Skip to content
L
lylehmann
·
AI Interface Conversational UX Retail

OBI Chatbot: AI-Driven Customer Empowerment

Designing a conversational interface to help DIY enthusiasts find exactly what they need for their home projects.

Conversational AI interface for OBI home improvement guidance.

Working with OBI, we designed a chatbot that acts as a virtual floor assistant — translating complex hardware inventory into a simple, helpful conversation that guides users through project planning and product selection.

The core insight driving every design decision: DIY users don’t search for products, they search for solutions to problems. The chatbot bridges that gap through progressive disclosure, contextual product cards, and a natural language interface that feels approachable for both novice DIYers and experienced tradespeople.

0% Faster
Search Performance
0% Conversion
Sales Growth
0 Languages
Market Reach
0M+ Users
Active Monthly

Core Features

Experience the power of intelligent conversational UI.

Problem-First Search

Users describe projects in natural language ("sink is leaking") and are guided to solutions.

Progressive Spec Disclosure

Technical specs are surfaced only when relevant, preventing DIYer cognitive overload.

Visual Product Cards

Rich cards with ratings and stock availability are embedded directly in the chat flow.

Expert Escalation

Seamless handoff to human advisors with full conversation context preserved.

The Process

From workshop alignment to a production-ready AI interface.

1

Stakeholder Alignment

Ran workshops with digital and customer service teams to identify legacy search failure patterns.

2

Conversation Architecture

Mapped 50+ user intents into a structured decision tree for optimized conversational flow.

3

Persona & Tone

Defined the "knowledgeable neighbour" personality to build trust and approachability.

4

Prototype & Test

Validated Figma prototypes with novice and expert users to refine information architecture.

5

API Integration

Designed the data contract between chat layers and real-time inventory systems.

6

A/B Rollout

Phased launch in the German market with rigorous KPI tracking and iteration.

Design Elements

Crafting a cohesive and trustworthy conversational environment.

Chat Bubble System

Visual hierarchy for user/bot messages and rich product interactions.

Typing Rhythm

Indicators simulate a natural human pace to build user trust.

Inline Quick-Replies

Context-relevant buttons to reduce typing friction and guide intent.

The Solution over The SKU

"Home improvement is daunting. By focusing on the user's problem rather than the part number, we unlocked significant conversion lift."

Project Outcomes

Clear, measurable impact for OBI and their customers.

40% reduction in average search-to-cart time
18% improvement in add-to-cart conversion rate
52% autonomous resolution without human support
Global deployment across 10 European markets

Frequently Asked Questions

Expert insights into the design and technical challenges.

How did you handle ambiguous queries? +
We used a single-turn clarification flow. When ambiguity is detected, the bot asks a targeted follow-up instead of listing options, keeping the flow natural.
What were the biggest technical constraints? +
Real-time inventory syncing was critical. We designed graceful degradation for out-of-stock items that suggests the nearest available store.
How was accessibility handled? +
Built to WCAG 2.1 AA with proper focus management and ARIA live regions for screen readers to announce new messages.