Cookidoo Add a feature project
The goal was to design an interface that supports users in a hands-on environment, where attention is divided and interactions need to be simple, precise, and reliable.
Role: UX Researcher - UX/UI Designer
Scope: Interaction design, feature design, user flows, wireframes
Timeline: 4 weeks
Thermomix is a multifunctional smart cooker that combines heating, blending, and weighing into a single device. It works closely with the Cookidoo® app, which provides guided cooking through step-by-step recipes and smart features that help users cook with confidence.
Because cooking is a hands-on activity where attention is divided, the interaction between the device and the digital interface needs to be simple, clear, and reliable.
In this project, I explored how the Cookidoo experience could be improved by designing an additional feature for the app, with the goal of making the cooking process easier to follow and more intuitive in real-time.
Problem Statement 1: Limited Recipe Variety & Customization
Users often experience the recipe library as repetitive, with limited everyday options.
There is little flexibility to adapt recipes or create and share personal variations, which makes the experience feel restrictive over time.
Problem Statement 2: Language and ingredient limitations
The platform is often limited to a single language, and recipes are not easily translated.
At the same time, there is no clear way to adapt ingredients or find substitutes, making it difficult for international or bilingual users to cook flexibly.
Project Goals
The goal was to make Cookidoo more flexible, usable, and relevant for everyday cooking.
User goals
-
Find and plan meals more easily, especially for everyday cooking
-
Access a wider variety of recipes and cuisines
-
Adapt recipes to personal preferences or available ingredients
-
Follow recipes clearly and confidently without confusion
-
Use the platform flexibly across languages and contexts
-
Feel in control of the cooking process through simple, guided steps
Product goals
-
Increase daily use by making the platform more useful for everyday cooking
-
Improve perceived value of the subscription through better content and features
-
Strengthen Cookidoo as a central hub for recipes and cooking workflows
-
Support international users through multilingual and flexible content
-
Encourage long-term engagement and loyalty
UX Research
Background
Problem
The current Cookidoo experience limits flexibility and personalization in everyday cooking.
Research Goal
I want to: Understand how users discover and interact with the recipe-management experience and identify improvements that help retain subscribers beyond the free trial period.
Research Plan - Add a feature
-
Define research personas
-
Conduct user interviews
-
Generate an affinity map
-
Analyze result
-
Determine problem statement & Project scope
-
Create low-fidelity wireframes and run initial test
-
Iterate to mid-fidelity wireframes
-
Testing & Prototype
-
Analyze results
-
High-fidelity wireframes prototype
-
Last testing & Iterations
Meet Miguel & Sofia
Research Personas
.png)
.png)
Miguel and Sofia represent Cookidoo users. Users who like to try international dishes and elaborate recipes use their Thermomix once or twice a week, and users who like simple everyday recipes make their cooking experience fast and easy.
*Photos generated with AI
Affinity map Results
All participants participating in the interview own a Thermomix; they all have different habits with their Thermomix. However, they have common frustrations regarding the Cookidoo frustration they experience.
-
Full speed features are scary to use.
-
Multi-language search and translation are too rigid.
-
There's no remote control for the TM
-
Instructions can be confusing.
Information Architecture
User Flows
The primary tasks developed were: “Edit an existing recipe” and “Create a new recipe.” Research shows these are the flows users need most.
To support flexibility and clarity, I restructured the content into a clearer and more scalable information architecture.
.png)
.png)
UI Design
Low-Fidelity Wireframes
The Low-Fidelity wireframes are a first draft of how the user/task flows will integrate into the existing app.
The Cookidoo app has a complex layout to find the different recipes. One can access a specific recipe through the main page, the collections, the search bar, or through the app suggestions.
For this case study - Modify an existing recipe- we approach the recipe through the search bar.
For the flow - Create a new recipe- the path is through "My Recipes."
The brand identity and coherent layout are respected, and the final product should look like a natural insert to the app.
_edited.png)
Edit a recipe Thermomix Cookidoo
_edited.png)
Create a recipe Thermomix Cookidoo
Mid-Fidelity Wireframes
This is an example of the discrete modified design of the current app.
The inserts feel natural and almost invisible.
The design of the Mid-fidelity wireframes follows the organic paths that users showed in the low-fidelity testing results.
For this case study, the necessary screens were designed. Such as:
Home
Search
My recipes
Recipe main page
Editing menu input
Creating menu input
Saved message pop-up screen
Iterations
Iterations resulting from mid-fidelity testing focused on reducing the amount of information users needed to complete in order to continue editing.
Before
4 Screens Step
.png)
.png)
After
1 Screen Step
.png)
.png)
.png)
Key Takeaways
.png)
A major challenge during the interviews was reconciling contradictory feedback from participants, making it difficult to identify which insights were most relevant to inform design decisions.
To resolve this, feedback was synthesized by identifying recurring patterns and validating them against usability observations rather than relying on individual opinions.
.png)
I initiated this case study after identifying a language barrier in Cookidoo that limited users’ access to recipes. While exploring automatic AI translation as a solution, a new pain point emerged: users lacked intuitive ways to manipulate and customize recipes.
An unexpected insight from user research revealed concerns I had not initially anticipated. For example, several participants found using the TM at full speed intimidating—highlighting a gap between system capability and user confidence.
A key trade-off in this case study involved simplifying the number of adjustable recipe parameters. Users were overwhelmed by details that were not critical to their goals, many of which could be handled automatically through AI-driven assistance.
AI played a central role in organizing and enhancing the experience by grouping similar inputs, reducing cognitive load, and generating supportive visuals where appropriate.
These decisions ensured customization remained flexible while reducing complexity for less confident users.
.png)
.png)