UX RESEARCH / CASE STUDY
Nego
Most people decide what to cook by scrolling. FoodFeed is a recipe app built that way on purpose.
CONTEXT
M.A. Thesis
UX/UI Interaction Design
SCOPE
14 screens
research → hi-fi prototype
METHODS
Interviews, SWOT
heuristic, competitive
ROLE
Designer & builder
solo, with Claude Code
THE PROCESS
01
Problem
Dinner decisions stall on the same few meals, over and over.
02
Research
Interviews, SWOT, heuristic and competitive analysis.
03
Insight
People already use social media to try new dishes.
04
Design
14 screens, built only from what users asked for.
05
Outcome
A thesis product rooted in research, not assumption.
01 / THE PROBLEM
Everyone eats. Almost no one knows what to make.
I found a real problem and design for it. I started with my own — cooking the same four meals on rotation — then went to check whether it was only mine.
02 / THE INSIGHT
Sitting with the data long enough to know what was valid.
METHOD 03
Heuristic evaluation
Usability gaps in the sites people already use daily.
METHOD 04
SWOT analysis
Where a new product could realistically win.
METHOD 01
User interviews
Ages 21–45, on how they really decide what's for dinner.
METHOD 02
Competitive analysis
Where existing recipe products stop short of the decision.
03/ WHAT I FOUND
Three findings that decided the product.
01
Social media is already the recipe box
Takes developers' data directly — no new format to learn.
02
Health was the hidden driver
Crisp at any size, fully editable — design styles, not re-enters data.
03
Restrictions block trying anything new
Content imports data; design exports and finishes. No overlap.
04 / THE PRODUCT
A feed of food — filtered by who you are.
What you see is what you get. Onboarding learns your constraints once, then the feed does the deciding for you.
Allergies & avoids, kept separate. Allergies raise a flag on any recipe that contains them. Dislikes just quietly drop out of the feed.
Step-by-step, optional
Confident cooks skim ingredients; newer cooks follow steps. Both paths exist
The feed. Two-card recipe rows under Trending / For You, with swipeable meal categories for users who want to get in and out fast.
Intake in view
Nutrition surfaced on the recipe, so eating well doesn't need a second app.
Dietary preferences. Faith-based diets, low sodium, low sugar — captured as identity, not a filter buried in settings.
Image-forward by request
Users judge a dish visually first. Photography leads every card.
05 / OUTCOME
14
✓
✓
✓
screens, every one traceable
to something a user said
A behavior-led concept
Built on how people already discover food instead of a new habit they'd have to learn.
Restrictions treated as first-class
Allergies, diets and dislikes handled in onboarding — so exploring feels safe, not risky.
A research practice I still use
Defensible design decisions, documented — the habit that carried into client and product work.
REFLECTION
This is where I learned to work structurally with ambiguity. Not to guess faster — to sit with the data until the real problem showed itself.
Most design work moves fast and agile. FoodFeed was the opposite, and it taught me what a project gains when it's genuinely rooted in research and documented well enough to defend.