Search Experience Discovery
As a user, I want to have a more efficient search engine and find the content faster
Project Context
Our Vision, to become the world’s #1 performance coach, is partially connected to our ability to understand the user’s needs and show the best content possible in our search results.
Context & Objectives
Happiness — Users will be able to find the most suitable content for their needs in an easy way, improving their happiness when using the product.
Task Success — Users will be able to find the content that they were looking for completing the task of searching, finding, and viewing
Adoption & Retention — Users will trust that they can use our application to find solutions to their problems instead of using Google, for example.
My Role & Responsibilities
• UX Research & Analysis: Broke down search into query processing, candidate retrieval, ranking, and result delivery
• Feature Definition: Defined delivery models for “no typing,” “typing,” and “no results” states
• Prototyping & Testing: Created interactive prototypes for each state; ran usability sessions
• Stakeholder Alignment: Worked with engineering, data science, and product teams to translate UX patterns into search-engine rules
Process & Activities
Decomposition & Audit
Mapped the search pipeline:
• Query Processing: String normalization, typo tolerance
• Candidate Retrieval: Matching algorithms
• Ranking: Scoring rules based on relevance and user preferences
• Results Delivery: UI patterns per scenario
• Audited current zero-result rate and average clicks to success
• Query Processing: String normalization, typo tolerance
• Candidate Retrieval: Matching algorithms
• Ranking: Scoring rules based on relevance and user preferences
• Results Delivery: UI patterns per scenario
• Audited current zero-result rate and average clicks to success
Delivery-Model Design
No Typing (Empty State):
• “Recommended for You”: Top 3 keywords based on persona preferences.
• “Recent Searches”: Quick access to last 3 queries.
• “Recommended for You”: Top 3 keywords based on persona preferences.
• “Recent Searches”: Quick access to last 3 queries.
Typing (Suggest State):
• Live “Related Keywords” suggestions ranked by preference and substring match.
• “See All Results” CTA to full results once the query is submitted.
• Live “Related Keywords” suggestions ranked by preference and substring match.
• “See All Results” CTA to full results once the query is submitted.
No Results (Zero-State):
• Friendly feedback (“We couldn’t find ‘X’—try these topics…”)
• “Search Instead For”: Top 3 recommended keywords to steer users back on track.
• Friendly feedback (“We couldn’t find ‘X’—try these topics…”)
• “Search Instead For”: Top 3 recommended keywords to steer users back on track.
Prototyping & Validation
• Built Figma prototypes covering all three states.
• Conducted 12 remote usability tests: measured time to first result and user satisfaction (5-point scale).
• Iterated on wording (“Recommended for You” vs. “Popular in Your Role”) and visual hierarchy.
• Conducted 12 remote usability tests: measured time to first result and user satisfaction (5-point scale).
• Iterated on wording (“Recommended for You” vs. “Popular in Your Role”) and visual hierarchy.
Implementation Collaboration
• Documented UI specs and edge-cases in Confluence.
• Partnered with data-science to tune scoring weights for personalized relevance.
• Held biweekly syncs with engineering to monitor zero-result trends and refine algorithms.
• Partnered with data-science to tune scoring weights for personalized relevance.
• Held biweekly syncs with engineering to monitor zero-result trends and refine algorithms.
Exploring Scenarios
Playable Prototype - Search Experience feature (no audio)
My Process, Outcomes, and Impact
Research
• Conducted 5 stakeholder interviews (product, data, support)
• Ran 12 remote usability tests on search tasks
• Analyzed 3 months of search analytics (zero-result rate, CTR)
• Benchmarked 4 top search-UI implementations
• Ran 12 remote usability tests on search tasks
• Analyzed 3 months of search analytics (zero-result rate, CTR)
• Benchmarked 4 top search-UI implementations
Ideation
• Facilitated 2 ideation workshops with UX + data-science
• Produced 15 sketch concepts across three search states
• Built 12 low-fidelity wireframes (empty, suggest, zero)
• Created 2 end-to-end search journey maps
• Produced 15 sketch concepts across three search states
• Built 12 low-fidelity wireframes (empty, suggest, zero)
• Created 2 end-to-end search journey maps
Validation
• Developed 3 mid-fidelity prototypes in Figma
• Tested with 12 users completing 3 core tasks each
• Logged 28 usability issues and 60 satisfaction scores
• Iterated through 2 design versions
• Tested with 12 users completing 3 core tasks each
• Logged 28 usability issues and 60 satisfaction scores
• Iterated through 2 design versions
Execution
• Delivered 5 high-fidelity mockups
• Authored 6 pages of design specs and edge-case documentation
• Held 8 design-dev syncs and supported 3 QA cycles
• Added 4 search-widget components to the design system
• Authored 6 pages of design specs and edge-case documentation
• Held 8 design-dev syncs and supported 3 QA cycles
• Added 4 search-widget components to the design system
Key Learnings
• State-Driven Design Matters: Tailoring UI to each search state significantly reduces user friction.
• Language & Tone Build Trust: Friendly zero-state copy keeps users engaged rather than frustrated.
• Data & UX in Tandem: Close collaboration with data teams ensures that UI patterns align with back-end capabilities and metrics.
• Continuous Monitoring: Post-launch analytics are essential to catch regressions and iterate further
- - -
Check another project done for MindTools
MindTools