Lola’s Cupcakes
Design Sprint🧁
5-day design sprint targeting product page drop-off. Redesigned the customisation flow to reduce decision fatigue and increase conversion confidence.
Context: Design sprint workshop, Allwomen Academy
My Role
Sprint participant
UX/UI Designer
Tools
Figma, FigJam, Adobe Illustrator, Photoshop, Notion, Loom
Timeline
5 days
Team Members
Dulinda Marques
Zelal Şahin
Aitana Barredo Valero
Lyna Santos
Introduction
This was my first design sprint, five days of structured collaboration to tackle a real business challenge for Lola's Cupcakes.
We were a group of eight, split into two teams of four. After a quick warm-up and short introductions revealing our diverse backgrounds, we immediately jumped into the sprint.
CHALLENGE 🎯
We were asked to improve the online shopping experience of Lola’s Cupcakes to boost conversion rates. The main issue was a large drop-off on the product page, so our goal was to uncover pain points and design a solution.
THE FEELING 🦋
The atmosphere was full of enthusiasm, curiosity, and determination, with everyone eager to dive in. From the start, the sprint moved at a fast, structured pace that kept us fully focused.
Understand 📊🧐
The sprint began with immersing ourselves in the problem. We were given a detailed brief, survey results, and interview data, and each of us wrote personal observations on sticky notes.
The surveys revealed patterns in user behavior and highlighted pain points. The interviews added depth, letting us empathize with users’ frustrations and motivations.
We also noted our own experiences as users. Each person simulated a specific “expert” perspective: Design, Marketing, Business, or Tech. I took the Designer Expert role, focusing on who the users are, their goals, and where the experience breaks down.
👉 Most users are young adults (25–34) buying cakes as gifts. Some seek quick purchase options, others prefer personalized experiences. On Lola’s website, categories were confusing, customization was limited, and the information structure contributed to drop-offs.
We presented our findings in lightning talks, shared notes, and clustered similar ideas into themes. From there, we collaboratively set a long-term goal to align the team around a shared direction.
Next, we defined a user persona. The user is looking for a cake purchase suitable for a professional environment, with specific dietary requests and a specific delivery date. They wants to give a good impression, values an easy ordering process and wants to meet colleagues expectations.
Mapping their user journey helped identify moments of frustration, emotional highs and lows, unmet needs, confusing tasks, and opportunities for delight.
Definition Phase 💬💼
With empathy built and the problem space mapped, we moved into definition. We turned insights into How Might We (HMW) statements and voted for one:
👉How might we ensure users feel reassured and excited that the cake customization options will align with their expectations?🧁🍰
We chose this HMW because survey data and interviews revealed users felt overwhelmed by cake options and uncertain about customization. Tackling this had the greatest potential to improve the experience and reduce drop-offs.
Next, we ran Lightning Demos, researching competitors and other industries for inspiration. Sharing our top findings showed how successful patterns could be adapted to make Lola’s experience more reassuring and enjoyable.
-
Domino’s Pizza offers a real-time customization preview,
-
Crumbs&Doilies provides a clear step-by-step flow
-
Sprinkles.com guides users seamlessly from browsing to purchase.
Sketch Phase ✍️✏️
With inspiration fresh, we focused on quantity over quality, quickly drawing rough sketches and writing concepts. Once we had a pool of ideas, we selected the stronger ones and moved into Crazy 8s, iterating variations of one idea.
We then expanded our favorite idea into a 1–3 frame storyboard, showing how it would work in practice.
Decide Phase 🤝✨
After sketching, we reviewed all concepts together. Each person explained their idea, and we discussed what worked and what didn’t. Using sticker voting, everyone placed two votes on sketches or parts they found most promising, explaining their choices briefly. The ideas with the most votes were carried on in the next stage.
Prototype Phase 🛠️📐
In this phase, our goal was to turn the concept into a testable prototype, focusing only on the key steps needed to validate our idea. We created a storyboard and task flow, outlining step by step what the user would see and do, then transformed it in Figma into an interactive prototype.
We prioritized redesigning the product page and customization flow, since that’s where most users dropped off.
Task flow: customize your cake and purchase it 🧁
.png)
Storyboard
Finally stepping into Figma, the key design decisions we implemented were:
Step-by-step process to make customization clear and enjoyable | Renders for transparency | Address + delivery date requested upfront to confirm availability early | Allergy info + ingredients visible, with customer photos for reference |Instant price updates based on selections | Dynamic real-time preview reflecting the exact custom combination | Pre-set popular designs to inspire and speed up decisions | Feedback section with photos for transparency and building trust
Validate ✅🔑
The last step was testing our prototype with three users. We observed how they navigated the flows, asked them to think aloud, and noted their reactions, iterating the design after.
Key learnings included:
-
Users loved seeing changes to their cake in real time - it gave them confidence in their order.
-
Simplified filters reduced decision fatigue.
-
The step-by-step process, guided by a top navigation bar, helped users feel supported throughout personalization.
-
Some microcopy still needed minor adjustments to reduce confusion
Conclusions📌💡
This sprint was a whirlwind of learning, collaboration and creativity. Highlights included:
-
The most impactful solution was restructuring the existing customisation flow to reduce choices at each step
-
Testing with just three users was enough to validate the direction and catch copy issues we'd missed
-
Working within a tight constraint (five days, four people, one problem) forces sharper prioritisation
-
Small usability fixes create relevant business impact