One of the most awe-inspiring institutions on the planet had a website making it genuinely difficult to find anything. I set out to understand why — and to design something better.
Most personas are invented. Lisa was built from evidence. I started with NASA's own government analytics data to understand who actually visits the site — then cross-referenced degree programs and career pathways that would realistically attract women into science fields in the late 1990s.
That research discipline matters. A persona grounded in real data creates design decisions you can defend, not just justify.
Note: While a proto-persona is informed by secondary research rather than primary interviews, grounding it in government analytics data gives it more credibility than assumption alone.
Built using NASA government website analytics data and researched educational pathways that would realistically support women entering meteorology and climate science in the late 1990s.
Five moderated usability sessions over Zoom. Four distinct tasks, each representing a realistic goal a NASA visitor might have. The results weren't subtle.
Average success rate: 60% · Average website opinion score: 2.8 / 5 · Moderated via Zoom · Transcribed via Otter AI
"Yeah… I never would have found it, if you got like 6 billion things out there."
"The filters aren't very helpful. 'Last year' must mean something different to me."
"You don't know how people classify stuff. I think the search algorithms are a little too loose."
"That felt oddly difficult. I just think they should be able to do better."
I scored NASA.gov against Nielsen's 10 Usability Heuristics — the industry standard for identifying structural UX issues without user testing. The site has genuine strengths. It's visually stunning, loads quickly, and houses an extraordinary volume of content. The problems cluster around navigation, information scent, and search.
Notably, the site's readability skews high-academic: a Dale-Chall score of 9.90–10.13 places homepage articles at a college-level reading threshold — appropriate for researchers like Lisa, but worth flagging for general public outreach goals.
Sampled key UI elements against WCAG 2.0 AA standards (4.5:1 normal text, 3:1 large text and UI components). All tested elements passed — several significantly exceeded the minimum.
Across heuristic evaluation, usability testing, competitor analysis, and card sorting, three root problems emerged consistently. Every design decision in the prototype traces back to one of these.
Every change in the prototype is traceable to a specific research finding. Here's how the evidence translated into design.
Card sorting exercise revealed that users grouped NASA content into three primary mental models: topic-based exploration, media (multimedia), and news/events. The existing 5-section nav mixed formats with subjects, creating ambiguity.
Redesigned top-level nav to: "Explore by Topic," "Multimedia," and "News & Events." Logo moved to standard top-left position. Explore search field renamed to "Explore Topics" and repositioned for clarity.
No usability test participant scrolled to find the sitemap. All participants used search rather than scroll-to-browse. Homepage scrolling was identified in the 2x2 matrix as High Value / Low Effort to eliminate.
Wireframed a condensed homepage with a single hero image and drop-down navigation exposing content categories on hover — reducing scroll requirement and surfacing information scent at the top level.
Task 3 (podcast search) had the lowest success rate at 50%. Users who found the Podcasts page saw only 5 series titles with no episode listings — leading them to question whether the content even existed.
Recommended a Podcasts page restructure that shows both series names and individual episode listings — matching users' mental model for how podcast content is typically organized.
Multiple participants called out ambiguous date filters ("Last year — does that mean calendar year or 12 months?") and search returning articles without publication dates. The 2x2 matrix flagged search as High Effort / High Value — a longer-term investment, not a quick fix.
Framed as a phased recommendation: immediate — add publication dates to all article cards. Longer term — invest in metadata tagging and filter UX improvements that give users meaningful control over search results.
I benchmarked NASA against both direct competitors (organizations competing for the same audience) and indirect ones (organizations NASA could learn navigation and search patterns from).
| Organization | Feature Analysis | Competitive Advantage | Heuristic Analysis |
|---|---|---|---|
| NASADirect | Tons of articles, press releases, multimedia, and podcasts on science and space exploration for all ages. | Vast collection of information on science and space exploration across many media formats. | It's difficult to find information without simply using a keyword search, and even then it's difficult to know if the correct result is being returned. The filtering system is difficult to use and articles are missing dates. |
| Library of CongressDirect Competitor | Excellent top-level search filters, followed by search by catalog, collections, visitor-type (researcher, visitor, teacher), Blogs, US Copyright Office, Trending, service & Programs, etc. The photo on their main page brings you to their events page. Very nicely organized site. | More information than NASA and extremely well organized. | Efficient navigation, organized, clear labeling, graphics are well-thought out, easier to navigate and filter than NASA's site, consistent design, visually appealing with a highly robust filtering system to refine search results. |
| Space ForceIndirect Competitor | Not really a competitor. Their website is really a military recruiting site, not an educational site. | Directly targeting US Space Force recruiting. Animation contains good quality scientific drawings of photos and fact sheets, but is very limited in scope compared to NASA. | The site contains a fraction of the content carried by NASA, so the navigation is very limited in scope. |
| SpaceXIndirect Competitor | Heavy on cool animations, light on content and in depth science articles. | Much more of a commercial site. Papers on payload (User's Guide). | Efficient navigation (even with heavy animation and high-end graphics), organized, clear labeling, consistent design, matched my expectations of what a SpaceX website might look like, visually appealing, easy to scan and read. |
Key insight: The Library of Congress — comparable in content volume — achieves dramatically better findability through audience-type navigation and robust search filtering. NASA's closest model for what "better" could look like.