The confusion is understandable. The only element of the service the client sees every day is the guard: at the gatehouse, in the lobby, at the car park barrier, at night in front of the warehouse. They talk to them, judge them, complain about them when they sleep and forget them when all is well. But a guard alone, without orders, without a means of communication, without relief and without anyone to call, is not a security service: it is a presence. The guarding agency exists precisely to turn that presence into a system, and that is what its contract must describe.
Post orders: the document that makes the post
Every post — every place a guard is assigned to — must have its written orders: what they watch, what they check (people, vehicles, goods), what they record, what they do in the event of intrusion, fire, accident, altercation, power cut, and above all whom they call, in what order, on which number. Post orders are drawn up with the client during a site visit, from its real risks: a warehouse of electronic goods does not have the same orders as a residence or a school.
A client who has never read them does not know what they have bought. An agency that has never written them does not know what it is selling. The International Code of Conduct for Private Security Service Providers, overseen by ICoCA, makes this formalisation — clear rules of engagement, strictly limited use of force, incident reporting — a basic obligation of the provider, whatever the country.
Recruiting, training, supervising: the invisible part
The agency does not sell labour; it sells selected, trained and supervised labour. Selection starts with identity and background checks, within the limits of what local law allows; initial training covers conduct, communication, access control, what to do in an incident, first aid and fighting an incipient fire; continuing training keeps those reflexes alive. A trained guard knows what they are not allowed to do — search a person without consent, detain someone, use force beyond self-defence — and that knowledge protects the client as much as the guard.
Supervision is the point on which agencies differ most. A mobile supervisor who visits sites at unannounced hours, checks uniform, logbook and the guard’s alertness and answers their questions, makes the difference between a post that is held and a post that is occupied. The client should ask how many sites one supervisor covers, and how often they visit.
The logbook, patrols and proof
The logbook is the post’s register: shift start and end times, notable entries and exits, incidents, calls made, supervisor visits. It is the system’s memory and, in a dispute, its evidence. A logbook empty for a week signals a post where nothing happened — or a guard who does not write.
Patrols, at varying times so as not to be predictable, are checked with tour-control points or apps that record each pass. The client receives a record. That record is more useful than a three-line monthly report: it says what was done.
Observe, deter, alert: the legal limits of the guard
A security guard is not a police officer. They do not arrest, do not search without consent, do not confiscate, do not interrogate. They observe, deter, control access according to their orders, raise the alarm, protect people while waiting for help, and report. An agency that lets the client believe its guards will “deal with” an intrusion by force puts everyone in danger: the guard, the intruder, the client and itself.
That limit has a practical consequence: the system must provide for who intervenes once the guard has raised the alarm. Police, a licensed response company where one exists, the client’s manager: the chain must be written, tested, and the numbers kept current. An alert that reaches no one is a useless alert.
The contract: what it must contain
The guarding contract describes the posts, the hours, the number of guards, the uniform, the attached post orders, the frequency of patrols and supervision visits, the equipment (radio, phone, torch, register), the arrangements for replacement in case of absence, handover times, reporting to the client, the agency’s insurance and the price. It also specifies what happens if the client requests an extra guard for an event, or a reduction in a quiet period.
The price deserves an honest explanation. A guarding service is made up almost entirely of wages and contributions; an agency that sells a twenty-four-hour post at a price that does not cover three full-time guards, their contributions, their training and their supervision is necessarily saving on one of those items. A client who buys the cheapest often buys an undeclared, untrained, underpaid and therefore unreliable guard.
Compliance with labour law as a quality criterion
Guarding is a labour-intensive sector where days are long, nights frequent and wages low. The International Labour Organization insists on social protection and safety at work for private security personnel: registration, contributions, working time, rest, suitable equipment. A guard who strings together two twelve-hour shifts to make up a wage watches nothing during the second. The client has a direct interest in the agency treating its guards well: it is the first condition of their vigilance.
In the African context
Private guarding is, in many of the continent’s cities, one of the largest employers in the service sector, and one of the most visible: villas, apartment blocks, banks, service stations, schools, construction sites, warehouses, headquarters of international organisations. This massive demand has two origins. The first is the perception of insecurity, real or felt, which drives households and businesses to protect themselves. The second is the requirement of foreign principals — companies, embassies, organisations — who impose on their local providers standards of recruitment, training and conduct often higher than national law requires, and who verify their application through audits.
This dual demand creates two very different markets. On one side, a low-price market where poorly trained, sometimes undeclared guards watch houses for a subsistence wage, with a stick and a whistle; on the other, a structured market where licensed, insured, trained agencies with mobile supervision and an alarm centre serve institutional clients. In between, an agency that wants to move up must invest in what cannot be seen: training, supervision, social registration, procedures.
The legal framework exists almost everywhere — agency licensing, guard authorisation, prohibition or strict regulation of carrying weapons, training obligations —, but its enforcement depends on the country and the city. A serious agency does not rely on public control: it applies the framework because that is what protects it in an incident, and because its most demanding clients require it.
The relationship with law enforcement is a decisive practical point. A guard raising the alarm must know which police station or unit covers the site, and the client must know how long an intervention may take. The system adapts to that reality: where response times are long, the agency strengthens deterrence (lighting, fencing, visible presence), the means of alert and, where authorised, its own response team.
Technology arrives through cameras, connected alarms, patrol and logbook apps, often installed by the same institutional clients; it does not replace the guard, but it changes their work: they become the operator of a system as much as a presence. An agency that trains its guards on these tools, and knows how to offer them alongside a post rather than instead of it, gains a real lead.
There remains the human question: guards are often men and women who have come from far away, housed on site, paid monthly, sometimes with deductions from their wages for a uniform or a training course. An agency that registers its guards, pays them on time, respects rest periods and treats them with dignity earns a loyalty that low-price competition never sees — and loyal guards make reliable posts.
Creating the agency: the licence before the uniform
The prerequisites are administrative before they are material: the licence of the country of practice, individual authorisations for the guards, public liability insurance, social registration of employees. Then come uniforms, means of communication, registers, premises, a supervision vehicle. The first contract often comes from a relationship of trust; the second, from the way the first is held.
The agency’s website gains from presenting the system rather than the guard: the types of sites guarded, how supervision is organised, guard training, insurance, licences, and how a client is kept informed of what happens on their site. Photos of uniformed guards reassure; the description of an escalation procedure convinces.
The ten minutes that follow
“If your guard sees someone climb over my fence at three in the morning, what happens in the following ten minutes?” An agency that answers with a precise description of an alert chain — who calls whom, who comes, in how long — sells a system. An agency that answers “they intervene” sells a guard, and that is too little.