Reframing a Health Platform Around Structured Data

Reframing a Health Platform Around Structured Data

Restructured navigation, content architecture, and product hierarchy to shift from content sprawl to a focused, PHR-first experience. Removed unused community features, aligned taxonomy to MedlinePlus standards, and introduced AI-enhanced search to improve clarity and findability.

Project scope

Role

UX Designer

Timeline

2024

Team

Design, Content, Engineering, CEO

Status

Shipped

Context

StoryMD began as a broad health content platform layered onto a developing personal health record (PHR). The product hierarchy placed library content, communities, and PHR at equal weight, creating confusion around the core business.


Content was organised alphabetically within loosely defined “Channels,” and search relied heavily on keyword matching, leading to inconsistent discovery and poor information hierarchy.

Problems

  • Feature sprawl diluted product focus

  • Library structure mirrored internal CMS logic rather than user mental models

  • “Channels” terminology confused users

  • Search returned HealthJournals instead of specific articles

  • Community boards were unused but retained structural prominence

  • Navigation hierarchy did not reflect product priorities

Strategic shift

I advocated for repositioning the PHR as the core product experience, with the library serving as contextual support. This required rethinking navigation hierarchy, removing unused community features, and restructuring the content taxonomy to align with established health topic standards.


The solution was not a visual redesign — it was a structural realignment of the product’s core hierarchy and priorities.

Original StoryMD high level navigation

New StoryMD high level navigation

Remove noise

Community boards removed
Free-tier account removed, streamlining sign up

Align to mental models

Top-level taxonomy aligned to MedlinePlus health topics

Restructure navigation

Side navigation (desktop)
Bottom navigation (mobile)
App-like immersion

Architecture & navigation redesign

With the strategic shift to PHR-centred design agreed, I iterated primary navigation designs with my fellow designer, landing on an immersive app-like left-hand desktop navigation and lower mobile navigation. The site no longer placed the PHR on a level with Library and Communities, forcing users to drill down within the PHR to access medical information.

Original and new landing pages and navigation

Original and new library organisational structure

Improving search through structured logic

Before

Initially, the search results returned whole HealthJournals. StoryMD HealthJournals had been designed to mimic a book with a table of contents directing users to articles within the HealthJournal. Users were confused as the links on the search results page directed them to the HJ’s hero page and not to the specific article applicable to their search.


And this wasn’t the only issue we encountered. Due to the search matching titles and keywords in the CMS, there was little to no hierarchy in the results returned and at times the results seemed entirely unrelated. Add to this the fact that simple glossary definitions were hidden in a third tab and you had search results that frustrated and confused users.

After

We implemented a layer of AI-powered search, allowing us to generate the most accurate and relevant search experiences. It fed search queries through a proprietary large language model (LLM) to understand the intentions behind searches and serve personalised results for each person.


With appropriate results being returned we also showed results for individual articles, and returned relevant glossary results as a card at the top of the screen, ensuring users got the answers they were looking for.

After addressing problems with the library search results we also fixed the issue of having search bars in different areas of the site. If a user wanted to search for items in their health record the field was treated as a filter. We combined them all into a global search that allowed users to search any area of the site, but that defaulted to the user’s current location.

Original and new search results pages

New global search autocomplete field

Outcomes

  • Re-centered the product around its core value proposition

  • Eliminated unused features to reduce structural noise

  • Established a scalable taxonomy foundation

  • Increased precision in search result targeting

  • Created alignment between product architecture and monetisation model

  • Re-centered the product around its core value proposition

  • Eliminated unused features to reduce structural noise

  • Established a scalable taxonomy foundation

  • Increased precision in search result targeting

  • Created alignment between product architecture and monetisation model

Reflection

This initiative required balancing founder vision with user clarity. By prioritizing structure over expansion, the platform became more cohesive and strategically aligned — setting the foundation for subsequent AI and workflow innovations.