Aria Welcome State

• High Drop-off Rate

  57% of users close the Aria panel immediately upon landing on the canvas without initiating a message.

• Blank-Input Paralysis

  First-time self-creators often don't know what to ask or what Aria is capable of doing beyond basic Q&A.

• Context Disconnect

  The welcome state feels static and disconnected from the user's entry flow—Template Creation versus AI Site Creation—and ignores active canvas selections.

• Perceived Intrusion

  Canvas-first editors can feel pushed into AI interactions when they simply want to visual-edit, leading to silent disengagement.

• Increase First-Session Engagement

  Boost initial interaction rates and conversation depth within the Aria panel.

• Intent-Aware Guidance

  Help users instantly understand Aria's capabilities through prompt buckets categorized by user intent: Explore, Learn, and Smart Actions.

• Reduce Friction & Paralysis

  Replace blank inputs and generic chips with dynamic, selection-aware suggestion starters.

• Balanced UX

  Deliver an attractive, playful, and inviting starting state that complements visual editing without aggressively taking over the canvas.

• Primary

  First-time self-creators entering Wix Harmony via the Template Flow, arriving with a complete design on canvas and Aria automatically open.

• Secondary

  Second-time returning users reopening Aria to continue chatting, tweaking, or expanding their site.

Understanding the Product Area

Analysis showed that while Aria is in context in the editor, beginners struggle to formulate prompts from scratch. However, driving engagement with Aria yields a direct business uplift, as active Aria users are more likely to convert to premium subscriptions.

The Project Scope

Restructure the welcome layout, re-organize text hierarchy, establish intent-based suggestion bucketing, execute an end-to-end content workflow, conduct rapid AI prototyping, build Figma specs, and submit an initial draft PR.

Business Strategy

Align AI assistance directly with user intent at critical onboarding moments to accelerate site completion, directly driving Wix's core KPI of premium purchases.

Key Insights

• Generic three-chip layouts fail because they don't cater to different user intentions, such as learning versus immediate action.
• Clear visual hierarchy separating greetings, dynamic guidance, and action starters improves scannability.
• Selection-aware prompts make AI assistance feel relevant rather than intrusive.

• Selected Concept

  Following internal UX reviews and dependency alignment, the final design retained the main input field at the bottom while placing intent-grouped prompt chips centrally. Clicking a prompt chip triggers a pop-up bottom-sheet widget with tailored prompt lists that feed directly into the bottom input bar.

Re-organized Text Hierarchy

• Personalized Greeting

  A prominent Hello {{user_name}} headline grabs attention and establishes rapport.

  

• Session-Aware Guidance

  Dynamic subtitles adapt by session: “How do you want to start?” for first-time visitors and “What would you like to do today?” for returning users.

  

• Intent-Grouped Buckets

  Prompts are categorized into core user intents—Add, Generate, Design, and Learn how to do things—reflecting real user data.

  

• Bottom-Sheet Prompt Widgets

  Central prompt chips trigger interactive bottom-sheet lists populated with data-backed starter prompts that auto-fill into the main chat input.

  

Content Writing & Localization Workflow

At Wix, creators now take end-to-end ownership of interface content, adding three dedicated steps to the product lifecycle:

• Early Concept Copywriting

  Drafted initial explanatory copy during early ideation and component drafting so concepts communicated design intent clearly to management and stakeholders.

• AI-Assisted Refinement & Terminology Alignment

  Mapped visual content keys on finalized components and used internal Claude skills aligned with Wix terminology and tone-of-voice guidelines, refining options with the Content Writer Lead.

• Feature Keys Spec & Localization Handoff

  Built a structured Feature Keys Document with content-key tables, screenshots, and visual callouts, enabling direct code integration and multi-language localization handoff.

• Design Specs

  Created comprehensive Figma component specifications defining first-time and returning-user states, prompt-bucket behaviors, content-key mapping, and bottom-sheet widget transitions.

• Draft PR Development (Cursor AI & Claude Code)

  As Product Creator, took full ownership of drafting the initial development branch. Connected to the company repo and used Cursor Agent and Claude Code to implement UI changes, micro-interactions, and behavioral logic while adhering to Design System rules and developer requirements.

• Enhanced Technical Handoff

  Pushed the completed draft to a new Pull Request and handed it to engineers alongside the Figma specs. The functional draft served as a technical first layer that made cross-functional communication and development smoother.

• AI-Assisted Prototyping Accelerates Decisions

  Using Claude to build functional, interactive low-fidelity prototypes drastically shortens review cycles with leadership.

• Value-Driven AI Onboarding

  Meeting users with context-aware, intent-driven options is far more effective at overcoming blank-input paralysis than presenting an empty search box.

• Integrated Content & Code Ownership

  Expanding creator ownership into UI copy mapping, localization keys, and draft code PRs bridges design and development, resulting in faster execution and higher product quality.