UX Case Study · Automating Onboarding for Unbxd
UX Case Study: Automating Onboarding for Unbxd
Automating Onboarding for an AI-Powered Commerce Platform
Unbxd is an e-commerce product discovery platform that applies artificial intelligence and advanced data sciences to connect shoppers to the products they are most likely to buy. Onboarding is a critical moment in any product experience — especially in complex, AI-driven platforms where value is not immediately visible.
At Unbxd, despite offering powerful product discovery capabilities, users struggled to get started and realize value early. This project focused on redefining the onboarding experience to make it more guided, intuitive, and outcome-driven — ultimately improving activation, engagement, and retention.
Netcoreunbxd.com- Stakeholder discussions
- Competitive benchmarking
- User interviews
- User feedback analysis
- Personas & user journeys
- Problem definition
- Brainstorming sessions
- Design strategy
- Feature identification
- Information architecture
- Wireframes & prototypes
- Visual design
- Usability testing
- Analytics & iteration
Research
The research phase combined stakeholder insights and industry best practices to understand onboarding challenges, define user needs, and identify opportunities to improve activation and conversion.
The research phase combined stakeholder insights and industry best practices to understand onboarding challenges, define user needs, and identify opportunities to improve activation and conversion.
The primary users of the platform were merchandisers, catalog managers, and marketing managers, who are responsible for configuring search, optimizing product discovery, and driving conversions, often without deep technical expertise.
- Users struggled to understand how to get started and often relied on external support for setup
- The existing onboarding lacked structure, leading to confusion and delayed time-to-value
- Heavy dependency on custom implementations limited scalability and slowed customer adoption
- Users needed clearer guidance to experience the product's core value early in the journey
Research also highlighted that onboarding should not end after initial activation, but should evolve into a continuous engagement system.
In SaaS ecosystems, ongoing onboarding plays a key role in:
- Driving feature adoption
- Encouraging upgrades and expansion
- Improving retention and long-term customer value
This aligns with the customer engagement funnel, where continuous guidance and education help transform users into long-term advocates.
Analysis revealed that 27% of users dropped off during the configuration stage, making onboarding a critical point of churn, where users struggled to complete setup and failed to experience early value. This highlighted a clear opportunity to improve retention by introducing a more structured, guided onboarding experience that not only reduces friction and accelerates time-to-value but also enables continuous optimization through user behavior insights while aligning the experience more closely with user goals.
Strategy
The strategy focused on guiding users through structured onboarding flows, delivering value early in the experience, and reducing decision-making effort through defaults and intelligent recommendations.
Ideate
Based on research insights, the ideation phase focused on exploring ways to transition from a service-led model to a guided, semi-self-serve onboarding experience, generating solutions that simplify complex setup workflows through step-by-step guidance, contextual support, and pre-configured defaults, while ensuring users can quickly understand the product, make confident decisions, and reach their first value moment with minimal effort.
Design
The design phase focused on translating insights into a scalable, guided onboarding experience by structuring the journey into clear, step-by-step flows that reduce complexity and cognitive load, while enabling users to confidently complete setup and reach value faster through progressive disclosure, contextual guidance, and pre-configured defaults, supported by intuitive information architecture, wireframes, and high-fidelity designs aligned with both user needs and business goals.
Faster Onboarding Using AI Powered Search
It's all about automation
Faster onboarding, flash feed-indexing, automated relevance setup, and self-optimizing algorithms take care of the search experience for the entire customer lifetime.
The Result of It All
The new onboarding experience launched across three regions in Q3 2023.
Amazon wanted to reduce seller time-to-live and dramatically improve the quality of the onboarding experience without compromising on compliance or data accuracy. The redesign had to work across radically different regulatory environments and seller profiles — from sole traders to large multi-entity corporations.
The experience had to scale. Whatever we built for Germany needed to be configurable for Japan. Whatever worked for a sole trader had to flex for an enterprise with 50 SKUs and three tax registrations.
- Automation was broadly accepted once we made it transparent and reversible
- Sellers in user testing described the flow as "surprisingly easy" and "fast"
- Engineering found the modular architecture easier to maintain than the original monolith
- Compliance team sign-off was faster because each step's regulatory mapping was documented
- Expand AI pre-fill to cover bank account verification using open banking APIs
- Allow sellers to save and resume from any step across devices
- Design a "re-onboarding" flow for sellers updating their business structure post-activation
After launch, we tracked the redesigned flow closely for 90 days. The data confirmed what our testing had predicted — sellers moved through onboarding faster, support contacts dropped by 44%, and first-week GMV per new seller increased by 22% compared to the cohort that had used the old flow.
The transparency of the automation layer was the single biggest trust driver. Sellers who saw "we found this from your business registration — confirm or edit" were significantly more likely to accept the pre-filled data than sellers in earlier prototypes where the same data appeared without explanation.
Speed without trust is friction in a different form. The insight that slowed us down in design ultimately sped everything up in production.
The project became an internal reference for how to apply automation to high-stakes UX flows. It is now part of the onboarding pattern library used by teams across Amazon's seller experience organisation.