A guard can be punctual, polite and fully uniformed yet still be ineffective when something unusual happens. That is the central problem with how to assess guard performance. If assessment measures attendance, patrol completion and incident numbers alone, it may confirm activity without proving protective capability.

Good guard performance is not about catching people out or building a disciplinary file. It is about establishing whether the security presence is reducing risk, recognising change, making sound decisions and helping the organisation respond properly when pressure rises. This matters in every environment, from a busy venue or retail estate to construction, corporate premises and critical infrastructure.

Start with the job the guard is actually expected to do

Before assessing an individual, be clear about the operational purpose of the post. A concierge-style reception role, a mobile patrol, a control room operator and a gatehouse officer may all be described as security guards, but their performance requirements are different.

The question is not simply, “Did they follow the assignment instructions?” It is, “Can they identify and manage the risks this location genuinely faces?” A guard at a public-facing site may need strong engagement, observation and escalation skills. A lone officer on a remote site may need disciplined communication, accurate reporting and the judgement to avoid becoming isolated by a developing issue.

This is where many contracts go wrong. Generic key performance indicators are imposed across several locations because they are easy to report. The result is an attractive spreadsheet that says little about readiness. Measures must reflect the risk assessment, the operating environment and the decisions the officer may need to make.

Assess guard performance in three areas

A useful assessment considers behaviour, judgement and outcomes. Each matters. None is enough on its own.

Observe behaviour, not just task completion

Task completion has a place. Patrols, access checks, equipment tests, visitor procedures and handovers should be completed properly. But a patrol confirmation only proves that a point was scanned. It does not prove the guard looked, listened or understood what was in front of them.

Direct observation is essential. Supervisors should spend time on site, at different hours and in normal operating conditions. They should look at how officers approach people, manage access, position themselves, communicate with colleagues and maintain awareness while dealing with routine demands.

A capable officer does not become so focused on a queue, delivery or visitor log that they stop noticing the wider environment. They use professional curiosity without becoming confrontational. They recognise when something does not fit the normal pattern and know what to do next.

Assessment should also include the less visible moments. Is the officer alert during a quiet period? Is the handover specific or a vague exchange of keys and information? Are control room updates timely and useful? These details often expose whether standards are embedded or performed only when supervision is expected.

Test judgement through realistic discussion

Real incidents are relatively rare in well-managed environments. That is a good outcome, but it means incident data alone cannot tell you whether guards are capable. A team can have no reported incidents because there were none, because issues were resolved quietly, or because officers failed to recognise and report them.

Use structured conversations based on credible site scenarios. Ask what the officer would notice first, what information they would gather, who they would inform and what factors would change their response. The value is not in securing a rehearsed answer. It is in hearing how the person thinks.

For example, an officer may say they would call a supervisor when a visitor becomes agitated. That is only the beginning. Can they explain how they would create space, maintain communication, protect others nearby and record the facts without exaggeration? Do they understand their authority and its limits? Can they distinguish an inconvenience from a concern requiring escalation?

Scenario assessment should be proportionate and psychologically safe. Its purpose is development, not embarrassment. If an officer cannot explain a basic response, that is useful evidence of a training, briefing or supervisory gap. Leaving the gap unaddressed because the individual is otherwise reliable is poor risk management.

Review outcomes with care

Reports, logs, access records, patrol data and feedback all provide evidence. They should be reviewed for quality, not simply volume. A short report can be excellent if it records clear facts, actions taken, decisions made and relevant times. A lengthy report can be useless if it is vague, speculative or missing the point.

Look for patterns. Are similar concerns being identified across shifts? Are there repeated failures in access control, key management or escalation? Do reports show that officers identify developing issues early, or only record events once someone else has intervened?

Low incident figures should never be treated automatically as success. At a site with a known history of nuisance behaviour, public disorder or repeated unauthorised access attempts, a sudden fall in reporting may warrant examination. It may reflect better control. It may also reflect officers who no longer challenge, record or escalate.

Use standards that can be seen and heard

Vague expectations produce vague assessment. “Be vigilant” and “provide a visible presence” are not performance standards. They offer no reliable basis for feedback.

Standards should describe observable conduct. A guard maintains control of an access point by confirming identity and authority before entry, applying visitor procedures consistently and escalating exceptions. A guard provides an effective handover by identifying unresolved issues, changes to the environment, outstanding actions and any relevant welfare concerns.

The same principle applies to communication. Rather than judging an officer as a good communicator, assess whether they give concise, accurate information; whether they adapt their manner to the person and situation; and whether they report promptly through the agreed route.

This helps supervisors give useful feedback. “You need to be more alert” is unlikely to change performance. “During the delivery issue, you focused on the driver’s paperwork but did not maintain sight of the open gate. Next time, pause the process, control the access point and request support if needed” gives the officer a practical adjustment.

Separate individual weakness from system failure

Not every poor outcome is a guard failure. Officers cannot perform well if instructions are contradictory, staffing is inadequate, communications fail, induction is superficial or supervisors are absent. Equally, a competent officer can be undermined by a culture that rewards keeping problems quiet to avoid inconvenience for the client.

This distinction is uncomfortable because it prevents easy blame. If several officers make the same error, examine the system before treating it as a series of individual shortcomings. Was the procedure clear? Was it practised? Did the shift receive a proper briefing? Was there enough time and support to do the job correctly?

Contract managers should be especially cautious about measuring guards solely against client satisfaction. Courtesy and responsiveness matter, but security staff are sometimes required to create friction. A professional challenge to an unauthorised visitor may be inconvenient. It may also be exactly what the site needs.

Build assessment into normal supervision

An annual appraisal is too remote from operational reality. Guard performance should be assessed through routine supervision, targeted observations, report reviews and short capability discussions. The process should be consistent enough to be fair, but flexible enough to reflect different roles and risk levels.

A simple assessment record can capture the task or situation observed, the required standard, evidence of performance, the officer’s own account, agreed actions and a review date. The record matters, but the conversation matters more. Officers need to understand what good looks like and why it protects people, property and organisational continuity.

Where weaknesses appear, respond with the right intervention. Some issues require clearer instructions or site familiarisation. Others require coaching, further training, closer supervision or a decision that the person is not suited to the role. Repeating generic refresher training for every problem is wasteful. Development should address the actual gap in knowledge, judgement, behaviour or confidence.

For organisations preparing for Martyn’s Law duties, this is particularly relevant. A security plan and staff training record will not compensate for guards who cannot recognise suspicious behaviour, communicate concern or act within a coordinated response. Capability has to be visible in practice.

Make improvement measurable

The strongest assessment programmes show whether development has changed performance. After coaching or training, observe the same behaviours again. Review whether reports have improved in quality, whether escalation is more appropriate and whether officers can explain their decisions with greater confidence and accuracy.

Mildot Group’s approach to capability assessment is built on this principle: turn theory into action and test whether it holds up in the working environment. The aim is not perfect answers. It is dependable judgement when information is incomplete and the consequences matter.

A guard force should not be judged by how quiet it looks on a normal day. Judge it by whether its people can notice the unusual, make proportionate decisions and give the organisation reliable information before a manageable concern becomes a larger problem.