ForIKEAGermany,wedevelopedtheirfirstinteractiveConversationalServiceAssistant,designedtohelpuserseasilyfindtherightmattresswhileenhancingtheoverallcustomerexperience.
Client
IKEA
Industry
Real Estate & Interior
Services
Technical ConsultingUX DesignVisual DesignSoftware DevelopmentDevOpsProducing
An interactive conversational service assistant that guides customers to the right mattress.
Customers are often overwhelmed when choosing a mattress.
Many users get lost in the complexity of the sleep category. The goal was to put advice at the center and measurably increase trust and purchase intent; success was measured by conversion rate, session duration, and wishlist additions.
The application needed to function on mobile and be relevant both at home and in the furniture store.
Advice should focus on people, not products.

A personal conversational sleep assistant that asks questions and recommends suitable mattresses.
The solution replaces traditional product search with a guided conversation, connecting a knowledge-based recommendation engine with a brand-compliant persona, animated interface, and direct links to product pages or wishlists.
The persona of the assistant is playful, curious, and aligned with IKEA's core values, accompanied by a reduced, modular design and fluid animations that emotionally engage users and make the advice tangible.
The introduction led to measurable improvements: conversion increased by about 12 percent and the average session duration grew by approximately 30 percent, while user feedback praised the quality of the advice.
The solution delivered measurable business benefits.
How we approached it.
We worked iteratively from research to launch.
The implementation followed the phases of discover, prototyping, build, and launch; close collaboration with IKEA's product management and repeated user tests ensured that each iteration significantly improved the advice.

The real magic moment occurs when the assistant resolves uncertainty with a few targeted questions and leads users to a clear product suggestion; the app-like mobile interface and animated responses make the interaction easy and trustworthy.
Technically, the solution was based on a lean, scalable architecture.
A lean architecture ensured performance and ease of maintenance.
Our key learnings help in scaling the assistant.
Modularity in persona, questions, and content allows for quick expansions into other categories; the integration of data, design, and retail advice increases relevance and provides valuable insights for conversational commerce.