The project itself :
Project Overview
AI-Powered Synthesis: This search will generate a dynamic contextual summary. The summary doesn't just list records; it explains the historical context of the collections and why they are exceptionally relevant to your specific research goal.
Problem:
Users, particularly genealogists, may not know which specific collections to search within for their research needs, especially when starting with a broad idea or a non-standard query. This creates a gap between broad research intent and specific collection discovery, leading to frustration and underutilized content.
Goal:
If we introduce an AI-powered collection recommendation tool, then we will increase user engagement with specific collections and improve discovery action volume, because the tool will provide intuitive guidance that bridges the broad search intent to focused collection discovery.
My role:
UX designer leading the Card Catalog search page redesign and AI feature
Responsibilities:
Final responsive designs (Web, Tablet, Mobile)
Content strategy reviews
UX design reviews
Discovery & research
Accessibility reviews
Interactive prototypes
All about the research :
Discovery
Because our legacy results view in search was already crowded with records and filter controls, dropping in a prominent AI feature risked overwhelming our users. I researched usage data on AI features and explored comparative products with existing AI summarization tools.
UX Research Plan
Quantitative audit
Audit current search initiation volume, query latency, and detail-view drop-offs across the legacy search flow to identify critical friction points and establish empirical benchmarks for redesign opportunities.
Qualitative Sentiment & AI trust
Synthesize in-depth researcher feedback to uncover primary discovery pain points and validate user trust and readiness regarding AI-assisted tools within their research workflows.
Competitive Benchmarking
Analyze industry leaders to evaluate how they integrate generative AI, specifically examining translating complex information into digestible summaries.
Baseline Metrics
To establish a baseline for the Card Catalog Search, I partnered closely with Product and 2 data analysts to audit search initiation rates and any other friction points that could shape our iterative design roadmap.
As a result, our team conducted a Deep Dive Analysis on understanding where users may struggle to engage with our Search experience.
Metrics were initially calculated over Day 0-7 after first exposure. Out of 280k test users, 2.6k [0.93%] clicked on search results. This shows early adoption and engagement. However, over time, the Search Click Rate showed [-1.25%] and Search Success rate showed a statistically significant drop with [-2.25%]. Despite low volume, Search demonstrates a strong repeat value among users with a 7-day return rate.
Customer Research
Collaborated with our UX Researcher to synthesize qualitative sentiment around AI-assisted research tools, focusing specifically on how genealogists uncover breakthrough family discoveries.
These user insights provided essential validation for our product strategy, ensuring our push to integrate generative AI features into the search experience was anchored in genuine research trust, workflow relevance, and high-value discovery moments.
Deep dive-interviews were conducted with 10 Genealogist power users, who have deeply integrated AI into their family history workflows. This is what we found.
Transcribing & analyzing records
Users upload screenshots or files of handwritten documents, census records, and old newspapers for text analysis.
Breaking "Brick-Walls"
Researchers query AI for leads on missing burial sites, parents, or ancestors. In addition, they use AI to generate spreadshets for DNA matches found in Ancestry.
Contextualizing lives
With AI, users like to generate a summary of an ancestor's life based on entered events, and provides historical context for events like the 1918 pandemic.
Competitive Analysis
To establish an intuitive layout strategy for the Card Catalog, I analyzed 6 AI-generated component patterns and transcription architectures across Google, YouTube, and Meta—extracting high-performing interaction models to integrate seamlessly into our results view.
I created all the app pages mockups, incorporating the right design elements such as typography, color, and iconography. Also I developed all the necessary components and needed elements. The goal was to demonstrate the final Suplan app in as much detail as possible.
The project schematically :
Define
Leveraged quantitative usage metrics and qualitative researcher feedback to establish the UX architecture—mapping comprehensive end-to-end user flows, aligning on terminology with Content Strategy, and creating journeys tailored to common habits of our power users - Genealogists.
Interaction flows
I designed two primary interaction flows and component mockups for the Card Catalog's core additions: the Contextual Search keyword input and the AI-generated summary container.
These components were meant to to pressure-test interactive decisions with Product and engineers prior to full visual polish.
I documented the behaviors of two main components - detailing default, interactive, and conditional states to streamline technical handoff and eliminate ambiguity. This was especially helpful for our engineers were offshore and most of our communication was async.
Content Strategy Workshops
Collaborated with Content Strategy across two workshops and async design reviews to architect clear UI copy for search headers, help tooltips, interactive slider states, and AI summary modules—optimizing visual hierarchy and onboarding prompts to prevent cognitive overload.
This is a mockup of the three-panel onboarding modal designed to introduce Contextual Search (Card Catalog search via natural language). The copy covers three core capabilities: free-form text input, an adjustable search scope slider (broad to exact), and an option to tab over to legacy keyword search. The copy was carefully crafted to drive immediate user adoption.
Persona Journey Map
Defined 5 distinct use cases anchored around 2 power user personas, mapping a specialized end-to end user flow across historically rich research paths.
This helps to evaluate the card catalog search architecture against the workflows of power-user subscribers-specifically genealogists who rely heavily on the Catalog for primary-source discovery.
I created all the app pages mockups, incorporating the right design elements such as typography, color, and iconography. Also I developed all the necessary components and needed elements. The goal was to demonstrate the final Suplan app in as much detail as possible.
The clear version :
Design
With the end-to-end user flows and core researcher goals established, I moved into rapid design exploration for the Card Catalog search experience. This phase progressed through multiple high-fidelity iterations shaped by peer design critiques, accessibility audits to ensure strict WCAG compliance, and cross-functional reviews with Product and Engineering to pressure-test technical feasibility.
Mockups
Designed an initial suite of 11 high-fidelity mockups for the Card Catalog search experience, encompassing the primary entry point, a four-screen onboarding walkthrough, four search component states, and two AI-summarized results views.
These early designs reviews were crucial to get to engineers, product, and content alignment. Not only to get feedback for iterations, but to also launch the product in a timely manner.
I created all the app pages mockups, incorporating the right design elements such as typography, color, and iconography. I also included captivating and visually appealing images, and developed all the necessary components and elements.
The goal was to demonstrate the final Voo's app in as much detail as possible.
The project schematically :
Outcome
The biggest highlight was integrating a new, effective and exciting AI tool at the search results level that our power users can get exited about.
Takeways
The series of hand-drawing frames that visually describe and explore a user's experience with a product.
Impact:
Qualitative feeeback
The engineering team initially deployed a functional proof-of-concept for live testing at RootsTech conference, earning overwhelmingly positive feedback from advanced genealogists who struggle with valuable records being buried under obscure naming conventions.
Validated by strong customer enthusiasm and strategic insights from our Corporate Genealogist, Crista Cowan, we leveraged this momentum to greenlight the project and launch a targeted A/B test.
What I learned:
Researching and validating this project helped me understand the value of using the Card Catalog search tool and how expert genealogists conduct their research.
Next Steps
The series of hand-drawing frames that visually describe and explore a user's experience with a product.
Conduct follow-up usability testing on the new Card Catalog search experience
Review and monitor the outcome of the soft launch at the quantitative level
Redesign the search results page with a cleaner, less-cluttered layout