Dashboard KPI Reference

This page defines every KPI shown across GC Surge dashboards — what each metric means, how it is calculated, and which dashboard it appears on. Use it as the single reference when interpreting the Home dashboard, Analytics, Alarm Center (Admin View), Alarm Center (Operator View), or the weekly report. Covers: The Three Core KPIs, Alarm Center (Admin View) KPIs, Alarm Center (Operator View) KPIs.

For context on how these KPIs are used in practice, see Alarm Center — Admin View and Alarm Center — Operator View.

The Three Core KPIs

GC Surge tracks three primary performance indicators across the platform:

KPI 1 — Alarm Processing Time (APT) per Operator: the time from when an operator takes ownership of an alarm to when all alarms on that site are closed. Lower is better.
KPI 2 — Cameras per Operator: the total number of cameras monitored per operator per shift. Measures monitoring capacity.
KPI 3 — Alarm Closure Efficiency: the ratio of alarms closed within the APT target to the total alarms in the period. Measures how consistently the team meets its handling standard.

Alarm Center — Admin View

📷 [Screenshot: 11_admin-home-leaderboard.png — Alarm Center Admin View showing KPI bar and operator leaderboard]

The Alarm Center admin view translates the live operation into management decisions — whether onboarding is finishing cleanly, whether NOVA99x is reducing noise, and whether the team has enough capacity for the current site mix.

Cameras onboarded and active — count of cameras successfully reaching an active alarm-sending state.
% of imported cameras successfully reaching alarms — highlights onboarding friction if any cameras are stuck.
Cameras by onboarding mode — breakdown across Mode 1 (Public IP), Mode 2 (Private/VPN), and Mode 3 (Edge).
Cameras stuck beyond 24 hours — cameras in a connecting or pending state for more than a day, with recommended fix actions surfaced inline.
Total alarm events received — across all cameras in the selected period.
Alarms forwarded to the operator queue — count passed from the Alarm Viewer into the operator monitoring workflow.
Alarms auto-filtered by NOVA99x — suppressed before reaching the operator queue.
Estimated operator-hours saved — alarms filtered by NOVA99x multiplied by the average manual review time per alarm (review time is configurable).
Time saved × operator hourly cost — the monetary value of saved hours, shown in the account currency (hourly rate configurable).
Current cameras per operator and additional capacity — how many more cameras the team could handle given the time saved by filtering, at the same APT target.
Potential revenue opportunity — potential capacity increase × average revenue per camera (configurable).
Active cameras per active operator — for the selected shift or period.
Average takeover-to-closed time per operator — aggregated across all operators for the admin view.

Alarm Center — Operator View KPIs

The Alarm Center operator view tells a shift worker whether they are keeping up with the live workload, without turning the experience into scorekeeping noise.

ALARMS HANDLED — total alarms the operator brought to a closed state in the selected viewing period. Counts only alarms the operator personally took to closure.
COMPARATIVE DATA — the same metric for a comparable reference period (prior shift or equivalent window), shown alongside the current value so operators can benchmark their own performance. A notably lower current figure may indicate a busier queue or reduced focus time.
ALARMS/HOUR — throughput rate: alarms handled divided by active shift time. Tracks processing speed normalised for shift length. Sharp drops mid-shift can signal a complex alarm batch, a logged break, or a network disruption.
ALARM PROCESSING TIME — the operator’s average time from acknowledging an alarm to closing it. Aligns with KPI 1 (APT per Operator). Shorter is better; very short (<10s) may indicate bulk-closing without proper investigation.
ALARMS — count of currently open alarms in the operator’s active site queue. A high and rising count during a shift signals the operator may need a site transferred.

Operator Workspace — Site Controls

The Alarm Center operator view groups site management into three panels and a site table. Operators use these to find, claim, and enter sites during a shift.

Viewing Period and Last Shift

The Viewing Period control (top-right) shows the active window for KPI data — defaults to the current shift. The Last Shift button switches the KPI display to the previous completed shift for self-review without opening reports. Clicking it again returns to the current period.

QUEUED SITES ASSIGNED TO YOU

Sites that an admin pre-assigned to this operator. These appear here rather than in the shared Unassigned Sites table. When the panel shows “There are no queued sites for you”, check the Unassigned Sites table below.

PICKED SITE

The site the operator is currently handling. Shows Site Name, Lock Status (In Progress — blue), Last Event, and Events count. The View ZenMode button (eye icon) in the Actions column opens ZenMode scoped directly to this site. When the panel reads “There are no picked sites for you”, claim a site from Queued Sites or Unassigned Sites.

Unassigned Sites

Lists all sites with active alarms that no operator has claimed yet, sorted by alarm urgency. Columns: Site, Last event, Events, Lock status, Actions.

Search site — text input at the top-right of the panel; filters the table by site name.
View ZenMode (eye icon) — opens ZenMode for the selected site so you can assess alarm urgency before taking ownership.
Assign to me (clipboard icon, tooltip “Assign to me”) — locks the site to you and moves it into your Picked Site panel. Only Available (green) sites can be assigned; In Progress (blue) sites are already held by another operator.
Pagination — Rows per page (default 10) and page navigation (e.g., 1–1 of 1) at the bottom of the table.

ONLINE USERS

The right panel shows the count of online operators (e.g., 4 LIVE) and a scrollable list of active users with their names and status. Lets operators see who is on the shift without leaving the Alarm Center.

Alarm Center — Team Metrics

The Alarm Center (Admin View)’s top cards show Alarms Today, Filtered by NOVA99x, Passed to Operator, Avg. Processing Time, Time Saved, Cost Saved, and Cameras Handled (defined in the sections above). A thin summary strip across the very top of the dashboard also shows:

Alarms Handled — a count of alarms the team brought to a terminal state: closed with a tag, escalated, or released to another operator. Alarms an operator opened and walked away from without closing don't count (those fall under SLA Breaches if they exceed the response threshold). It's a throughput figure — how much closure work the team completed — not how many alarms appeared.
Responsive Rate — the share of alarms acknowledged within your SLA: alarms acknowledged within SLA ÷ total alarms × 100, where SLA is your account's configured response time (typically 30–120 seconds). Acknowledgment is the operator opening the alarm modal, not closing it. A rate below 90% is a leading indicator of understaffing or coverage gaps at certain hours — investigate with the Real Alarm Density by Hour chart and the shift roster.
Avg. Acknowledge — the average reaction time across handled alarms: the mean of (alarm opened by operator − alarm ingested by platform). It measures reaction speed, not closure speed. A healthy target is under 30 seconds; above 60 suggests too many alarms per operator, an unfocused workspace, or a noisy queue. Read it as a weekly trend rather than a single value.
SLA Breaches — a count of alarms whose acknowledgment time exceeded your account's response-time threshold. Common causes: under-staffed peak windows (cross-check Peak window in Analytics), operators batching alarms instead of opening them immediately, or NOVA99x letting through too much noise. Prevention is mostly scheduling — match coverage to the Real Alarm Density curve — plus training operators to respond first and investigate second. The leaderboard breaks breaches down per operator for targeted coaching.

Alarm Center — Leaderboard Columns

The operator leaderboard has these columns — Shift, Picked sites, Avg. processing, Alarms handled, Response rate, Activity rate, Unique cameras, and Status. Read them together; no single column is a fair ranking on its own:

Alarms handled — the number of alarms an operator brought to closure. A high count can mean an efficient operator or one auto-closing without investigating, so pair it with Avg. processing, Response rate and closure-tag distribution.
Avg. processing — the operator’s average handling time: the mean of (alarm closed − alarm opened by operator). 30–90 seconds is typically healthy. Very fast (<15s) suggests auto-closing without proper investigation (cross-check their false/true closure ratio); very slow (>3 min) suggests workflow friction or a training need.
Response rate — an individual operator's version of Responsive Rate (their alarms acknowledged within SLA). When one operator's is much lower than peers: they may be on noisy/high-volume sites and overwhelmed at peaks, newer and over-investigating before acknowledging, or on a shift that doesn't match the alarm-volume curve. Cross-check against the Peak window and their assigned site count — coaching fixes the second, scheduling the first and third.
Activity rate — the fraction of an operator's logged-in time spent actively handling alarms; a measure of engagement, distinct from Response rate (which is about speed). An operator can have a high Response rate and a low Activity rate (answered every alarm fast, but long stretches with nothing to handle). Consistently low rates across operators suggest more coverage than needed at that time of day.
Picked sites — how many distinct sites an operator handled events from. Higher means broad coverage, lower means a focused subset — neither is automatically good (focused suits dedicated-customer operators, broad suits floaters). Watch for someone who consistently picks far fewer sites than peers.
Unique cameras — the number of distinct cameras an operator handled events from, shown separately from Picked sites because a site can have one camera or fifty. 5 picked sites with 80 unique cameras is a dense portfolio; 5 sites with 7 cameras means they touched only a sliver of each. Helps keep heavy-camera-count sites from always landing on the same operator.
Status — an operator’s current availability for alarm routing. Online: signed in and within their active shift window, eligible to receive alarms. Offline: signed out, no alarms routed. Off duty: signed in but excluded from assignment (a configured break or a Focus Zone setting that takes them out of the active queue). Off-duty operators still appear on the leaderboard so their historical stats stay visible; they just don’t get new work.

Alarm Center — Site Queue

The Alarm Center admin view has two site panels side by side — Unassigned sites (active alarms with no operator) and Sites Assigned to operators — each with Site, Last event, Events, Lock status, Assigned to, and Actions columns.

Queue — each row is an unowned site (a site with active alarms and no operator handling them). “Queue is clear” means every site with active alarms has an operator — the normal state during well-staffed hours. The panel exists for the opposite: operators signing out without releasing sites, or a peak overwhelming coverage. Clicking a row lets a supervisor take over or reassign.
Lock status — shows who currently holds each site. When an operator starts handling a site’s alarms, the platform locks that site to their session so no one else can take alarms from it concurrently, preventing duplicate work and inconsistent closures. The lock releases automatically when the operator signs out, their session times out, or a supervisor re-routes the site.
Assigned to — which operator owns a site, however that ownership arose. Sites can be assigned three ways: static (an admin fixes the operator at all times), round-robin (alarms pull the next available operator from the shift roster as they arrive), or manual (a supervisor assigns from this panel during a peak or coverage gap). Once assigned, the lock holds until released or the session ends.

My Status and Team Online

My Status — your personal session view: your own alarm counters, your assigned sites, and how long you've been in shift. It's a self-check (“am I keeping up?”), as opposed to the Leaderboard, which is the team view of the same kind of data. Operators usually keep My Status visible during a shift; supervisors live on the Leaderboard.
Team Online (Active count) — the live staffing level: operators signed in and in an active alarm-handling state (Online, not Off duty). Compare it to your roster — 8 rostered but only 5 Active means two are signed out and one is on break. During a peak, watching this number drop is the signal to call in additional coverage. It’s recomputed continuously, not on a fixed poll interval.

Shared KPI Definitions

These definitions apply wherever the metric appears — Home dashboard, Analytics, or the weekly report. The exact formulas may be refined over time, but the operational meaning stays stable.

Alarm Processing Time (APT) — time from an operator taking ownership of an alarm to its closure.
Alarm Closure Efficiency — proportion of alarms closed within the target handling standard.
Cameras per Operator — total active camera workload divided by the operators handling that shift.
Filter Ratio — proportion of total alarms filtered by NOVA99x before reaching an operator. A high filter ratio shows NOVA99x is actively removing noise.
Time Saved — cumulative operator hours avoided through NOVA99x filtering.
Cost Saved — Time Saved valued at the configured operator hourly cost.
Potential Capacity — the estimated number of additional cameras your current team could handle at today’s alarm rate and NOVA99x filter rate, without adding headcount. Calculated from free operator time (hours saved by filtering) divided by the operator-time cost of one additional camera at current load. Treat it as growth guidance — it shifts if alarm rate per camera or filter rate changes.
Potential Revenue — the estimated revenue potential from the Potential Capacity figure, calculated as additional cameras × your configured average revenue per camera.

Reading APT and Cameras per Operator Together

APT (Alarm Processing Time) and Cameras per Operator (CPO) measure different dimensions of performance. APT measures how fast an operator handles each alarm. CPO measures how many cameras the team can cover per shift. Lower APT frees up operator time, which creates the headroom for CPO to grow — the relationship is causal, not mathematical. When reading dashboard results, treat APT improvement and CPO growth as two separate signals that together confirm the platform is working, not as a formula where one directly produces the other.