Still struggling to choose the right mattress at IKEA? Too many terms to understand, no time to visit the store, and not in the mood to talk to sales?
Still struggling to choose the right mattress at IKEA? Too many terms to understand, no time to visit the store, and not in the mood to talk to sales?
Here’s an IKEA AI Assistant can help you select perfect mattress !
Here’s an IKEA AI Assistant can help you select perfect mattress !
TEAM
3 AI team Designers + Senior Designers from other teams
3 AI team Designers + Senior Designers from other teams
4+ Developers
4+ Developers
1 Product Manager
1 Product Manager
2 Product Owners
2 Product Owners
TIMELINE
3 months
3 months
CONTEXT
Customers usually have problems in decision making during shopping process due to lack of valuable information and have problems in finding the products they need.
Customers usually have problems in decision making during shopping process due to lack of valuable information and have problems in finding the products they need.
The small mini program goal is to develop an AI shopping buddy that functions as a "friend" to customers, understanding their needs and assisting them in making purchase decisions, thereby bridging IKEA brand expertise with customer requirements.
The small mini program goal is to develop an AI shopping buddy that functions as a "friend" to customers, understanding their needs and assisting them in making purchase decisions, thereby bridging IKEA brand expertise with customer requirements.
Visual Design / Hi-fi Prototype
I balanced Scandinavian and Chinese design styles while maintaining IKEA’s brand and addressing local users’ needs. I also contributed to IP design and Poster design, ensuring consistency with the overall user experience.
User Flow
I quickly finished the current UX/UI and visual design of the current landing page with senior designer so the co-workers can use AI toolbox immediately in their work.
Usability Testing & Iteration
I designed the interview questions, organized usability testing results in Miro, and finished final Usability Testing report with a Senior Designer.
Visual Design / Hi-fi Prototype
I balanced Scandinavian and Chinese design styles while maintaining IKEA’s brand and addressing local users’ needs. I also contributed to IP design and Poster design, ensuring consistency with the overall user experience.
User Flow
I quickly finished the current UX/UI and visual design of the current landing page with senior designer so the co-workers can use AI toolbox immediately in their work.
I finished the user flow exploration, and prioritized final design to provide smooth experience for users while shopping.
Usability Testing & Iteration
I designed the interview questions, organized usability testing results in Miro, and finished final Usability Testing report with a Senior Designer.
Final Design
Final Design
Homepage
Homepage
Homepage
Product Pip Reply
Product Pip Reply
Product Pip Reply
Product Comparison
Product Comparison
Product Comparison
View User Comments
View User Comments
View User Comments
Scan Label to view product details
Scan Label to view product details
User Journey Exploration based on Local Shopping Behavior
User Journey Exploration based on Local Shopping Behavior
I contributed to designing the homepage and exploring possible user flows with two other designers based on different user scenarios. This involved conducting a SWOT analysis and helping to create a seamless journey from initial queries to mattress selection.
I contributed to designing the homepage and exploring possible user flows with two other designers based on different user scenarios. This involved conducting a SWOT analysis and helping to create a seamless journey from initial queries to mattress selection.
Balance Business and User’s Needs
Balance Business and User’s Needs
We chose the final version because the AI shopping buddy assists rather than replaces in-store staff recommendations. We also aligned the design with common chatbot UX patterns for a smoother user experience. Removing scanned products from the homepage helped streamline the user flow, focusing on guiding users through the selection process without overwhelming them with too many options upfront.
We chose the final version because the AI shopping buddy assists rather than replaces in-store staff recommendations. We also aligned the design with common chatbot UX patterns for a smoother user experience. Removing scanned products from the homepage helped streamline the user flow, focusing on guiding users through the selection process without overwhelming them with too many options upfront.
Maintain Brand Identity through Visual Design
Design System
Utilized the IKEA SKAPA Design System and China Subsystem to maintain brand identity.
Homepage Design
The other designers and I explored different design styles. I created the final homepage design, balancing Scandinavian and Chinese aesthetics while maintaining IKEA’s brand identity and addressing local users’ needs.
IP Design
I explored over 20 versions and successfully combined IKEA’s classic imagery (the IKEA man) with a robotic element, expressive facial expressions to emphasize the chatbot’s identity, and make users think it’s a “friend”.
Visual Design
Maintain Brand Identity through Visual Design
Maintain Brand Identity through Visual Design
Design System
Utilized the IKEA SKAPA Design System and China Subsystem to maintain brand identity.
Homepage Design
The other designers and I explored different design styles. I created the final homepage design, balancing Scandinavian and Chinese aesthetics while maintaining IKEA’s brand identity and addressing local users’ needs.
IP Design
I explored over 20 versions and successfully combined IKEA’s classic imagery (the IKEA man) with a robotic element, expressive facial expressions to emphasize the chatbot’s identity, and make users think it’s a “friend”.
Poster Design
I designed the final marketing poster and various promotional slogans, which were successfully displayed in mattress sections across stores, attracting numerous customers to try the mini program.
Poster Design
I designed the final marketing poster and various promotional slogans, which were successfully displayed in mattress sections across stores, attracting numerous customers to try the mini program.
Usability Testing
Usability Testing
For the beta-testing version of the Shopping Buddy, I designed the interview questions, while other designers and the PM conducted interviews with one co-worker and nine groups of IKEA customers from different age groups and occupations. I then synthesized the interview findings, extracted key insights, and compiled them into a comprehensive report including the user journey map.
For the beta-testing version of the Shopping Buddy, I designed the interview questions, while other designers and the PM conducted interviews with one co-worker and nine groups of IKEA customers from different age groups and occupations. I then synthesized the interview findings, extracted key insights, and compiled them into a comprehensive report including the user journey map.
Read Full Report
Read Full Report
Learnings & Insights
Learnings & Insights
In-store scenario relies strongly on physical interaction🤝
In-store scenario relies strongly on physical interaction🤝
The AI Shopping Buddy helps customers explore and understand mattresses, simplifying early choices. However, it can't assist with in-store mattress location, making staff support necessary. For in-store mode, a physical guide is essential, while its strength lies in online use without location constraints.
The AI Shopping Buddy helps customers explore and understand mattresses, simplifying early choices. However, it can't assist with in-store mattress location, making staff support necessary. For in-store mode, a physical guide is essential, while its strength lies in online use without location constraints.
Opportunity in non-staff zones and pre-store journey 💐
Opportunity in non-staff zones and pre-store journey 💐
Majority of customers have the need to seek answers to their queries while in-store, suggesting a potential use of AI in non-staff areas to fulfill this demand. Additionally, customers enter stores with unclear shopping goals, offering a chance to employ AI for inspiration and guidance in the pre-store journey.
Majority of customers have the need to seek answers to their queries while in-store, suggesting a potential use of AI in non-staff areas to fulfill this demand. Additionally, customers enter stores with unclear shopping goals, offering a chance to employ AI for inspiration and guidance in the pre-store journey.
More effort in digitalization accessing 📱
More effort in digitalization accessing 📱
The low uptake of digital tools highlights the need for improved accessibility and tailored education to match customer behaviors (photoshoot, measuring) during in-store digitalization progress.
The low uptake of digital tools highlights the need for improved accessibility and tailored education to match customer behaviors (photoshoot, measuring) during in-store digitalization progress.
People trust human experience over machine output👩🏻💻
People trust human experience over machine output👩🏻💻
Customers see AI as impersonal and purely informational, trusting store staff more for their real-world experience. To bridge this gap, AI should prioritize human-like interactions and experience-based knowledge over mere intelligence.
Customers see AI as impersonal and purely informational, trusting store staff more for their real-world experience. To bridge this gap, AI should prioritize human-like interactions and experience-based knowledge over mere intelligence.
Possible Future Improvements
Possible Future Improvements
Possible Future Improvements
Next Steps
Next Steps
Improve AI shopping buddy for in-store usage based on recommendations above.
Improve AI shopping buddy for in-store usage based on recommendations above.
Improve AI shopping buddy for in-store usage based on recommendations above.
Test AI shopping buddy in App or SMP to verify its effectiveness in online journey without physical or location constraints
Test AI shopping buddy in App or SMP to verify its effectiveness in online journey without physical or location constraints
Test AI shopping buddy in App or SMP to verify its effectiveness in online journey without physical or location constraints
Use AI to covert pictures from customer phone to a shopping list in store companion
Use AI to covert pictures from customer phone to a shopping list in store companion
Use AI to covert pictures from customer phone to a shopping list in store companion
Extend knowledge base into more ranges to prep for all-ranges serves.
Extend knowledge base into more ranges to prep for all-ranges serves.
Extend knowledge base into more ranges to prep for all-ranges serves.
Personal Takeaways
Personal Takeaways
UX/UI Localization
UX/UI Localization
This is my first time designing a 2C app that contains Chinese characters, which differs from an English layout. Also, due to cultural differences, local users prefer different app styles. Therefore, it is important to consider local user needs while maintaining IKEA's brand identity.
This is my first time designing a 2C app that contains Chinese characters, which differs from an English layout. Also, due to cultural differences, local users prefer different app styles. Therefore, it is important to consider local user needs while maintaining IKEA's brand identity.
Emotional Design
Emotional Design
Although it's an AI chatbot, we want it to feel like a friend to customers. The IP design emphasizes a warm, joyful, and friendly character. In UX writing, it incorporates tone, phrasing, and thought patterns that make users feel more connected.
Although it's an AI chatbot, we want it to feel like a friend to customers. The IP design emphasizes a warm, joyful, and friendly character. In UX writing, it incorporates tone, phrasing, and thought patterns that make users feel more connected.
Feasibility Matters
Feasibility Matters
Due to AI technology limitations and developer scheduling constraints, we had to make trade-offs between features, which taught us how to better determine user needs' priority and ensure the best user experience within limited resources.
Due to AI technology limitations and developer scheduling constraints, we had to make trade-offs between features, which taught us how to better determine user needs' priority and ensure the best user experience within limited resources.