
Multi-Device AI Assistant for Personalized Home Improvement Support
Design Timeline
1.5 Months (Sep'23 - Oct'23)
Highlights
Muse Award Gold Winner
Team Members
Audrey Riley, Grace Zhu, and Junyang Hao
My Role
Lead Product Designer: Secondary Research, Research Synthesis, Design System, UX, UI & Prototypes for Mobile and Smart Cart
Project Summary
01
Problem
How might we improve customer retention at Lowe's by facilitating the home improvement process for DIYers?
Solution
Lo, an AI-driven CUI offering end-to-end, personalized support from problem diagnosis to troubleshooting.

30% ↓
Customer Support Requests
20% ↑
User Retention
Anticipated Impact in Numbers
Context
02
Lowe's is Behind Home Depot in Market Share
A CUI would enable Lowe to provide efficient customer service in the face of staff shortages and give it a competitive advantage. It can use AI to provide personalized support for each user's unique home improvement query.

Sales of the leading home improvement chains in the United States in 2022. Statista 2024.
Why a CUI?

Hands-free
experience

Technical information
simplified

Personalized assistance

Accessible for visually impaired users
Research Methods to Derive to User Pain Points

15+ Web Articles & Papers
About Lowe’s & DIYing culture in the US

Lowe's App Audit
To identify user journeys & opportunities

12+ Interviews
With store employees, customers, and DIYers

Fly On The Wall Observation
To learn about the store layout & customer behavior
Research Findings
03
The Lowe's App Today has Limited Capabilities

Chat has limited capabilities and opens in a browser.

How-to guides exist but are text-heavy and difficult to follow.
Lowes' Main Users

Primary User
Chris, the DIYer
Chris, a DIY enthusiast with intermediate skills, seeks budget-friendly home improvement solutions.
Troubles: finding optimal solutions, underestimating timelines, and troubleshooting challenges.

Secondary User
Amy, the Professional
Amy, a highly skilled, experienced interior designer, seeks quality products to match client preferences.
Troubles: product availability, budget constraints, and time-consuming product research.
User Journey & Pain Points of the DIYer
Pre-Store
Identify Problem

Action Planning

Pain Points
-
Get a detailed understanding of the problem & how to fix it.
-
Uncertain about the skillset & time the project requires.
Solutions
-
Uses AR & AI to analyze the problem and lists various issues from most to least likely.
-
Provide difficulty level and ETA with a tutorial.
-
Suggest product with quantity/measurements.
In- Store
Store Navigation

Product Comparison

Pain Points
-
Exhausting to locate products.
-
Sift through countless products & get help to find a durable, in-budget one with the right specs.
Solutions
-
Directs user through the store according to their shopping list.
-
Compares multiple products, their purposes, and prices to recommend the best one for the user's specific DIY need and explains the choice.
Post- Store
Plan and Kickstart

Trouble Shooting

Pain Points
-
Mixing up steps.
-
Figuring out how, where, and why things go wrong when they go wrong.
Solutions
-
Follow-along tutorials for advanced DIY projects with the flexibility to skip steps.
-
AI-driven adaptive tutorial that analyzes the problem to troubleshoot and offers a solution.
Ideation
04
Fluid Storyboarding for Unrestricted Ideation
Fleshing Out Ideas with Impact-Effort Analysis




Testing Insights
05
'Smart Cart' for In-Store Use in Addition to Mobile
Findings from the 'Speed-dating' research with 6 designers prompted us to rethink the in-store CUI modality to enhance the effectiveness of the 'product comparison' feature.

Voice is avoided in public, making it difficult for elderly users with vision issues to read on phones.

Users must carry multiple products and find and walk to the kiosk to compare products.
.webp)
A smart cart allows integration to mobile and in-spot comparison with a barcode scanner.
Designing the User Interface (UI)
06
Logo Form - 'The Helper'
We took inspiration from the form of Lowe's staff jacket to create an icon that looks like a person in overalls.

+


Animations for the CUI Voice Interaction States
Keeping the Form Consistent yet Expressive to Depict the State

New Design System with 75+ Components in Under 1 Week
We maintained Lowe’s existing color palette for consistency across channels but adjusted the brightness and saturation for a fresher look. We redesigned the UI elements with softer corners for a more approachable feel while preserving a strong, knowledgeable aesthetic and introduced the consistent use of the 'Inter' font.

Tone of Voice of the Creator Archetype, Like Our User

Final Designs
07
Say Hel-Lo to the Future of DIY
Pre-Store
Instant Issue Diagnosis with AR and Computer Vision



Pre-Store
Personalized Action Plan with Step-by-Step Tutorials and Shopping Lists
In-Store
Avoid Backtracking with Seamless Aisle Navigation based on Your Shopping List


In-Store
Scan and Compare Products to Find the Perfect Fit for Your Needs
Post-Store
Step-by-Step Audio and Video Guides



Post-Store
AI-Powered Adaptive Tutorials for Effortless Troubleshooting
Impact
08
Anticipated Impact
30% ↓
In Customer Support Requests.
20% ↑
User Retention
Key Learnings
-
Conversation Design to inform UX: Thinking of UX as a conversation helps us find more organic solutions.
-
Motion Design: designing a form that accommodates various motion states and communicates the state (i.e., speaking/listening/error, etc.)
-
Solution-Oriented Mindset: I realized that in a situation of stress, it's key to focus on finding a solution rather than being stuck on the problem.
