Local NVR or cloud video: what you’re really choosing between

The comparison

The bandwidth question decides a lot of this

Cloud video means every camera uploads continuously, all day. That is the constraint people underestimate, because most internet connections are built assuming you download far more than you upload.

A handful of cameras at moderate resolution can consume a meaningful share of a small business connection’s upload capacity. Add more cameras, or raise the resolution, and you reach a point where the cameras are competing with your staff’s video calls and your cloud accounting software.

Before choosing cloud, check your actual upload speed, not the number on the marketing material, and ask the vendor for the upload requirement per camera at the resolution and frame rate being proposed. If the answer is vague, that is informative.

Two things soften this. Most cloud systems record locally at the camera or a small appliance and upload continuously in the background or only on events, which lowers the sustained requirement considerably. And upload requirements have come down as compression has improved. Ask which model applies to the system you are being quoted.

If your site is on a connection where upload is genuinely constrained, such as a rural property on Starlink, local recording is close to mandatory.

The stolen recorder problem

This is the argument for cloud that is hardest to answer.

A break-in at a small business is often not sophisticated, and an obvious black box with blinking lights in a cupboard is a recognisable thing to take. If somebody takes the recorder, they have taken the evidence of themselves taking it.

There are partial answers on the local side. Put the recorder somewhere not obvious and not in the same room as the entry point. Lock the cupboard. Configure the system to also push event clips offsite, which many local recorders can do. None of those is as complete as footage already being somewhere else.

Cloud removes the problem entirely. That is a real benefit, and for a site that has already been broken into, it usually settles the decision.

The multi-site question

If you have one building, local recording is straightforward and cloud’s main advantage does not apply.

If you have two or more sites, cloud changes the experience substantially. One login, all cameras, one search across everything. The alternative is a separate connection to each site’s recorder, each with its own login, which nobody does regularly and which means nobody is really watching.

Smart search across sites is genuinely useful here. Finding a white van across three locations in a two-hour window is a cloud strength, and it is the sort of thing analytics does reliably as opposed to the claims that are oversold.

The total cost over time

Do this arithmetic before deciding, because it is usually the deciding factor and it is rarely presented.

Take the monthly per-camera fee, multiply by your camera count, multiply by twelve, then by seven years. Compare that to the difference in up-front cost between the two quotes.

For a small camera count, the numbers can be close and cloud’s convenience wins easily. For a larger installation, the seven-year cloud total is often a large multiple of the local hardware cost. Neither result is wrong. What is wrong is not knowing which situation you are in.

One honest point in cloud’s favour on cost: the local option is not actually zero ongoing. Disks fail, typically within a few years of continuous writing. A recorder eventually needs replacing. Firmware needs updating. And somebody needs to check periodically that all the cameras are still recording, because cameras stop recording quietly. Budget for that or the local option is cheaper than it turns out to be.

The vendor lock-in question

Worth thinking about for both, in different ways.

With cloud, your footage and your system depend on a company continuing to exist and continuing to price reasonably. Cameras are often proprietary to the platform, so leaving means replacing hardware. Ask what happens to your footage if you cancel, and how you export it.

With local, you own it, but you own the obsolescence too. A recorder from a vendor that has since abandoned the product line still works and will never be updated again.

The middle path exists: standards-based cameras recording to a local system, with the option to add a cloud layer later. It costs a little more in hardware and keeps your options open. For a business unsure which way to go, it is often the sensible answer.

The short answer

One site, decent camera count, good physical security for the recorder, tight budget: local NVR. Put it somewhere hidden, and configure event clips to push offsite so a theft does not take everything.

Multiple sites, or a site that has already been broken into, or nobody available to maintain hardware: cloud, having first checked your upload capacity.

A clinic or professional office in Whitby with a handful of cameras: either works. Cloud usually wins on convenience because nobody there wants to think about disk health.

A warehouse in Courtice with twenty cameras: run the seven-year arithmetic before choosing cloud. It is often a bigger number than expected.

Whichever you pick: the resolution, lighting, framing and retention decisions matter far more than where the footage is stored. A camera system is only worth what its footage is worth.