+56% improvement in onboarding readiness

PROBLEM
Inconsistent onboarding resources created confusion and slowed new starter productivity


ROLE
Research synthesis, journey mapping, information architecture, UI & content design


CHALLENGES
9 teams with conflicting priorities, existing resource dependency, conflicting persona needs, leadership requirements


OUTCOMES
Improved onboarding readiness by 56%, 2 independent teams created contextual version

UNDERSTANDING THE PROBLEM

Inconsistent onboarding experiences affected day-to-day work

Early conversations surfaced the problem stemmed from inconsistency - new starters received different levels of guidance depending on team structure, which affected confidence and productivity in the first weeks.

Sequential 1-1 interviews with 10 new starters across 6 teams narrowed down key themes where inconsistencies lay across.

  • People & Support

  • Logistics & Set-up

  • Resources & Structure

  • Culture & Confidence

ANALYSIS & PRIORITISATION

I mapped the onboarding journey from pre-start to Week 2. This revealed that friction clustered around resource delivery and structure, particularly when informal support wasn’t available.

Resources & Structure emerged as the biggest friction point

A quantitive assessment matrix (which scored each user across each onboarding activity), revealed “Resources and Structure” as the lowest at 51/100. I narrowed my scope to focus on this area within the project’s time constraints, and provided a baseline metric to test improvement.

USER INSIGHTS

Research revealed distinct onboarding needs

While users experienced similar onboarding challenges, their need for structure and guidance varied significantly.

A fully standardised onboarding journey risks slowing experienced joiners, while a self-directed approach leaves others unsure of expectations. A solution needed to provide a consistent foundation while allowing flexibility for different levels of autonomy.

Excerpt: Four lightweight archetypes summarised the different learning onboarding learning styles from research

CONCEPT EXPLORATION

A single, flexible model best supported different onboarding needs

Two main concepts emerged through user flows assessing flow and friction points - either a centralised resource delivery, or improving the existing personalised onboarding checklist. Friction with the existing method lay in reliance on tool licence, and memory to duplicate and send a personalised version.

The resource-model delivery format was the most flexible to suit all archetype needs. I chose to create it as a Figma prototype to test a hypothesis that a structured, interactive model would improve clarity and drive higher engagement. Additionally,

PROTOTYPING & VALIDATION

Early testing exposed friction in content and pacing

Low-fidelity prototypes helped experiment with structure, content and tone. Using Figma Make allowed me to rapidly test ideas and validate the “linear with flex” model over a hub-and-spoke approach - keeping decisions simple - whilst still suiting autonomous users and returners.

Alignment through a Crit and a resources workshop with recent new starters helped find the balance between user expectations and current available resources. I designed sections and resources based on relevancy across all teams, critical information for a new starter, and availability, to avoid the guide being diluted and losing trust with the user.

FINAL DESIGN

Refining the structure to reduce cognitive load

Key design decisions included:

  • a calm, minimal entry point showcasing tone and brand

  • accessibility checks for guided colour contrast

  • side-nav over progress bar, to reduce friction for revisits

  • accordions for dense content, if not relevant for all

  • checklists for guidance of suggested tasks

OUTCOMES & REFECTION

New starters reported clearer expectations and understanding

Outcomes

  • A 56% increase on the “Resources and Structure” score, improving from 51 to 80 following implementation.

  • The onboarding template was organically adopted by 2 other teams (1 internal; 1 external), illustrating ease of customisation, and relevancy of chosen artefacts

  • Positioning the tool as a Figma prototype over a static guide was validated through user feedback and cross-team adoption.

Reflections

  • Future iterations will focus on identifying remaining learning resource gaps and improving long-term maintainability.

  • The interactive format enabled experimentation, but scaling further would require clearer version control and ownership across teams.

Reduced user licence request time from ~5 days to 90 mins

Redesigned the licence request experience by introducing clearer decision guidance, structured request flows and lightweight automation to reduce accidental requests and improve approval speed

View Case Study