Qatalyst 2.0 · Preview traction

Who is exploring the Qatalyst 2.0 preview

preview.qatalystcarbon.com · — · external visitors only (our team and the design tool excluded) · updated · refreshes daily at 10:00 SGT

— people opened the preview from — countries, and — of them engaged. Only — got past the sign-in screen (— of engaged visitors), so sign-in is the biggest drop-off. Engaged sessions averaged — minutes.

Reach and engagement

People reached—opened the preview
Engaged—— of people reached
Signed in and explored—— opened product screens
Countries—— cities
Came back—more than one session
Engaged sessions over 2 min—of — engaged sessions

External visitors per day

People per day, external only. Lines mark releases.

From opening to exploring

People who reached each step, 7–23 Oct. Red marks the biggest drop.

Page speed for visitors

Time until the main content shows, for the slowest quarter of loads (LCP p75).

—
—
—
—

— faster since the site was split into smaller files on 16 Oct.

Who engaged

Where visitors come from and what they use

Where engaged visitors are

Engaged people by country.

Roles explored

People who viewed screens as each role. One person can try several.

Most-visited screens

People who opened each screen.

Language and device

People who switched language, and devices of engaged visitors.

Patterns

When people visit, and what each role explores

When people visit

Engaged sessions by the weekday and hour they started, Singapore time (UTC+8). Darker means more sessions.

What each role explored

People who opened each screen while viewing as each role. Darker means more people. Shown for the 12 most-visited screens.

Behaviour

Where people click

External visitors' clicks drawn over each screen, 7–23 Oct, desktop only. Orange shows where clicks gather; blue dots are clicks that did nothing (dead) and rage clicks. Positions are approximate: clicks are recorded in 16-pixel steps on screens 1,280–1,920 px wide and redrawn on the current v16 layout.

Sign-in screen

Before visitors enter the app. Most drop-off happens here.

— clicks · — did nothing · — rage clicks

Today

The home screen every role lands on.

— clicks · — did nothing · — rage clicks

Project overview

The four questions for one project.

— clicks · — did nothing · — rage clicks

Explore

The project catalogue and filters.

— clicks · — did nothing · — rage clicks

Library

Standards and methodologies.

— clicks · — did nothing · — rage clicks

Signals

News and events matched to projects.

— clicks · — did nothing · — rage clicks

How these numbers were counted

  • External only. Our own testing (this team's two Singapore networks and the Claude app) is excluded: — people and — sessions. The design team's working copy also reports to the same analytics (— people); it is excluded too.
  • Engaged means at least 10 seconds on the page or one action (a click, sign-in or screen view). — visitors did neither; most were single US hits typical of email link scanners and link previews.
  • Feedback: — click on Give feedback and — submitted forms (Tally).
  • No visits carried a referrer or campaign tag, so we cannot yet tell which emails, posts or videos drove traffic. The videos' QR codes are tagged, but nobody has arrived through them yet.
  • Click heatmaps come from PostHog's heatmap data. The preview is one web address, so each click is assigned to the screen the visitor had open at that moment, from the screen-view events. Clicks before sign-in are drawn on the sign-in screen.
  • The data refreshes every day at 10:00 Singapore time from PostHog. Each source in Sources shows when it was last updated.