← Back to Work 03 of 07
Analytics Heatmap Analysis Navigation Design UX Research Academic · 2020

Pratt News Analysis

A week-long behavioural analytics study of Pratt Institute's newly redesigned news website — combining Google Analytics, Hotjar heatmaps, and scroll maps to diagnose a 94% drop-off rate and surface four targeted design recommendations.

94%
Articles drop-off rate
43%
Clicks above fold
75%
Abandon search mid-way
4
Design recommendations
01 — Context

A new website. The same old problems.

Pratt Institute had recently relaunched its news website. Despite the redesign investment, early signals suggested users weren't engaging with the content — navigation felt unclear, and editorial content wasn't being discovered.

This study applied behavioural analytics to understand what users were actually doing on the site, not what they said they would do. The goal was to provide the Pratt communications team with specific, actionable recommendations grounded in observed behaviour rather than opinions.

Pratt News website overview

Pratt News website · Newly redesigned · Desktop view (64% of total traffic)

02 — Research

One week of data. Four tools. One clear story.

The study ran for one week (Sept 24–30) using Google Analytics for traffic and behaviour data, and Hotjar for heatmaps, scroll maps, and movement maps. The focus was desktop — accounting for 64% of all site visitors — though mobile findings were documented for follow-up.

Google Analytics
Traffic sources, session depth, bounce rates, and user flow through the site's primary navigation paths.
Hotjar Heatmaps
Click density mapping across homepage, article listing, and article detail pages — revealing where attention clustered and where it didn't.
Scroll + Movemaps
Scroll depth maps to identify where users stopped reading, and mouse movement maps to trace attention patterns on key pages.
Hotjar heatmap of navigation area

Hotjar heatmap · Navigation area · 43% of all page clicks occur above the fold on nav elements

03 — Findings

Users wanted the content. The architecture was blocking them.

The data told a consistent story: users arrived with intent and left without finding what they came for. Four specific patterns emerged across the week's data.

Finding 01
94% drop-off after clicking "Articles" or "View all articles" — the CTA was triggering a page that users immediately abandoned.
Finding 02
Search was the most-clicked element (6.6% of all visitors) but 75% didn't complete a search — the interface opened a field without surface-level affordance.
Finding 03
Only 46.8% of users scrolled far enough to see related articles — a key discovery pathway that the majority of users never encountered.

"Your findings also show us that the labels we chose for the navigation options are really working well."

— Pratt Communications Team, post-presentation feedback
Mouse movement map showing navigation patterns

Mouse movement map · Homepage · Users track navigation heavily but struggle to exit into article content

Scroll map showing drop-off before related articles

Scroll map · Article page · 53% of users never reach the related articles section at page bottom

04 — Recommendations

Small changes. Measurable impact.

Each recommendation mapped directly to a specific observed behaviour — no speculation, no "best practice" generalisations.

  • Reduce navigation height and make it sticky — with 43% of all clicks happening on nav elements, keeping it visible as users scroll eliminates a major friction point.
  • Add predictive search with nav visible — the search field's 75% abandonment rate indicates users need more surface-level guidance before committing to a query.
  • Move related articles to a sticky right sidebar — placing this content in persistent view rather than below the fold would expose it to the 53% of users currently missing it.
  • Relabel "View all articles" to "Browse all articles" — user mental model testing showed "browse" better matches the exploratory intent behind the click, likely reducing the 94% drop-off.
Recommended related articles sidebar design

Recommendation · Related articles · Sticky sidebar keeps discovery pathway visible throughout reading

"Thank you so much for these design recommendations. They look really thorough and professional."

— Pratt Communications Team

The client validated both the findings and the direction, noting that the navigation label analysis confirmed choices they had made intuitively. A follow-up mobile mockup was delivered addressing a secondary set of mobile-specific patterns identified in the data.

← Previous
Design Jedi
Next project →
COVID-19 Predictive Spatial Analysis