From installed agent to organisation-wide insight
Six deliberate stages take you from a 40 MB installer to a live workforce dashboard — with an admin approval gate, your own classification rules, and honest, counts-only capture at every step.
Install a lightweight agent on every desktop
One signed installer per platform — Windows, macOS, Linux — run by hand or pushed silently through your MDM. The agent works in the background at under 2% CPU, capturing locally and reporting to the central dashboard. No network? It buffers the day on-device and defer-syncs the moment the connection returns.
No machine reports until an admin says so
After install, each machine requests registration and lands in a Pending queue. Admins approve or reject every device — or switch on automatic approval for trusted rollouts. The same lifecycle controls handle the exit: remotely uninstall an agent, or disable its login, without ever touching the machine.
Teach Momentum what productive means — per department
Global App Rules mark every application Productive, Non-Productive, or Unmarked. Then departments override where reality differs: a design tool is productive for Design and unmarked for HR; a social app is productive for the Marketing team that runs your accounts. Scores are only as fair as the rules — so the rules are yours.
Set the baseline every score is measured against
Before the first report means anything, you define what a normal day looks like: workday hours, expected desk hours, target productive hours, office start and end times, and the productivity thresholds that separate green from red. Scores compare people to your standard — not to an arbitrary one.
An honest chronological record from first login
Once approved, the agent builds each person's Timeline from their first login of the day: green blocks for active work, red for idle stretches, in chronological order. Alongside it sit the Time Summary pie, App Usage — including offline applications — and Input Statistics: counts of mouse clicks and keystrokes with a per-app distribution. Counts, never content.
Three altitudes of insight, one source of truth
The same captured signal powers every view. Timeline answers 'how did this day unfold'. Activity answers 'did this person hit their expected desk time' with start/finish, active vs expected desk time, and a productivity score. The Workforce Dashboard answers the organisational question: an Organization Health Score rolling up productivity, attendance, security, and AI adoption, plus top and bottom performers and departments, and a master grid of every user with live agent status.
Agent → sync → classification → dashboards
A simple pipeline with one deliberate property: nothing is lost offline, and nothing is scored until it passes through rules you control.
Captures active/idle time, app usage, and input counts locally on Windows, macOS, and Linux.
Online, data streams to your tenant; offline, it queues on-device and back-fills on reconnect.
Every app-minute is weighed against Global App Rules and department overrides, then measured against your baseline settings.
Timeline, Activity, and the org Workforce Dashboard render the same signal at three altitudes.
Exactly what is captured — and what is not
Momentum measures activity, not content. The boundary is technical, not just policy: the agent records counts and categories, never what anyone typed or read.
Built for the people who run the fleet
Every agent reports its own operational telemetry back to the dashboard. Logs stream Info and Error entries per machine, tagged by component — Daemon, Installer, Task Scheduler, Upgrade — so a stuck rollout is diagnosed from your desk, not a desk visit. Resource Usage shows CPU and memory per device, proving the agent stays out of your users' way.
How-it-works FAQ
Does anything get captured before a machine is approved?+
No. After install, the machine requests registration and sits in the Pending state. Only once an admin marks it Approved (or your automatic-approval policy admits it) does the agent begin reporting time, app usage, and input statistics.
What happens when someone works offline?+
The agent keeps capturing locally — active/idle time, app usage including offline apps, and input counts — and performs a deferred sync when the machine reconnects. Timelines back-fill automatically, so a day on a plane still shows up complete.
Can the same app be productive for one team and not another?+
Yes — that is exactly what per-department overrides are for. A Global App Rule sets the default (Productive, Non-Productive, or Unmarked), and each department can override it. Figma can be productive for Design while staying Unmarked for HR.
How is the productivity score calculated?+
Against your own baseline. You define workday hours, expected desk hours, target productive hours, and productivity thresholds in settings; the score compares each person's actual active time in productive-classified apps to those targets. It's one of four pillars — alongside attendance, security, and AI adoption — rolled into the org-wide Organization Health Score.
What does 'input statistics' actually mean?+
Counts only: how many mouse clicks and keystrokes occurred, with a per-app distribution. Momentum never records which keys were pressed or the content of anything typed — it is an activity signal, not a keylogger.
How do we retire a machine or off-board a person?+
From the dashboard: admins can remotely uninstall the agent from any machine, or disable agent login so the device can no longer report. Historical data stays intact for reporting.