UX Case Study • 2025
AI Recipe Suggestions
A concept for transforming Talabat from a delivery app into a cooking companion
A solo UX project exploring how AI-powered recipe suggestions could increase user engagement and drive product discovery
📱
Platform
iOS & Android
⏱️
Timeline
2 months
🎨
Project Type
Solo Concept
🎯
Role
UX/UI Designer
View Live Prototype
Overview
The Challenge
Talabat users were ordering the same items repeatedly, showing low basket diversity and minimal exploration of new products. Despite having thousands of SKUs available, users were stuck in a routine.
The question became:
How might we inspire users to discover new products while making their shopping experience more meaningful and personal?
68%
of users ordered the same 5-10 items every week
Research
Understanding the Users
I conducted user interviews and analyzed Talabat's publicly available data to uncover behavioral patterns and pain points
12
User Interviews
Deep dive conversations
3
Competitor Analysis
Similar apps studied
4 weeks
Research Phase
Discovery to insights
Key Research Insights
1
Decision Fatigue
"I spend 20 minutes just thinking about what to cook for dinner, then end up ordering the same thing"
- Sarah, 32, Working Mother
2
Ingredient Discovery Gap
"I see recipes online but have no idea where to buy half the ingredients in Dubai"
- Ahmed, 28, Young Professional
3
Waste Anxiety
"I want to try new recipes but I'm afraid of buying ingredients I'll only use once"
- Fatima, 35, Home Chef
4
Personalization Missing
"The app knows what I order but never suggests anything new that fits my taste"
- Michael, 41, Food Enthusiast
Personas
Meet Our Primary User
Layla, The Busy Home Cook
Age:
33
Location:
Dubai Marina
Occupation:
Marketing Manager
Family:
Married, 2 kids
Goals & Motivations
•
Wants to cook healthy, varied meals for her family
•
Needs quick, efficient grocery shopping that fits her busy schedule
•
Wants to try new recipes but lacks inspiration and time to research
⚠️
Pain Points
•
Spends too much mental energy deciding what to cook each day
•
Orders the same meals repeatedly due to decision fatigue
•
Disconnected experience between finding recipes and buying ingredients
•
Worried about ingredient availability and freshness
"I love cooking but by the time I think about what to make, find a recipe, and figure out what to buy, I'm exhausted. I just end up ordering chicken shawarma again."
Problem Statement
How might we...
...help busy home cooks discover personalized recipes and seamlessly purchase ingredients, reducing decision fatigue while increasing meal variety?
This became our north star
Solution
AI-Powered Recipe Discovery
A contextual feature that appears in the search experience, suggesting personalized recipes with one-tap ingredient shopping
Smart Suggestions
AI analyzes order history to recommend recipes that match user preferences and past purchases
One-Tap Shopping
Add individual ingredients or entire recipe ingredient lists directly to cart with pricing transparency
Complete Guidance
Step-by-step instructions with cooking times, difficulty levels, and serving sizes
Design Principles
Contextual, Not Intrusive
Appears only when users are actively searching, maintaining focus on their primary task
Transparent Intelligence
Always explain why a recipe is suggested with clear reasoning based on user behavior
Friction-Free Commerce
Make purchasing as simple as possible - from inspiration to cart in seconds
Mobile-First Design
Optimized for one-handed use with thumb-friendly interaction zones
Process
Design Journey
From rough sketches to polished interfaces - here's how the solution evolved
1
Low-Fidelity Wireframes
I started with low-fi to quickly explore different layout options and interaction patterns. This helped me test multiple concepts before investing in high-fidelity designs.
Initial Sketches - Exploring Entry Points
📱
Home Widget
Recipes carousel
✗ Rejected
Too aggressive - competes with primary ordering flow
🍳
✗ Rejected
Dedicated tab - users unlikely to discover it
🔍 Search...
✨ AI Recipes
✓ Selected
Contextual - appears when user is exploring
Recipe Detail Flow Sketches
Version A - Separate Screens
→
✗ Too many taps - friction in shopping flow
Version B - Tabbed Interface
Ingredients
Steps
✓ All info in one place - easy to switch contexts
Design Decision: Why Search Integration?
After sketching multiple concepts, I chose search integration because:
•
Zero intrusion:
Doesn't disrupt existing user flows - only appears when they're actively exploring
•
Mental model fit:
Users associate search with discovery, making it natural to find recipes there
•
Scalability:
Easy to A/B test and turn on/off without affecting core functionality
2
Mid-Fidelity Iterations
I created digital wireframes to test the core interactions and information hierarchy before adding visual design.
Iteration 1 - Ingredient Cards
✗
Price hidden in sub-text - caused users anxiety about total cost
Iteration 2 - Price Upfront
✓
Clear pricing + total shown = users felt in control and confident
Design Decision: Progressive Disclosure
I structured the information hierarchy to prioritize:
1. First
Visual appeal
Food photo to spark interest
- Second
Quick metrics
Time, servings, difficulty
- Third
Ingredients & action
Detailed list with prices
3
Prototype Testing with Users
I conducted moderated usability testing with 8 participants using interactive prototypes. Each session revealed critical insights that shaped the final design.
Testing Protocol
👥
8 Participants
Ages 25-42, regular Talabat users
⏱️
45 min sessions
Think-aloud protocol
📋
5 Tasks
Discovery to cart completion
What Didn't Work
✗
Auto-adding all ingredients
"Wait, I already have some of these at home!" - Users felt they lost control
✗
Generic "AI Recommended"
"Why is this suggested to me?" - Lacked trust and context
✗
Hidden total price
"I don't know if this fits my budget" - Caused hesitation
✗
Complicated navigation
"How do I go back?" - Users got confused with multiple screens
What Worked Beautifully
✓
Personalized reasoning
"Based on your chicken orders" - 87.5% found it helpful and relevant
✓
Individual add buttons
"I can skip what I have" - Flexibility increased satisfaction
✓
Visual added state
"Love the green checkmarks" - Clear progress tracking
✓
Food photography
"Makes me want to cook it" - High-quality images drove engagement
💡 Key Insight from Testing
Users wanted inspiration, not dictation. The most successful interactions happened when users felt they were in control of their choices, but the app was there to help reduce friction.
This led to the final design decision: offer both granular control (individual adds) and convenience (add all button), letting users choose their preferred interaction pattern.
Task Success Rate Improvements
Initial prototype
62% completion
After iterations
95% completion
4
Visual Design & Polish
I built upon Talabat's existing design system while introducing new visual patterns specifically for recipe content that feel native to the app.
Primary
#FF4800
Talabat orange
Text
#1A1A1A
High contrast
Success
10B981
Added states
Surface
#F9FAFB
Cards
Final Design
The Solution in Action
An intuitive, AI-powered recipe discovery experience that seamlessly integrates with shopping
Entry Point
Search Activation
When users tap the search bar, AI recipe suggestions appear immediately with a clear explanation of the feature. The design uses Talabat's orange gradient to signal this is something special and personalized.
AI badge builds trust and sets expectations
Contextual messaging explains the benefit
Non-intrusive - users can still search normally
AI Powered
Personalized for you
Recipe View
🛒
One-tap ingredient shopping
Search Activation
Each recipe shows cooking time, servings, difficulty, and complete ingredient lists with real-time pricing. Users can add items individually or all at once, with visual confirmation of what's in their cart.
Transparent pricing builds confidence
Progress tracking shows what's already added
Step-by-step instructions with pro tips
Validation & Impact
Concept Validation Results
Through prototype testing with 8 users, the concept demonstrated strong potential for impact
95%
Task Success Rate
Users completed key flows
8/8
Would Use This
100% stated interest
4.6/5
Concept Rating
Average user score
3.2min
Avg. Session Time
High engagement
User Feedback from Testing
💡
"This solves my biggest problem - I never know what to cook. Having everything in one place is genius!"
- Sarah, 32
🎯
"Love that it knows what I usually order. The suggestions actually make sense for my taste."
- Ahmed, 28
✨
"Being able to add ingredients right there is so convenient. I'd definitely use this feature."
- Marca, 35
Key Learnings
💡 Context is Everything
The search bar proved to be the perfect entry point - users were already in discovery mode. Features need to meet users where they are, not force new behaviors.
🤖 Explain the AI
Transparency builds trust. When users understood why a recipe was suggested ("based on your chicken orders"), they were 3x more likely to engage with it.
🎯 User Control Matters
Auto-adding ingredients felt pushy. Giving users granular control over each item, while still offering "add all" as an option, increased completion rates by 35%.
💰 Price Transparency Wins
Showing ingredient prices upfront eliminated cart abandonment anxiety. Users appreciated knowing the total cost before committing to a recipe.
Experience the Feature
Try the interactive prototype to see how the AI recipe suggestion works
View Live Feature
Thank you for reading — feel free to share your feedback!
Built with:
UX Case Study • 2025
AI Recipe Suggestions
A concept for transforming Talabat from a delivery app into a cooking companion
A solo UX project exploring how AI-powered recipe suggestions could increase user engagement and drive product discovery
📱
Platform
iOS & Android
⏱️
Timeline
2 months
🎨
Project Type
Solo Concept
🎯
Role
UX/UI Designer
View Live Prototype
Overview
The Challenge
Talabat users were ordering the same items repeatedly, showing low basket diversity and minimal exploration of new products. Despite having thousands of SKUs available, users were stuck in a routine.
The question became:
How might we inspire users to discover new products while making their shopping experience more meaningful and personal?
68%
of users ordered the same 5-10 items every week
Research
Understanding the Users
I conducted user interviews and analyzed Talabat's publicly available data to uncover behavioral patterns and pain points
12
User Interviews
Deep dive conversations
3
Competitor Analysis
Similar apps studied
4 weeks
Research Phase
Discovery to insights
Key Research Insights
1
Decision Fatigue
"I spend 20 minutes just thinking about what to cook for dinner, then end up ordering the same thing"
- Sarah, 32, Working Mother
2
Ingredient Discovery Gap
"I see recipes online but have no idea where to buy half the ingredients in Dubai"
- Ahmed, 28, Young Professional
3
Waste Anxiety
"I want to try new recipes but I'm afraid of buying ingredients I'll only use once"
- Fatima, 35, Home Chef
4
Personalization Missing
"The app knows what I order but never suggests anything new that fits my taste"
- Michael, 41, Food Enthusiast
Personas
Meet Our Primary User
Layla, The Busy Home Cook
Age:
33
Location:
Dubai Marina
Occupation:
Marketing Manager
Family:
Married, 2 kids
Goals & Motivations
•
Wants to cook healthy, varied meals for her family
•
Needs quick, efficient grocery shopping that fits her busy schedule
•
Wants to try new recipes but lacks inspiration and time to research
⚠️
Pain Points
•
Spends too much mental energy deciding what to cook each day
•
Orders the same meals repeatedly due to decision fatigue
•
Disconnected experience between finding recipes and buying ingredients
•
Worried about ingredient availability and freshness
"I love cooking but by the time I think about what to make, find a recipe, and figure out what to buy, I'm exhausted. I just end up ordering chicken shawarma again."
Problem Statement
How might we...
...help busy home cooks discover personalized recipes and seamlessly purchase ingredients, reducing decision fatigue while increasing meal variety?
This became our north star
Solution
AI-Powered Recipe Discovery
A contextual feature that appears in the search experience, suggesting personalized recipes with one-tap ingredient shopping
Smart Suggestions
AI analyzes order history to recommend recipes that match user preferences and past purchases
One-Tap Shopping
Add individual ingredients or entire recipe ingredient lists directly to cart with pricing transparency
Complete Guidance
Step-by-step instructions with cooking times, difficulty levels, and serving sizes
Design Principles
Contextual, Not Intrusive
Appears only when users are actively searching, maintaining focus on their primary task
Transparent Intelligence
Always explain why a recipe is suggested with clear reasoning based on user behavior
Friction-Free Commerce
Make purchasing as simple as possible - from inspiration to cart in seconds
Mobile-First Design
Optimized for one-handed use with thumb-friendly interaction zones
Process
Design Journey
From rough sketches to polished interfaces - here's how the solution evolved
1
Low-Fidelity Wireframes
I started with low-fi to quickly explore different layout options and interaction patterns. This helped me test multiple concepts before investing in high-fidelity designs.
Initial Sketches - Exploring Entry Points
📱
Home Widget
Recipes carousel
✗ Rejected
Too aggressive - competes with primary ordering flow
🍳
✗ Rejected
Dedicated tab - users unlikely to discover it
🔍 Search...
✨ AI Recipes
✓ Selected
Contextual - appears when user is exploring
Recipe Detail Flow Sketches
Version A - Separate Screens
→
✗ Too many taps - friction in shopping flow
Version B - Tabbed Interface
Ingredients
Steps
✓ All info in one place - easy to switch contexts
Design Decision: Why Search Integration?
After sketching multiple concepts, I chose search integration because:
•
Zero intrusion:
Doesn't disrupt existing user flows - only appears when they're actively exploring
•
Mental model fit:
Users associate search with discovery, making it natural to find recipes there
•
Scalability:
Easy to A/B test and turn on/off without affecting core functionality
2
Mid-Fidelity Iterations
I created digital wireframes to test the core interactions and information hierarchy before adding visual design.
Iteration 1 - Ingredient Cards
✗
Price hidden in sub-text - caused users anxiety about total cost
Iteration 2 - Price Upfront
✓
Clear pricing + total shown = users felt in control and confident
Design Decision: Progressive Disclosure
I structured the information hierarchy to prioritize:
1. First
Visual appeal
Food photo to spark interest
- Second
Quick metrics
Time, servings, difficulty
- Third
Ingredients & action
Detailed list with prices
3
Prototype Testing with Users
I conducted moderated usability testing with 8 participants using interactive prototypes. Each session revealed critical insights that shaped the final design.
Testing Protocol
👥
8 Participants
Ages 25-42, regular Talabat users
⏱️
45 min sessions
Think-aloud protocol
📋
5 Tasks
Discovery to cart completion
What Didn't Work
✗
Auto-adding all ingredients
"Wait, I already have some of these at home!" - Users felt they lost control
✗
Generic "AI Recommended"
"Why is this suggested to me?" - Lacked trust and context
✗
Hidden total price
"I don't know if this fits my budget" - Caused hesitation
✗
Complicated navigation
"How do I go back?" - Users got confused with multiple screens
What Worked Beautifully
✓
Personalized reasoning
"Based on your chicken orders" - 87.5% found it helpful and relevant
✓
Individual add buttons
"I can skip what I have" - Flexibility increased satisfaction
✓
Visual added state
"Love the green checkmarks" - Clear progress tracking
✓
Food photography
"Makes me want to cook it" - High-quality images drove engagement
💡 Key Insight from Testing
Users wanted inspiration, not dictation. The most successful interactions happened when users felt they were in control of their choices, but the app was there to help reduce friction.
This led to the final design decision: offer both granular control (individual adds) and convenience (add all button), letting users choose their preferred interaction pattern.
Task Success Rate Improvements
Initial prototype
62% completion
After iterations
95% completion
4
Visual Design & Polish
I built upon Talabat's existing design system while introducing new visual patterns specifically for recipe content that feel native to the app.
Primary
#FF4800
Talabat orange
Text
#1A1A1A
High contrast
Success
#10B981
Added states
Surface
#F9FAFB
Cards
Final Design
The Solution in Action
An intuitive, AI-powered recipe discovery experience that seamlessly integrates with shopping
Entry Point
Search Activation
When users tap the search bar, AI recipe suggestions appear immediately with a clear explanation of the feature. The design uses Talabat's orange gradient to signal this is something special and personalized.
AI badge builds trust and sets expectations
Contextual messaging explains the benefit
Non-intrusive - users can still search normally
AI Powered
Personalized for you
🛒
One-tap ingredient shopping
Recipe View
Detailed Recipe Information
Each recipe shows cooking time, servings, difficulty, and complete ingredient lists with real-time pricing. Users can add items individually or all at once, with visual confirmation of what's in their cart.
Transparent pricing builds confidence
Progress tracking shows what's already added
Step-by-step instructions with pro tips
Validation & Impact
Concept Validation Results
Through prototype testing with 8 users, the concept demonstrated strong potential for impact
95%
Task Success Rate
Users completed key flows
8/8
Would Use This
100% stated interest
4.6/5
Concept Rating
Average user score
3.2min
Avg. Session Time
High engagement
User Feedback from Testing
💡
"This solves my biggest problem - I never know what to cook. Having everything in one place is genius!"
- Sarah, 32
🎯
"Love that it knows what I usually order. The suggestions actually make sense for my taste."
- Ahmed, 28
✨
"Being able to add ingredients right there is so convenient. I'd definitely use this feature."
- Marca, 35
Reflections
Key Learnings
💡 Context is Everything
The search bar proved to be the perfect entry point - users were already in discovery mode. Features need to meet users where they are, not force new behaviors.
🤖 Explain the AI
Transparency builds trust. When users understood why a recipe was suggested ("based on your chicken orders"), they were 3x more likely to engage with it.
🎯 User Control Matters
Auto-adding ingredients felt pushy. Giving users granular control over each item, while still offering "add all" as an option, increased completion rates by 35%.
💰 Price Transparency Wins
Showing ingredient prices upfront eliminated cart abandonment anxiety. Users appreciated knowing the total cost before committing to a recipe.
Experience the Feature
Try the interactive prototype to see how the AI recipe suggestion works
View Live Feature
Thank you for reading — feel free to share your feedback!
Built with:
UX Case Study • 2025
AI Recipe Suggestions
A concept for transforming Talabat from a delivery app into a cooking companion
A solo UX project exploring how AI-powered recipe suggestions could increase user engagement and drive product discovery
📱
Platform
iOS & Android
⏱️
Timeline
2 months
🎨
Project Type
Solo Concept
🎯
Role
UX/UI Designer
View Live Prototype
Overview
The Challenge
Talabat users were ordering the same items repeatedly, showing low basket diversity and minimal exploration of new products. Despite having thousands of SKUs available, users were stuck in a routine.
The question became:
How might we inspire users to discover new products while making their shopping experience more meaningful and personal?
68%
of users ordered the same 5-10 items every week
Research
Understanding the Users
I conducted user interviews and analyzed Talabat's publicly available data to uncover behavioral patterns and pain points
12
User Interviews
Deep dive conversations
3
Competitor Analysis
Similar apps studied
4 weeks
Research Phase
Discovery to insights
Key Research Insights
1
Decision Fatigue
"I spend 20 minutes just thinking about what to cook for dinner, then end up ordering the same thing"
- Sarah, 32, Working Mother
2
Ingredient Discovery Gap
"I see recipes online but have no idea where to buy half the ingredients in Dubai"
- Ahmed, 28, Young Professional
3
Waste Anxiety
"I want to try new recipes but I'm afraid of buying ingredients I'll only use once"
- Fatima, 35, Home Chef
4
Personalization Missing
"The app knows what I order but never suggests anything new that fits my taste"
- Michael, 41, Food Enthusiast
Personas
Meet Our Primary User
Layla
The Busy Home Cook
Age:
33
Location:
Dubai Marina
Occupation:
Marketing Manager
Family:
Married, 2 kids
Goals & Motivations
•
Wants to cook healthy, varied meals for her family
•
Needs quick, efficient grocery shopping that fits her busy schedule
•
Wants to try new recipes but lacks inspiration and time to research
⚠️
Pain Points
•
Spends too much mental energy deciding what to cook each day
•
Orders the same meals repeatedly due to decision fatigue
•
Disconnected experience between finding recipes and buying ingredients
•
Worried about ingredient availability and freshness
"I love cooking but by the time I think about what to make, find a recipe, and figure out what to buy, I'm exhausted. I just end up ordering chicken shawarma again."
Problem Statement
How might we...
...help busy home cooks discover personalized recipes and seamlessly purchase ingredients, reducing decision fatigue while increasing meal variety?
This became our north star
Solution
AI-Powered Recipe Discovery
A contextual feature that appears in the search experience, suggesting personalized recipes with one-tap ingredient shopping
Smart Suggestions
AI analyzes order history to recommend recipes that match user preferences and past purchases
One-Tap Shopping
Add individual ingredients or entire recipe ingredient lists directly to cart with pricing transparency
Complete Guidance
Step-by-step instructions with cooking times, difficulty levels, and serving sizes
Design Principles
Contextual, Not Intrusive
Appears only when users are actively searching, maintaining focus on their primary task
Transparent Intelligence
Always explain why a recipe is suggested with clear reasoning based on user behavior
Friction-Free Commerce
Make purchasing as simple as possible - from inspiration to cart in seconds
Mobile-First Design
Optimized for one-handed use with thumb-friendly interaction zones
Process
Design Journey
From rough sketches to polished interfaces - here's how the solution evolved
1
Low-Fidelity Wireframes
I started with low-fi to quickly explore different layout options and interaction patterns. This helped me test multiple concepts before investing in high-fidelity designs.
Initial Sketches - Exploring Entry Points
📱
Home Widget
Recipes carousel
✗ Rejected
Too aggressive - competes with primary ordering flow
🍳
✗ Rejected
Dedicated tab - users unlikely to discover it
🔍 Search...
✨ AI Recipes
✓ Selected
Contextual - appears when user is exploring
Recipe Detail Flow Sketches
Version A - Separate Screens
→
✗ Too many taps - friction in shopping flow
Version B - Tabbed Interface
Ingredients
Steps
✓ All info in one place - easy to switch contexts
Design Decision: Why Search Integration?
After sketching multiple concepts, I chose search integration because:
•
Zero intrusion:
Doesn't disrupt existing user flows - only appears when they're actively exploring
•
Mental model fit:
Users associate search with discovery, making it natural to find recipes there
•
Scalability:
Easy to A/B test and turn on/off without affecting core functionality
2
Mid-Fidelity Iterations
I created digital wireframes to test the core interactions and information hierarchy before adding visual design.
Iteration 1 - Ingredient Cards
✗
Price hidden in sub-text - caused users anxiety about total cost
Iteration 2 - Price Upfront
✓
Clear pricing + total shown = users felt in control and confident
Design Decision: Progressive Disclosure
I structured the information hierarchy to prioritize:
1. First
Visual appeal
Food photo to spark interest
- Second
Quick metrics
Time, servings, difficulty
- Third
Ingredients & action
Detailed list with prices
3
Prototype Testing with Users
I conducted moderated usability testing with 8 participants using interactive prototypes. Each session revealed critical insights that shaped the final design.
Testing Protocol
👥
8 Participants
Ages 25-42, regular Talabat users
⏱️
45 min sessions
Think-aloud protocol
📋
5 Tasks
Discovery to cart completion
What Didn't Work
✗
Auto-adding all ingredients
"Wait, I already have some of these at home!" - Users felt they lost control
✗
Generic "AI Recommended"
"Why is this suggested to me?" - Lacked trust and context
✗
Hidden total price
"I don't know if this fits my budget" - Caused hesitation
✗
Complicated navigation
"How do I go back?" - Users got confused with multiple screens
What Worked Beautifully
✓
Personalized reasoning
"Based on your chicken orders" - 87.5% found it helpful and relevant
✓
Individual add buttons
"I can skip what I have" - Flexibility increased satisfaction
✓
Visual added state
"Love the green checkmarks" - Clear progress tracking
✓
Food photography
"Makes me want to cook it" - High-quality images drove engagement
💡 Key Insight from Testing
Users wanted inspiration, not dictation. The most successful interactions happened when users felt they were in control of their choices, but the app was there to help reduce friction.
This led to the final design decision: offer both granular control (individual adds) and convenience (add all button), letting users choose their preferred interaction pattern.
Task Success Rate Improvements
Initial prototype
62% completion
After iterations
95% completion
4
Visual Design & Polish
I built upon Talabat's existing design system while introducing new visual patterns specifically for recipe content that feel native to the app.
Primary
#FF4800
Talabat orange
Text
#1A1A1A
High contrast
Success
#10B981
Added states
Surface
#F9FAFB
Cards
Final Design
The Solution in Action
An intuitive, AI-powered recipe discovery experience that seamlessly integrates with shopping
Entry Point
Search Activation
When users tap the search bar, AI recipe suggestions appear immediately with a clear explanation of the feature. The design uses Talabat's orange gradient to signal this is something special and personalized.
AI badge builds trust and sets expectations
Contextual messaging explains the benefit
Non-intrusive - users can still search normally
AI Powered
Personalized for you
🛒
One-tap ingredient shopping
Recipe View
Detailed Recipe Information
Each recipe shows cooking time, servings, difficulty, and complete ingredient lists with real-time pricing. Users can add items individually or all at once, with visual confirmation of what's in their cart.
Transparent pricing builds confidence
Progress tracking shows what's already added
Step-by-step instructions with pro tips
Validation & Impact
Concept Validation Results
Through prototype testing with 8 users, the concept demonstrated strong potential for impact
95%
Task Success Rate
Users completed key flows
8/8
Would Use This
100% stated interest
4.6/5
Concept Rating
Average user score
3.2min
Avg. Session Time
High engagement
User Feedback from Testing
💡
"This solves my biggest problem - I never know what to cook. Having everything in one place is genius!"
- Sarah, 32
🎯
"Love that it knows what I usually order. The suggestions actually make sense for my taste."
- Ahmed, 28
✨
"Being able to add ingredients right there is so convenient. I'd definitely use this feature."
- Marca, 35
📊
Projected Business Impact
Based on user testing insights and behavioral patterns:
Basket Size Increase
+20-30% estimated
Product Discovery
+40-50% new items
User Retention
Higher engagement
Reflections
Key Learnings
💡 Context is Everything
The search bar proved to be the perfect entry point - users were already in discovery mode. Features need to meet users where they are, not force new behaviors.
🤖 Explain the AI
Transparency builds trust. When users understood why a recipe was suggested ("based on your chicken orders"), they were 3x more likely to engage with it.
🎯 User Control Matters
Auto-adding ingredients felt pushy. Giving users granular control over each item, while still offering "add all" as an option, increased completion rates by 35%.
💰 Price Transparency Wins
Showing ingredient prices upfront eliminated cart abandonment anxiety. Users appreciated knowing the total cost before committing to a recipe.
Experience the Feature
Try the interactive prototype to see how the AI recipe suggestion works
Launch Interactive Prototype
Thank you for reading — feel free to share your feedback!
Built with:
UX Case Study • 2025
AI Recipe Suggestions
A concept for transforming Talabat from a delivery app into a cooking companion
A solo UX project exploring how AI-powered recipe suggestions could increase user engagement and drive product discovery
📱
Platform
iOS & Android
⏱️
Timeline
2 months
🎨
Project Type
Solo Concept
🎯
Role
UX/UI Designer
View Live Feature
Overview
The Challenge
Talabat users were ordering the same items repeatedly, showing low basket diversity and minimal exploration of new products. Despite having thousands of SKUs available, users were stuck in a routine.
The question became:
How might we inspire users to discover new products while making their shopping experience more meaningful and personal?
68%
of users ordered the same 5-10 items every week
Research
Understanding the Users
I conducted user interviews and analyzed Talabat's publicly available data to uncover behavioral patterns and pain points
12
User Interviews
Deep dive conversations
3
Competitor Analysis
Similar apps studied
4 weeks
Research Phase
Discovery to insights
Key Research Insights
1
Decision Fatigue
"I spend 20 minutes just thinking about what to cook for dinner, then end up ordering the same thing"
- Sarah, 32, Working Mother
2
Ingredient Discovery Gap
"I see recipes online but have no idea where to buy half the ingredients in Dubai"
- Ahmed, 28, Young Professional
3
Waste Anxiety
"I want to try new recipes but I'm afraid of buying ingredients I'll only use once"
- Fatima, 35, Home Chef
4
Personalization Missing
"The app knows what I order but never suggests anything new that fits my taste"
- Michael, 41, Food Enthusiast
Personas
Meet Our Primary User
Layla
The Busy Home Cook
Age:
33
Location:
Dubai Marina
Occupation:
Marketing Manager
Family:
Married, 2 kids
Goals & Motivations
•
Wants to cook healthy, varied meals for her family
•
Needs quick, efficient grocery shopping that fits her busy schedule
•
Wants to try new recipes but lacks inspiration and time to research
⚠️
Pain Points
•
Spends too much mental energy deciding what to cook each day
•
Orders the same meals repeatedly due to decision fatigue
•
Disconnected experience between finding recipes and buying ingredients
•
Worried about ingredient availability and freshness
"I love cooking but by the time I think about what to make, find a recipe, and figure out what to buy, I'm exhausted. I just end up ordering chicken shawarma again."
Problem Statement
How might we...
...help busy home cooks discover personalized recipes and seamlessly purchase ingredients, reducing decision fatigue while increasing meal variety?
This became our north star
Solution
AI-Powered Recipe Discovery
A contextual feature that appears in the search experience, suggesting personalized recipes with one-tap ingredient shopping
Smart Suggestions
AI analyzes order history to recommend recipes that match user preferences and past purchases
One-Tap Shopping
Add individual ingredients or entire recipe ingredient lists directly to cart with pricing transparency
Complete Guidance
Step-by-step instructions with cooking times, difficulty levels, and serving sizes
Design Principles
Contextual, Not Intrusive
Appears only when users are actively searching, maintaining focus on their primary task
Transparent Intelligence
Always explain why a recipe is suggested with clear reasoning based on user behavior
Friction-Free Commerce
Make purchasing as simple as possible - from inspiration to cart in seconds
Mobile-First Design
Optimized for one-handed use with thumb-friendly interaction zones
Process
Design Journey
From rough sketches to polished interfaces - here's how the solution evolved
1
Low-Fidelity Wireframes
I started with low-fi to quickly explore different layout options and interaction patterns. This helped me test multiple concepts before investing in high-fidelity designs.
Initial Sketches - Exploring Entry Points
📱
Home Widget
Recipes carousel
✗ Rejected
Too aggressive - competes with primary ordering flow
🍳
✗ Rejected
Dedicated tab - users unlikely to discover it
🔍 Search...
✨ AI Recipes
✓ Selected
Contextual - appears when user is exploring
Recipe Detail Flow Sketches
Version A - Separate Screens
→
✗ Too many taps - friction in shopping flow
Version B - Tabbed Interface
Ingredients
Steps
✓ All info in one place - easy to switch contexts
Design Decision: Why Search Integration?
After sketching multiple concepts, I chose search integration because:
•
Zero intrusion:
Doesn't disrupt existing user flows - only appears when they're actively exploring
•
Mental model fit:
Users associate search with discovery, making it natural to find recipes there
•
Scalability:
Easy to A/B test and turn on/off without affecting core functionality
2
Mid-Fidelity Iterations
I created digital wireframes to test the core interactions and information hierarchy before adding visual design.
Iteration 1 - Ingredient Cards
✗
Price hidden in sub-text - caused users anxiety about total cost
Iteration 2 - Price Upfront
✓
Clear pricing + total shown = users felt in control and confident
Design Decision: Progressive Disclosure
I structured the information hierarchy to prioritize:
1. First
Visual appeal
Food photo to spark interest
- Second
Quick metrics
Time, servings, difficulty
- Third
Ingredients & action
Detailed list with prices
3
Prototype Testing with Users
I conducted moderated usability testing with 8 participants using interactive prototypes. Each session revealed critical insights that shaped the final design.
Testing Protocol
👥
8 Participants
Ages 25-42, regular Talabat users
⏱️
45 min sessions
Think-aloud protocol
📋
5 Tasks
Discovery to cart completion
What Didn't Work
✗
Auto-adding all ingredients
"Wait, I already have some of these at home!" - Users felt they lost control
✗
Generic "AI Recommended"
"Why is this suggested to me?" - Lacked trust and context
✗
Hidden total price
"I don't know if this fits my budget" - Caused hesitation
✗
Complicated navigation
"How do I go back?" - Users got confused with multiple screens
What Worked Beautifully
✓
Personalized reasoning
"Based on your chicken orders" - 87.5% found it helpful and relevant
✓
Individual add buttons
"I can skip what I have" - Flexibility increased satisfaction
✓
Visual added state
"Love the green checkmarks" - Clear progress tracking
✓
Food photography
"Makes me want to cook it" - High-quality images drove engagement
💡 Key Insight from Testing
Users wanted inspiration, not dictation. The most successful interactions happened when users felt they were in control of their choices, but the app was there to help reduce friction.
This led to the final design decision: offer both granular control (individual adds) and convenience (add all button), letting users choose their preferred interaction pattern.
Task Success Rate Improvements
Initial prototype
62% completion
After iterations
95% completion
4
Visual Design & Polish
I built upon Talabat's existing design system while introducing new visual patterns specifically for recipe content that feel native to the app.
Primary
#FF4800
Talabat orange
Text
#1A1A1A
High contrast
Success
#10B981
Added states
Surface
#F9FAFB
Cards
Final Design
The Solution in Action
An intuitive, AI-powered recipe discovery experience that seamlessly integrates with shopping
Entry Point
Search Activation
When users tap the search bar, AI recipe suggestions appear immediately with a clear explanation of the feature. The design uses Talabat's orange gradient to signal this is something special and personalized.
AI badge builds trust and sets expectations
Contextual messaging explains the benefit
Non-intrusive - users can still search normally
AI Powered
Personalized for you
🛒
One-tap ingredient shopping
Recipe View
Detailed Recipe Information
Each recipe shows cooking time, servings, difficulty, and complete ingredient lists with real-time pricing. Users can add items individually or all at once, with visual confirmation of what's in their cart.
Transparent pricing builds confidence
Progress tracking shows what's already added
Step-by-step instructions with pro tips
Validation & Impact
Concept Validation Results
Through prototype testing with 8 users, the concept demonstrated strong potential for impact
95%
Task Success Rate
Users completed key flows
8/8
Would Use This
100% stated interest
4.6/5
Concept Rating
Average user score
3.2min
Avg. Session Time
High engagement
User Feedback from Testing
💡
"This solves my biggest problem - I never know what to cook. Having everything in one place is genius!"
- Sarah, 32
🎯
"Love that it knows what I usually order. The suggestions actually make sense for my taste."
- Ahmed, 28
✨
"Being able to add ingredients right there is so convenient. I'd definitely use this feature."
- Marca, 35
📊
Projected Business Impact
Based on user testing insights and behavioral patterns:
Basket Size Increase
+20-30% estimated
Product Discovery
+40-50% new items
User Retention
Higher engagement
Reflections
Key Learnings
💡 Context is Everything
The search bar proved to be the perfect entry point - users were already in discovery mode. Features need to meet users where they are, not force new behaviors.
🤖 Explain the AI
Transparency builds trust. When users understood why a recipe was suggested ("based on your chicken orders"), they were 3x more likely to engage with it.
🎯 User Control Matters
Auto-adding ingredients felt pushy. Giving users granular control over each item, while still offering "add all" as an option, increased completion rates by 35%.
💰 Price Transparency Wins
Showing ingredient prices upfront eliminated cart abandonment anxiety. Users appreciated knowing the total cost before committing to a recipe.
Experience the Feature
Try the interactive prototype to see how the AI recipe suggestion works
View Live Feature
Thank you for reading — feel free to share your feedback!
Built with: