Why I Almost Always Recommend IP Cameras for Business CCTV

Why I Prefer IP Cameras With PoE

My biggest frustration with traditional analog CCTV is the cabling and connections. Separate video and power paths can mean more components to install, test and diagnose. A properly installed analog system can work well, but those extra connections become more noticeable across a large business site.

With a compatible IP camera and PoE switch or recorder, one network cable provides data and power. From an installer’s perspective, that means one cable to install, test and replace if necessary.

PoE doesn’t eliminate faults. Cameras, cables and switches can still fail. What it gives me is a more consistent infrastructure to plan and support. The customer sees the camera, the picture and the mobile app; I also see the cabling and equipment that have to keep working behind them.

The Real Advantage Shows Up When Something Fails

Image quality matters, especially when you need to identify a person or understand an incident. But it isn’t the biggest reason I recommend IP cameras. The difference becomes clearer months or years later, when a camera goes offline or a client cannot find a recording.

At that point, I need to locate the fault. With an IP system, I can work through the camera, its network connection, PoE supply, switch and recording configuration. That gives me a logical troubleshooting path, provided the installation has been planned and documented properly.

CCTV is long-term infrastructure. The initial quotation is only part of the cost: someone will need to maintain it, replace failed equipment and add cameras as the business changes. I would rather support a simple, well-planned system than one that only looks good on installation day.

What Happens If the PoE Switch Fails?

A failed PoE switch can take several cameras offline at once. That’s a real limitation and should be part of the design. In my experience, though, replacing a confirmed failed switch is often easier than spending hours tracing individual power faults through a poorly installed analog system.

Check the power supply, uplink and connected equipment before assuming the switch itself has failed. Think about the impact of losing each switch, and plan suitable power protection and recovery arrangements for the site.

When Does Analog CCTV Still Make Sense?

I’m not saying everyone should replace their analog CCTV. If the system works reliably, provides useful footage and meets your requirements, there may be no reason to replace it. Existing cabling, replacement costs and a limited budget can all make analog the practical choice. If equipment has failed or the business needs to expand, compare the cost and support implications of repairing the existing system with moving to IP.

Mixing CCTV Brands: What ONVIF Does and Does Not Guarantee

I prefer a consistent CCTV environment where practical. HiLook and Hikvision are brands I work with regularly, and familiar tools and management platforms can make support easier. That is a practical preference, rather than a reason to assume every model will work with every other model.

Mixed-brand systems can work. ONVIF profiles define shared capabilities for conformant devices and clients: Profile T covers advanced video streaming, Profile G covers recording and storage, and Profile M covers analytics metadata and events. Check the exact models, supported profiles and features you need. Basic video working across brands does not guarantee that every analytics feature, event or configuration option will work through the recorder.

In October 2025, ONVIF announced the phase-out of Profile S and recommended Profile T. Existing Profile S systems do not stop working because of that announcement, but new purchases should account for current authentication and security requirements. When I’m troubleshooting a client’s system, I want to spend time finding the fault rather than discovering that two products implement a feature differently. Confirm compatibility before buying, even within one brand.

Camera Resolution: How Many Megapixels Do You Actually Need?

This is probably one of the first specifications people look at. You’ll commonly see resolution labels such as:

• 2MP (2 megapixels)
• 4MP
• 5MP
• 8MP (often marketed as 4K UHD)

Higher resolution can provide more detail, but more megapixels don’t automatically solve every CCTV problem. A higher-resolution image is only useful if the rest of the system allows you to make use of that detail. The camera still needs:

• Suitable lighting.
• The correct lens.
• Good positioning.
• Appropriate viewing distance.
• Sufficient storage.
• Appropriate recording settings.

For example, putting an extremely high-resolution camera on the corner of a building and using it to cover an enormous area doesn’t necessarily mean you’ll be able to identify someone standing 30 or 40 metres away. The person might simply occupy too little of the image. This is why I don’t like the approach of:

“Let’s just buy the highest megapixel camera.”

The better question is:

“What level of detail do we actually need at this location?”

That is a much more useful way to select CCTV equipment.

Turret, Dome or Bullet?

The physical design of the camera matters too. You’ll commonly come across turret, dome and bullet cameras, and each has situations where it can make sense.

Fixed Lens vs Varifocal

Another important decision is the lens. A fixed-lens camera has a fixed focal length, which determines its field of view together with the sensor. A varifocal camera allows that field of view to be adjusted. This can be extremely useful when you’re dealing with a location where the exact viewing requirement isn’t perfectly predictable until the camera is installed and positioned.

For example, you might know that you need to monitor a gate, but the exact amount of surrounding area you want in the image may only become obvious once the camera is mounted. A varifocal lens gives you more flexibility to fine-tune the view. Some motorised models also allow remote zoom and focus adjustments. The trade-off is coverage versus detail: a wide view shows more of the scene but makes a person or object smaller in the image. A narrower view covers less area while giving the important subject more of the frame.

Camera Placement Can Matter More Than Camera Price

This is something I would encourage anyone planning CCTV to think about carefully. A R5,000 camera in the wrong location can be less useful than a R2,000 camera in the right location. The exact prices will obviously vary, but the principle remains.

If the camera is pointed at the wrong angle, has the wrong field of view or is positioned too far away from the area of interest, upgrading the camera may not solve the underlying problem. I’ve seen this kind of thing in real installations. Sometimes the answer isn’t:

“We need a better camera.”

Sometimes the answer is:

“We need to move the camera.”

That is why site planning is so important. Before installing a camera, you should consider:

• What area needs to be covered?
• Where will people approach from?
• Where will the light come from?
• Is the camera facing directly into sunlight?
• Is there anything that could obstruct the view?
• How far away will the important subject be?
• What level of detail is actually required?

These questions can have a bigger impact on the final result than the specification printed on the box.

Daylight, Shadows and Difficult Lighting

Lighting is one of the biggest challenges in CCTV. A camera can produce a fantastic image during the day and a much less useful image at night. The same can happen when the camera is positioned incorrectly relative to the sun. Imagine a business entrance where the camera is looking toward a bright outdoor area.

During part of the day, the camera may be dealing with bright sunlight outside and a relatively dark entrance underneath a roof. Without suitable image processing, you can end up with one part of the image being correctly exposed while the important part is too dark. This is where features such as Wide Dynamic Range (WDR) become useful. WDR is designed to help cameras deal with scenes containing significant differences between bright and dark areas.

It is particularly relevant for locations such as:

• Entrances.
• Vehicle gates.
• Loading areas.
• Windows.
• Reception areas.
• Outdoor-to-indoor transitions.

This is another reason I wouldn’t choose a camera purely by megapixel count. A camera with the right features for the environment can produce more useful footage than a camera with a higher headline specification that isn’t suited to the scene.

Night Vision and Low-Light Performance

The useful question is not simply whether a camera has night vision, but whether its footage will show the detail you need. Infrared (IR) can illuminate a scene, but results depend on the sensor, lens, distance, movement, ambient light and reflective surfaces.

A stated infrared range is not a promise that you can identify a person at that distance. Check recorded footage under the conditions that matter for the site, including moving subjects at night.

Colour Night Vision

Some cameras can retain colour in low light, while others use additional visible lighting. If clothing or vehicle colour matters, plan the lighting as part of the installation. The camera still needs enough light to produce useful detail.

AI Detection Is Becoming More Useful

Another area that has changed considerably is camera analytics. Modern cameras can do more than simply record everything they see. Depending on the equipment, cameras may offer analytics such as:

• Person detection.
• Vehicle detection.
• Line crossing.
• Intrusion detection.
• Loitering detection.
• Object detection.
• Event-based alerts.

The important distinction is between recording video and understanding events within video. A basic motion detection system might trigger because a tree moved in the wind, a shadow crossed the image or an insect passed close to the camera. More advanced analytics can attempt to distinguish between people, vehicles and other movement. That can make notifications more useful. But I still wouldn’t treat AI detection as magic.

The quality of the camera, positioning, lighting, configuration and the specific analytics available all matter. A badly positioned camera doesn’t suddenly become a good security camera because it has AI written on the box. Again:

The installation comes first.

How Far Can an IP Camera Be From the Network?

This is another area where a little planning goes a long way. For conventional copper Ethernet, plan around a 100-metre channel limit between active network devices, including patch leads. That doesn’t mean every camera installation stops at exactly 100 metres. Suitable PoE extenders and other network designs can support longer distances, subject to equipment and power requirements. You may use:

• PoE extenders.
• Intermediate network switches.
• Fibre optic links.
• Additional network cabinets.
• Different network architectures.

The important thing is not to simply assume that a longer cable will solve the problem. If a camera is a significant distance away from the main network cabinet, that should be considered during the design stage. This is particularly relevant on:

• Farms.
• Warehouses.
• Large industrial sites.
• Schools.
• Complex office buildings.
• Large residential properties.
• Multi-building sites.

The physical size of the property can have a major impact on the correct CCTV architecture.

Do You Need Internet for IP Cameras?

This is another question I hear fairly often. And the short answer is:

Not necessarily.

An IP CCTV system can operate on a local network without needing an internet connection for the cameras to communicate with the NVR. Internet connectivity becomes important when you want features such as:

• Remote viewing.
• Remote notifications.
• Remote management.
• Cloud services.
• Access to footage while away from the site.

This distinction is important because an IP camera system and an internet-connected camera system are not necessarily the same thing. The cameras can communicate locally. The NVR can record locally. Your internal network can carry the CCTV traffic. The internet is then an additional connection used for remote functionality where required. For businesses, I actually think this is a useful distinction because it highlights something that often gets overlooked:

Your CCTV system is part of your network.

It doesn’t automatically need to be exposed directly to the internet.

What About Wi-Fi Cameras?

For homes, I think Wi-Fi cameras can be perfectly acceptable. They’re convenient, relatively easy to install and can be a practical solution where running network cable isn’t realistic. For businesses, however, I look at them differently. Many people buy Wi-Fi cameras because they assume that removing the network cable automatically makes installation easier. Sometimes it does. Sometimes it simply moves the problem.

The camera may no longer need a network cable, but it still needs power. You can then end up with:

• Power adapters.
• Extension cables.
• Power points in inconvenient locations.
• Wireless coverage problems.
• Signal interference.
• Batteries that need charging.

At that point, the supposedly “easy” installation can become less professional than simply installing a proper wired camera. For a business environment, I usually prefer wired CCTV whenever possible. It gives me a more predictable infrastructure and, with PoE, the network connection and power can travel through the same cable. That makes sense to me not only from an installation perspective, but also from a support perspective.

How Much CCTV Storage Do You Actually Need?

There isn’t one universal answer. Two systems with the same number of cameras can require very different amounts of storage. For example, consider two businesses:

They may both have eight cameras, but their storage requirements could be dramatically different. That’s why I prefer designing CCTV around the actual requirement rather than simply saying:

“Eight cameras? Buy an 8-channel NVR.”

The number of channels tells you how many cameras the recorder can support. It doesn’t tell you whether the storage capacity is appropriate.