Carvel is an 85-year old ice cream franchise that has over 200 locations providing premium soft-serve ice cream and ice cream cakes. Carvel's cake sales account for a significant amount of its revenue but its online cake sales have been declining over the years. Carvel currently uses a third-party platform for its online cake-ordering experience and the Carvel team suspects that the complexity of navigating the platform has been contributing to the decline of its online cake sales. Our team worked directly with Carvel and its parent company Focus Brands to analyze the issues with the current online ordering platform and create a new experience that aligns with their brand mission of bringing simple joy to people.
Duration: August - December 2019
Team: Avinandan Basu, Kenneth Huang, Monica Jeon, Chongling Zhou
1. Coordinated with the industry partner and was responsible for scheduling meetings and providing deliverables.
2. Actively participated in all user research and analysis, ideation and testing.
3. Created the pre-made cake ordering flow from wireframes to high fidelity mockups.
Client: Focus Brands
Course: Research Methods for Human-Computer Interaction at Georgia Tech
Problem Statement: A user who wants to order a cake online and feels overwhelmed by the complex online ordering system needs a more efficient way to navigate and order a custom cake.
A modified store-locator flow for users to view available pre-made cakes at their nearest store.
A simple customization flow divided into three main stages: size & shape, inside, outside. The visibility of system status is clear, providing users with visual information of their cake building process.
To gain a holistic understanding of the problem space and understand user needs, our team created a research plan that incorporates the following methods. Tap image to view details.
Before delving into our research, we performed task analysis to a) understand user goals by breaking down each step involved in completing the task of ordering a cake from Carvel’s cake-ordering system, b) identify which subtasks users are struggling the most in completing, and c) generate preliminary ideas to tackle usability issues.
By comparing and contrasting the strengths and weaknesses of the platforms of Carvel's competitors, we gained insights on the general landscape in which Carvel's cake-ordering system competed. We identified primary competitors and gauged Carvel's presence in the market.
Some direct competitors include: Baskin Robbins, Dairy Queen, Ben and Jerry's, Cold Stone Creamery and these competitors were evaluated based on information delivery (information architecture, navigation and interaction), appearance and learning (tools and suggestions that aid user's decision-making process.)
Performing heuristic evaluation gave us specific examples of the usability issues that current users might be experiencing. We used Jakob Nielsen's 10 General Principles for Interaction Design as our guidelines and categorized each usability issue into one of the ten heuristics with a severity rating (minor, major, catastrophe). We also defined whether each issue is related to Navigation, Content, Presentation or Interaction of the website.
After we compiled our data we created graphs to visualize these issues and figure out where most of them lie on the platform. The graphs showed that a significant portion of the usability problems are from Heuristics 1 (Visibility of system status) and 7 (Flexibility and efficiency of use), many of which were minor usability issues that are simply irritant to users. However, there were serious issues that could potentially result in user drop-offs. Most catastrophic issues dealt with the text-heavy customization form and general navigation around the website, which we will be addressing in the ideation phase. These results suggest that the new system should be visually clear about the user’s process and their current status in the process, and help them navigate the platform efficiently.
Before conducting user interviews we decided to visit local bakeries and grocery stores (Kroger and Walmart bakeries) and talk to the managers and bakers for expert interviews for a high level understanding of user behaviors. Since they have experience in dealing with a lot of custom cake orders they were able to provide us insights on the general cake ordering process with their own online ordering platform, user preferences in customization options, and communication methods.
Then we conducted six semi-structured user interviews with Georgia Tech students who have ordered custom cakes before to understand specific needs and wants during the cake-ordering process. We had a list of questions about their experiences in ordering cakes online (such as their attitudes towards online cake ordering, possible pain points, and particular features that they enjoyed) and also had them go through Carvel's current cake ordering experience. We were able to gather insights on what users might want in an online cake ordering platform. Some quotes from our user interviews and field studies that revealed recurring themes:
Based on the interviews with students and experts, we created affinity notes and turned them into an affinity diagram to gain a better understanding of user needs. We first started with yellow sticky notes to capture details from the interviews and grouped them into blue sticky notes that have design relevance, which then are grouped into pink sticky notes, which reveal key issues in our data. Through this bottom-up process we were able to find the underlying implication of our data and we wanted to make sure we addressed these higher level issues in the new system. Some of the key issues captured in the pink sticky notes about design implications are:
1. Users are frustrated with the disconnect between the Carvel website and third-party ordering platform.
2. The new system should clearly show the visualization of the final product.
3. Users want sufficient information (price, reviews, options, ingredients) before ordering a cake.
4. Users want simple customization options (small inscriptions and decorations) but more control over these options (such as choosing the right shade of color for frosting).
5. Users need a more efficient way to communicate their errors.
6. Users want a more efficient way to order on mobile.
Based on our collected data, we analyzed our findings and created personas and journey maps to put ourselves in the users' shoes and understand their needs, wants and emotions to guide us through the ideation phase.
We conducted a focus group session to collect general feedback on different design alternatives for the navigation styles, pre-made cake order flows and store locator. Through this session, we received very constructive feedback on how we can improve the designs. Some main findings include:
1) Users prefer the accordion navigation style but they want to see the final visualization of the cake better.
2) Users want to see the pricing right away.
3) Users don't want to see unavailable cakes.
4) Users want tips/recommendations embedded in the process.
After our focus group sessions, we had an expert review with one of our stakeholders who is a UX Specialist at Focus Brands. We wanted to share our findings from the focus group session and also gain his opinions on the design alternatives which would determine our direction for the next round of iteration. We discovered that there were a few points that might be in conflict with user needs such as showing prices before a store is selected because different stores might have different prices. Some main findings from this review session include:
1) Prices can only be shown after a store is selected.
2) The cake visualization in the accordion style needs to be more prominent.
3) There should be an option to see all cakes provided by Carvel.
For the final iteration, I closely followed Carvel's style guide to keep our designs consistent with Carvel's visual language. Some of the assets for the customization flow had to be created for the cake visualization feature which is one of the crucial features of the new design.