1,216 inbound (1,165 deduplicated), 1,730 guarding-desk outbound, 115 officer briefings. Test, internal, silent and never-connected calls excluded.
Someone on the way in about a minute, not five to fifteen.
An alarm comes in. The desk verifies the site, checks the reference and works the dispatch procedure. Tool calls, failed attempts and the wait for a person stay on the case record.
Set up with you on a call — not self-serve.
For alarm receiving, security dispatch, facilities emergency response and out-of-hours desks.
look_up_officerverify_sitearrange_coverMeasured from our own live security control room · 1 September – 14 October 2026. See the figures and their basis ↗
What the desk handles.
Alarms, keyholder and access, intercom, and the calls that arrive at 3 a.m.
- Alarms
Identify the site and follow its response procedure.
- Keyholder & access
Work access questions and keyholder contact under the agreed SOP.
- Intercom
Verify the caller’s request and record what happened.
- Officer calls
Brief the officer and record the response.
- Patrol & office traffic
Keep operational calls attached to their case.
- Briefings & voicemail
Record the next step, including a failed attempt to reach someone.
522 alarm-desk calls by keyword classification: alarms 49%, keyholder and access 15%, intercom 9%, officers 5%, office and patrol 5%, brief or voicemail 16%. Approximate.
Measured on the desk.
1 September – 14 October 2026. The figures describe real security operations, not a forecast for another sector.
Median over 108 dispatches on the current agent. An earlier dashboard shows 68 s over 121 including the legacy agent; the full-desk review shows median 71 s, P95 140 s. Before: the operator's previous control room, 5 min typically and up to 15.
Median from ring to the first exact site match.
The dispatch reference read back to the caller: 104 of 108.
Share of location checks that failed, latest two weeks. Before: 2%, and the operator's own figure was 98% verified.
The questions worth asking.
What happens if it cannot identify the site?
It does not guess. Location checks that fail are counted: 2% → 0.8% Measured vs operator. A call it cannot resolve goes to a person with what it already has. There is a recorded case either way.
Share of location checks that failed, latest two weeks. Before: 2%, and the operator's own figure was 98% verified.
Does it decide who to send?
It proposes, and committing someone is an action that waits for a person. The wait, the person and the answer are all on the record.
What if the caller is wrong?
It checks rather than accepts. On one call the caller gave a postcode that did not match the site: caught in 13 s; dispatched at 77 s Measured.
This is one measured call, not a response-time promise.
Measured from our own live security control room · 1 September – 14 October 2026. See the figures and their basis ↗
Nothing dropped under pressure.
closed with no human touch
Measuredof check calls with an outcome
Measured vs operator- Playbooks
- Versioned procedures
- Tools
- Granted per tool
- Answers in
- 2 s Measured vs operator
- Escalates to
- A named person
- Hours
- Set by your procedure
Book-ons, check calls, late attendance and arranging cover, across sites.
For security guarding, facilities, care, logistics and field service — wherever shifts run across sites.
See the module →One platform. Different procedures, different calls, and a separate set of numbers to hold the desk to.
Your operation,
with agents.
Start where the pressure is: a pilot on one desk, measured against your own baseline, with the numbers agreed before it starts.