IoT & AI

B2C

Deploying AI-powered fitting technology for mainstream e-commerce

Client

Motesque

Motesque

Model

Product accelerator

Product accelerator

Services

Product strategy

Product strategy

UX/UI design

UX/UI design

Rapid prototyping

Rapid prototyping

Proposition testing

Proposition testing

Info

Motesque, a German deep-tech startup, developed AI-powered technology that analyses body shape through computer vision to recommend perfect product fit.

Their first commercial partnership with Fahrrad.de, one of Germany's largest bike retailers, needed a consumer interface that could make complex science feel simple.

In Germany, return rates due to incorrect sizing reach 34%, nearly double the European average. For retailers, technology that solves fit before purchase addresses a critical margin problem.

Challenge

The brief appeared straightforward: visual design and prototyping for an AI-powered size recommendation widget.

The reality was more complex. Existing design work was superficial: linear journeys that ignored the sophisticated logic AI demands. Motesque had no design system, no comprehensive wireframes, and dozens of unmapped scenarios that would determine completion or abandonment.

The widget needed seamless cross-device functionality, image validation handling, differential treatment of new versus returning users, and Motesque AI integration that respected Fahrrad.de's brand.

Our analysis

AI sophistication is irrelevant if the interface can't handle real consumer behaviour: blurry photos, mid-journey abandonment, device switching. Existing flows assumed ideal paths. AI-powered experiences must work for every path.

We reframed the brief from visual design to systems integration. The widget wasn't decoration on top of AI—it was the bridge between breakthrough biomechanics and messy human behaviour.

First we needed to diagnose all possible scenarios before designing any of them, build reusable systems rather than one-off solutions, validate with enough fidelity to surface problems before code.

What we did

01.

Mapping the complete user journey

We discovered and documented dozens of scenarios previously not considered. Desktop users needed QR code handoffs to mobile where camera access made photo capture practical. Image uploads could fail validation for wrong body positions, multiple people in frame, or poor lighting. Users calculating size for one bike model then switching to another shouldn't repeat the entire process. These weren't edge cases. They were core journeys determining success or failure.

Mapping the complete user journey

We discovered and documented dozens of scenarios previously not considered. Desktop users needed QR code handoffs to mobile where camera access made photo capture practical. Image uploads could fail validation for wrong body positions, multiple people in frame, or poor lighting. Users calculating size for one bike model then switching to another shouldn't repeat the entire process. These weren't edge cases. They were core journeys determining success or failure.

02.

Building systematic design foundations

We created a comprehensive component library and style guide from scratch, balancing two requirements: widgets needed to feel native to each retailer's site while maintaining consistent Motesque brand expression. The system needed to scale beyond this initial implementation for bikes to cover their roadmap including mattresses and footwear without rebuilding for each category.

Building systematic design foundations

We created a comprehensive component library and style guide from scratch, balancing two requirements: widgets needed to feel native to each retailer's site while maintaining consistent Motesque brand expression. The system needed to scale beyond this initial implementation for bikes to cover their roadmap including mattresses and footwear without rebuilding for each category.

03.

Creating instructional clarity through illustration

Initial designs used generic stock photography that communicated nothing. We replaced these with custom illustrations demonstrating exactly what users needed to do: which body angles to capture, how to position themselves, what measurements the system required. This more functional approach to imagery reduced confusion at critical moments.

Creating instructional clarity through illustration

Initial designs used generic stock photography that communicated nothing. We replaced these with custom illustrations demonstrating exactly what users needed to do: which body angles to capture, how to position themselves, what measurements the system required. This more functional approach to imagery reduced confusion at critical moments.

04.

Prototyping for validation, not presentation

We created two high-fidelity interactive prototypes for desktop and mobile that simulated complete user journeys including error states, loading behaviours, and cross-device handoffs. These allowed real user testing before development investment and served as shared truth between Motesque's AI team, Fahrrad.de's e-commerce team, and our development partners. Concrete representations of what would be built reduced ambiguity and prevented expensive mid-development changes.

Prototyping for validation, not presentation

We created two high-fidelity interactive prototypes for desktop and mobile that simulated complete user journeys including error states, loading behaviours, and cross-device handoffs. These allowed real user testing before development investment and served as shared truth between Motesque's AI team, Fahrrad.de's e-commerce team, and our development partners. Concrete representations of what would be built reduced ambiguity and prevented expensive mid-development changes.

Project impact

De-risked first partnership

Our systematic approach prevented launch issues that would have damaged Motesque's credibility with Fahrrad.de, the kind of failed implementations that plague AI startups.

De-risked first partnership

Our systematic approach prevented launch issues that would have damaged Motesque's credibility with Fahrrad.de, the kind of failed implementations that plague AI startups.

Designed for scale

The design system works across bikes, mattresses, and footwear without rebuilding for each vertical. Motesque can demonstrate their technology to new retail partners with production-ready experiences.

Designed for scale

The design system works across bikes, mattresses, and footwear without rebuilding for each vertical. Motesque can demonstrate their technology to new retail partners with production-ready experiences.

Production-ready in 6 weeks

From brief to deployment within Fahrrad.de's e-commerce platform in a single accelerator project. Speed that matters when competing for retail partnerships against established providers.

Production-ready in 6 weeks

From brief to deployment within Fahrrad.de's e-commerce platform in a single accelerator project. Speed that matters when competing for retail partnerships against established providers.

De-risk AI innovation without sacrificing speed

Building AI-powered products, IoT platforms, or complex B2B2C technology?

The gap between breakthrough capability and consumer adoption determines commercial success. Sophisticated algorithms mean nothing if consumers abandon the interface before it provides value.

The virtual fitting room market is projected to grow from $5.71 billion to $24.30 billion by 2032, but only companies that solve the systems problem, not just the surface problem, will capture that opportunity.

Our accelerators compress months of discovery and design into focused sprints that prove value fast, because we understand both the technology and the human behaviour it needs to serve.

Book an accelerator

Previous

PetLink

Next

Vitesse PSP