and released on a yearly basis. If the user wants to find the newest release of the Agda Short-Sleeve, the repeat user will search up Agda on the bar. This would showcase the many available Agda Short-Sleeves from the current season and previous seasons.
Lole’s target demographic places the female shopper with disposable income at a range of 30-45 years old. A good portion of our users are repeat customers. Almost 80% of customer transactions come from users that have previously bought on the site. The rest are customers that are new and want to try the product.
Currently, the ability for us to capture new users for a return purchase is about 30% of the 20% traffic that we do convert.
Lole’s target demographic places the female shopper with disposable income at a range of 30-45 years old. A good portion of our users are repeat customers. Almost 80% of customer transactions come from users that have previously bought on the site. The rest are customers that are new and want to try the product.
Currently, the ability for us to capture new users for a return purchase is about 30% of the 20% traffic that we do convert.
Roles & Responsibilites
Lead Product Designer
Product Strategy, User Research & Analysis, User Interaction, Visual Design, Prototyping & Testing, Information Architecture
My role was the lead designer for this project. I was tasked revisiting the way our users interacted with our search bar. The mandate was to build an efficient user flow for the search function. I worked closely with the Head of Ecomm to set business requirements and build use cases (including edge cases). I had also worked with two front-end developers to ensure that the menu was pixel perfect and one back-end developer to ensure that the back-end references were accurate to the architecture of the product categories.
Product Strategy, User Research & Analysis, User Interaction, Visual Design, Prototyping & Testing, Information Architecture
My role was the lead designer for this project. I was tasked revisiting the way our users interacted with our search bar. The mandate was to build an efficient user flow for the search function. I worked closely with the Head of Ecomm to set business requirements and build use cases (including edge cases). I had also worked with two front-end developers to ensure that the menu was pixel perfect and one back-end developer to ensure that the back-end references were accurate to the architecture of the product categories.
Scope & Constraints
The project looks at the interaction from the search bar and the results that are recommended from the data architecture based on products (product description, colour and size). The main constraint is that there are overlapping tags/attributes within the data architecture so we have to be careful on how we set up recommendations based on product title or category.
Process
This design process spanned two weeks with the intent to have an aligned schedule with the upcoming sprint.
1. Understand
2. Analyze
3. Design
4. Stakeholder Validation
5. Developer Handoff
6. Gather Data and Iterate
1. Understand
2. Analyze
3. Design
4. Stakeholder Validation
5. Developer Handoff
6. Gather Data and Iterate
Outcomes
The results were outstanding. In terms of user experience and business results, this solution proved to be a no brainer and is something the brand will lean on moving forward. It looks like the feature allowed our users to view a lot more of our products based on their specific search. This effectively increased our pages per session by a whopping 2.3 pages. The effect on the visits on the Product Display Page (PDP) also proved to have a positive effect by increasing traffic on these pages by 18%.
Key Takeaways
In e-commerce, users arrive to the site looking to potentially spend their money (maybe not now, but maybe in the near future) - that’s the bottom line. If brands are wanting to capitalize on this idea, the notion of wanting an efficient user flow is key. If brands are able to showcase their products in a variety of ways (in this case being in the search bar), the faster the user decides if they will eventually purchase or be another bounce statistic.