A museum's worth of dinosaur data. One scroll to explore all of it.

A dataset this large overwhelms easily.

The brief was to turn a National History Museum dataset — hundreds of dinosaurs, their diet, habitat, taxonomy, and discovery history — into a single-page web app people would actually want to explore. The hard part wasn't the content, it was the volume: presenting that much information in a way that stayed informative and entertaining, stayed responsive across devices, and didn't fall over under its own data.

I reviewed two references: National Geographic Kids, built for casual browsing — bright cards, games, and videos, but no interactive map or way to see where a species actually lived, and no taxonomy detail. And DinosaurPictures.org, built as a research tool — real search and filters by period, region, diet, and type, plus an interactive globe, but no designed visual identity.

That gap became the brief: the audit showed search, filters, and a location map were genuinely useful, while a dry, list-only layout proved that data alone isn't engaging — so I invested in more visual content to keep the experience approachable.

Six weeks. One MoSCoW board.

Based on the competitive analysis and the MVP goals, we prioritized the product requirements using the MoSCoW framework. Since there wasn't enough time to implement every possible feature, we focused on the ones that delivered the most value.

This prioritization shaped the project's six sprints. We began with project kickoff and establishing a shared vision statement, followed by defining the MVP and prioritizing key features. Next, we set up the product backlog, created low-fidelity wireframes, and established the team's GitHub workflow. The following three sprints focused on design, development, testing, and deployment to production. The final sprint was dedicated to project wrap-up, bug fixes, and final refinements before launch.

I sketched low-fidelity wireframes and user stories early, before any visual design, to pin down the core flows — searching or filtering by name, country, weight, diet, and length; browsing dinosaurs by location on a world map; and drilling into a single dinosaur's full detail view. Getting this structure agreed with the team first meant the later visual and technical work had a stable shape to build against.

To move quickly against the deadline, I used Google Looker Studio to rapidly prototype how the dataset's diet and type breakdowns could be visualized. It let me show the team a working version of the charts almost immediately, which kick-started the rest of the interactive dashboard.

I conducted two rounds of usability testing with five potential users. Between each round, I refined the design based on the participants' feedback. During the first round, users completed key tasks while sharing their thoughts and challenges.

First round of usability testing

The navigation menu blends into the page header, making it hard to see.

The search button isn't clearly labelled (e.g. "Submit").

The details section is text-heavy and overwhelming.

In the detail view, the left column is too empty while the right side has too much information.

Second round of usability testing

In response, I gave the navigation menu its own visual weight, labelled the search button clearly, and rebalanced the detail page so information sat evenly across both columns. The second round confirmed that the initial issues had been resolved, but it also revealed a few new usability problems. These final issues were addressed and incorporated into the high-fidelity design.

There's no way to access favourited dinosaurs once they're saved.

It's difficult to choose an individual dinosaur from the map.

The map lacks a zoom feature to explore it interactively.

It's unclear what the "Show Charts" button does.

The detail view appears once a user selects a dinosaur from the list or search.

The taxonomy chart is crowded and confusing.

Simplify until it's legible.

1. Taxonomy visualization

The taxonomy data was more complex than expected — rendered in full, it produced a very deep evolutionary tree that would have overwhelmed rather than informed the user. Instead of showing the whole tree for every dinosaur, I simplified the visualization to highlight only the branches relevant to the dinosaur currently selected.

2. Map visualization

The dataset only recorded country-level location data. A literal pin per dinosaur would have piled multiple dinosaurs on top of each other in the same country, making individual dinosaurs impossible to select. I used a choropleth map instead — highlighting the density of dinosaurs per country — paired with a scrolling carousel listing the dinosaurs themselves, a simple way to pick one without needing a precise pin to click.

Everything, on one scroll.

After refining the design through two rounds of usability testing, I designed it as a single-page, all-in-one learning tool for dinosaur enthusiasts of any age — search, map, taxonomy, and detail views, plus a news feed, all reachable by scrolling rather than navigating between pages. Interactive modules cover dinosaur history, evolution, habitat, and taxonomy without asking the user to leave the page.

The annotations below show how each usability finding was addressed and resolved in the final design.

Introduced a "Favourite" button in the nav, so users can access their saved dinosaurs.

Adapted the map into a choropleth, showing dinosaur density per country and syncing with the filters and list.

Added a scrollable dinosaur list for quickly picking one without hunting on the map.

Detail view appears once a dinosaur is selected from the list or search.

Taxonomy is shown as a tree chart for the selected dinosaur.