Most commercial premises are secured in layers that were added at different times by different people for different reasons. The result is usually strong in one or two places and weak everywhere else, which is the same as being weak, because nobody attacks the strong part.
This is a walk through the five layers, in the order somebody approaching your site would meet them.
1. The perimeter buys you time
A fence does not stop anyone. It delays them and it makes the decision to enter deliberate rather than casual, which matters both for deterrence and for prosecution.
Walk your boundary and look for the three usual failures: a section where a bin, a pallet stack or a parked trailer provides a climbing aid; a gate that is secure when locked and routinely left open during the day; and a stretch that cannot be seen from anywhere. Fixing those three costs very little and removes most of the easy routes in.
2. Access control decides who gets past the door
Keys are the weakest common control, because a key cannot be revoked and cannot be audited. If your building is on keys and has more than a handful of staff, assume there are copies you do not know about.
Credential-based access, whether fob or mobile, gives you two things that matter more than the door itself: the ability to revoke access the day somebody leaves, and a record of who went where. When something goes missing, that record turns a suspicion into an answer.
Check your leaver process. The most common gap in UK commercial premises is not a technical one. It is that nobody owns the step where access is removed.
3. Lighting is the cheapest deterrent there is
Lit approaches, lit entrances, lit yards. It is cheap, it is passive, and it works, both because it removes cover and because it improves everything your cameras record.
The common mistake is glare. A bright light aimed outward blinds the camera behind it and creates deep shadow beyond its pool. Even, downward light across the whole approach beats a few bright floods every time.
4. Surveillance is only useful if it produces evidence
Two questions decide whether a CCTV system is worth its cost. Can you identify a face at the entrance? Can you still retrieve footage from six weeks ago?
A great many systems fail both. Cameras mounted high enough to see the whole yard and nobody’s face. Retention set to whatever the recorder shipped with. If your system fails either test, fixing it is usually a matter of one repositioned camera and a larger drive, not a new installation.
Make sure somebody on site knows how to export footage, and has done it at least once. Discovering the process during an investigation wastes the days that matter most.
5. Procedure is what makes the first four work
This is the layer that is almost never written down and almost always the reason an incident became expensive.
Name the keyholders and check the list is current. Decide who is called at night, and what they are authorised to do. Write down the alarm response: who attends, in what timeframe, and what happens if they find a door open. Agree what staff should and should not do if they find someone on site.
None of this requires spending. All of it determines whether the equipment you have already bought produces a good outcome or a bad one.
A short review you can do this week
Walk the boundary looking for climbing aids and blind stretches. Pull the access list and remove everyone who has left. Stand in the yard after dark. Check the CCTV retention setting and view last week’s entrance footage. Read your alarm response procedure, and if there is not one, write half a page.
That is an afternoon, and on most sites it will find more than a five-figure equipment upgrade would.