Lo - An AI Agent For Lowes
Designing a Conversational AI
for DIY Home Improvement
Project Details
Overview
Role
UX Research, Interaction Design, Voice User Interface Design, Prototype Development, User Journey Mapping,
& Motion Design
Timeline
Sep-Oct 2023
Team
Audrey Reiley
Jun Yang Hao
Shivani Kannan
Grace Zhu
My team and I developed Lo, a conversational AI agent system that empowers non-expert users to confidently navigate tool and supply selection and their uses for home improvement projects. Our solution addresses the entire DIY workflow—from initial problem identification to project completion—leveraging voice interaction to deliver personalized, hands-free assistance across mobile, tablet, and in-store kiosk platforms.
The Challenge
Through our research, we discovered that DIY enthusiasts face significant challenges throughout their home improvement journey. Many visitors to home improvement stores struggle with problem diagnosis, finding suitable products, navigating large retail spaces, and troubleshooting issues during project execution.
A recurring phenomenon we identified as "DIY overwhelm" represents customers' cognitive overload when confronting complex project requirements and extensive product catalogs. While Lowe's provides comprehensive product selection and expert staff, customers often struggle to access timely assistance and translate technical information into actionable guidance for their specific situations.
Traditional retail experiences create passive browsing rather than active problem-solving, with customers typically lacking the confidence and framework to tackle home improvement projects independently. These gaps risk turning customer visits into transactions rather than meaningful learning experiences that build lasting relationships and DIY expertise.
Hands-Free Experience
Customers often have dirty hands or are juggling tools and materials, making voice interaction ideal for the DIY context.
Why an AI Agent for Lowes?
Our research revealed four key opportunities for conversational user interfaces in the home improvement context:
Accessible for Vision Challenges
Many customers struggle with small text on product labels or technical specifications, especially in warehouse lighting conditions.
Personalized Assistance
Each DIY project is unique, requiring tailored guidance based on skill level, budget, and specific circumstances.
Complex product specifications and installation instructions can be translated into conversational, easy-to-understand guidance.
Technical Information Simplified
Research
Secondary Research
Onsite Observations & Interviews
Conducted systematic observations across different departments, documenting customer behavior patterns, pain points, and interaction needs.
Web Articles
Conducted background research on Lowe's business model and
the broader DIY culture in the United States through online articles
and industry reports.
App Audit
Conducted a comprehensive audit of the existing Lowe's mobile application to map user journeys and identify improvement opportunities.
Conducted 12 guerrilla interviews with customers and staff, plus 4 in-depth interviews with experienced DIYers to understand motivations, challenges, and decision-making processes.
Interviews
We began by analyzing DIY culture in America, discovering that 47% of Americans consider themselves somewhat handy, while only 18% consider themselves extremely handy. This gap indicated a significant opportunity to support intermediate DIYers who want to tackle projects but need guidance and confidence-building.
Primary Research
Key Research Findings
Through our research synthesis, we identified six critical pain points in the DIY journey:
-
Customers need experience to identify root causes and aren't sure if projects match their skill level.
-
Customers want to know if required products are in stock before making a trip to the store.
-
Large retail spaces create difficulty finding specific products and frequent walking back and forth.
-
Getting timely assistance from knowledgeable staff, especially during peak hours.
-
Finding products best suited for specific needs among overwhelming options and technical specifications.
-
Diagnosing and resolving issues that arise during the project execution phase.
Personas
Primary Users
Chris, the DIYer
Chris is an intermediate DIY enthusiast who seeks budget-friendly solutions but struggles with project planning and troubleshooting. He represents confident DIYers who need guidance without condescension.
Amy, the professional
Amy is an experienced interior designer who needs quality products for clients but faces availability issues, budget constraints, and time-consuming research. She represents professionals requiring efficient, detailed product information.
1
User Journey Mapping
We mapped the complete DIY experience across three key phases:
Pre Store
Problem Identification
-
Need experience to identify the root cause
-
Uncertain if project matches skill level
1
Pre Store
Problem Identification
-
Need experience to identify the root cause
-
Uncertain if project matches skill level
Action planning
-
Product availability uncertainty
-
Lack of clarity on necessary tools
2
In Store
Store Navigation
-
Difficulty finding products
-
Frequent walking back and forth
2
In Store
Store Navigation
-
Difficulty finding products
-
Frequent walking back and forth
Product Comparison
-
Overwhelming text and small print
-
Unsure which product is suitable
3
Post Store
Planning & Kickstart
-
Underestimating time
-
Mixing up project steps
-
Information overload
3
Post Store
Planning & Kickstart
-
Underestimating time
-
Mixing up project steps
-
Information overload
Troubleshooting
-
Struggling to diagnose when things go wrong
-
Need for adaptive problem-solving support
Designing an AI Personal Assistant
Based on our research findings, we developed "Lo" - a conversational AI assistant designed specifically for Lowe's customers. Lo embodies the helpful, inspiring, and community-centric values of the Lowe's brand while providing personalized guidance throughout the DIY journey.
Al Characteristics
Character Traits
-
Reliable and trustworthy
-
Encouraging and optimistic
-
Intelligent but approachable
-
Collaborative and adaptive
Tone and Language
-
Casual yet informative
-
Direct and personable
-
Optimistic and educational
Purpose
-
Educate customers about products and techniques
-
Inform with relevant, timely information
-
Enable successful project completion
Interaction States & Behaviors
Design Language System
Visual Identity
Color Palette
Lo's visual representation uses Lowe's established color palette and typography while introducing a distinctive helper icon that suggests both human assistance and technological capability.
Typography
Inter family (Light, Regular, Bold) for consistency with Lowe's brand standards while ensuring readability across devices.
Clean, minimal icons that support voice interaction without overwhelming the interface, following Lowe's established visual language.
Iconography
Final Outcomes
Our solution addresses the need for retail home improvement stores like Lowe's who would benefit from carefully designed conversational voice interactions that prioritize human needs while leveraging technology as an enabling tool rather than the primary focus. Lo employs augmented reality and conversational AI to guide customers through problem diagnosis, product selection, and project execution. The system combines machine learning with real-time inventory data to create personalized shopping experiences, enabling customers to navigate complex product decisions with confidence. This approach sets our work apart from typical retail technology implementations. Instead of adding digital layers that distance customers from products, our system promotes deeper engagement with the physical materials and tools central to DIY success.
Next Steps & Further Development
Given more time and resources, we identified several key areas for continued development and enhancement:
-
Social functionality allowing DIYers to share projects, ask questions, and learn from each other's experiences. This could include project galleries, peer mentoring systems, and collaborative problem-solving forums integrated directly into the Lo experience.
-
Develop algorithms that anticipate home maintenance needs based on customer purchase history, local weather patterns, and typical wear cycles where Lo could proactively suggest preventive projects and seasonal preparations.
-
Help customers understand the environmental impact of their choices, suggest eco-friendly alternatives, and track resource conservation across projects.