Choosing a home security camera system that lasts
Model recommendations date within a year. The criteria that separate a good system from a bad one do not.
Roundups of the best camera systems go stale in about a year. Model numbers change, product lines get discontinued, and a company that was excellent can be acquired and gutted. What does not change is the set of questions that separates a system you will be happy with in five years from one you will be replacing in two.
Decide the storage model first
This is the decision everything else follows from, and it is close to irreversible once you have bought into an ecosystem.
- Cloud — footage leaves the property immediately, so it survives the camera being stolen or destroyed. Almost always subscription-based, and the subscription usually scales per camera. A third party holds your footage.
- Local — a recorder or SD card on site. No recurring cost, nobody else holding your video, and it works with the internet down. Vulnerable to the recorder being taken or damaged.
- Hybrid — records locally, pushes clips to cloud on events. Best of both, more expensive, and increasingly the sensible default.
Questions that outlast any model list
- Does it record continuously, or only on motion? Motion-only leaves gaps, and the gaps land at the start of events — exactly the part you need.
- What happens when the internet drops? A system that stops recording during an outage is a system that stops recording when someone cuts your service line.
- What happens if the company stops trading, or ends support? Cameras tied to a proprietary cloud can become inert. Anything speaking a standard protocol like ONVIF can be adopted by other software.
- Can you export footage as an ordinary video file? If retrieval requires proprietary software an insurer or the police cannot open, the footage is much less useful than it appears.
- How long is the retention, and is it configurable?
- Does it get security updates, and for how long? An unpatched internet-connected camera is a device on your home network with a bad history.
Where to spend, and where not to
Spend on the front door camera and low-light performance. Those two decisions determine whether footage is usable when it matters. Spend on retention if there is any chance of discovering a problem days later.
Do not spend on resolution beyond what the position needs, on more cameras where better placement would do, or on features that require a subscription you will resent in year two.
The one that catches people out
Ecosystem lock-in. Buying four cameras from one manufacturer means the fifth will be from them too, at their price, on their subscription. That is a fine trade if you are happy with them — but it is a decision being made at camera one, usually without anyone realising it.
Systems built on open standards cost slightly more up front and leave the choice open. For anything you expect to keep beyond a few years, that flexibility is generally worth the premium.