Skip to content

Activity timeline & ranking

The Activity page is where most managers spend their time. It's the day-by-day, hour-by-hour record of what each employee actually did.

Two tabs

The page is split into two tabs at the top:

  • Timeline — one employee, one day, the full breakdown.
  • Ranking — the whole team, one week, sorted by productivity.

Controls at the top of the page

  • Employee selector — searchable dropdown of every employee in your organisation. Defaults to the first member alphabetically (or yourself, if you're not an owner — non-owners only see their own data here).
  • Date navigation — prev / next buttons, a Today shortcut, and a calendar picker for jumping to any day.

The same controls drive both tabs; switching tabs preserves your selected employee.

The Timeline tab

Header stats

A row of summary numbers for the selected employee + day:

  • Total tracked time — clock time the tracker was actively recording.
  • Productivity — the weighted average of all segments' 1–5 scores, surfaced as a single percentage.
  • Focus ratio — fraction of the day spent on segments scored 4 or higher.
  • Context switches — how often the foreground app/site changed.

The hour-by-hour timeline

Below the header is a vertical list of hours. Each hour expands to show the segments the AI grouped activity into. Every segment shows:

  • The app or site in focus (icon + name).
  • The window title or page title (when available).
  • The duration of the segment.
  • A productivity score from 1 (distraction, red) to 5 (highly productive, green).

Hovering a segment surfaces a tooltip with the AI's reason. Clicking opens a side panel with:

  • A screen recording clip (if your policy is set to Record + Save or Record + Save while unproductive and the segment fell into the "save" bucket).
  • The full text the AI extracted from the screen (titles, document text it could read, terminal output, etc.).
  • Buttons to correct the segment if the AI mis-scored it — your correction becomes a timeline correction and feeds back into future scoring for the same employee.

Legend

A small colour legend lives at the top of the timeline:

ColourScoreMeaning
Green5Highly productive — deep work
Light green4Productive — work tools and comms
Grey3Neutral — research, light admin
Light red2Low productivity — short breaks, scrolling
Red1Distraction — entertainment, gaming

The exact rubric is whatever your productivity prompt says — see Productivity prompt.

Filters

  • Flagged only — show only segments that match a suspicious-app rule from your policy. A small badge in the header tells you how many flagged segments are in the current day.
  • Score range — slide to "≥ 4" to see only deep work, or "≤ 2" to see only distractions.

Timeline corrections

The link Timeline corrections at the top right of the Activity page lists every correction you've recorded for the current employee. You can edit or remove them from there.

The Ranking tab

A leaderboard of the whole organisation for the selected week (the date picker snaps to a Monday).

Each row shows:

  • The employee's name and avatar.
  • A progress bar representing their average productivity that week.
  • Total tracked hours.
  • Number of recorded days.
  • A "weekly report" link that deep-jumps into the AI report for that employee + week.

On mobile screens the table collapses to stacked cards with the same information.

The Ranking tab uses the same data the weekly report email draws on, so the order you see here is the order the weekly digest will list employees in.

Tips

  • Use the calendar to compare days. Stepping back a week and picking the same day-of-week is a quick way to see whether yesterday's distraction patterns are recurring.
  • Open two tabs side-by-side — Timeline in one, Ranking in the other — when you're doing weekly review meetings.
  • Treat corrections as feedback to the AI, not just data cleanup. The more you correct, the more the scoring on that employee learns your idea of "productive".

ScreenJournal end-user documentation. For sales or account questions visit screenjournal.ai.