There are two ways to buy a video system. The first is to buy an ecosystem: cameras and recorder from one manufacturer, sold as a matched set, supported as a matched set, priced as a matched set. It works, it is usually cheaper on day one, and it is what most small commercial sites end up with.
The second is to buy an open-platform video management system — software written to record and manage cameras from many manufacturers, sold separately from the cameras themselves. Milestone is the archetype of this category; it is not the only member of it, and this article is about the category rather than any product in it.
The case for the open platform is genuinely strong, and we make it to clients regularly. But it is made badly almost every time, because it is made as though the freedom were free. It is not. The costs are real, and they are structural rather than line items — which is exactly why they do not appear on the quote you are comparing.
Strip away the marketing and an open platform buys one thing: the ability to choose a camera on its merits, and to keep choosing.
That sounds abstract until you are standing in year six. A closed ecosystem couples two decisions that have nothing to do with each other — which cameras suit your site, and which software you run them on — and every later decision inherits the weld. The camera you need for a difficult loading bay has to exist in that manufacturer's catalogue. The recorder that dies has to be replaced by the same brand or the cameras stop being manageable. A price rise, a licensing change, or a manufacturer you would rather not buy from any more all arrive as an all-or-nothing replacement, at that manufacturer's price, in a conversation where the leverage is entirely theirs.
An open platform separates those decisions. Cameras from different manufacturers can sit on the same system, the right camera can be bought for one difficult location alone, and the exit — changing part of the estate without replacing all of it — stays open. That is the real argument, and it is the strongest one against a closed ecosystem. Notice that it is not a features argument. It is a leverage argument, and leverage is worth most to the buyer who expects to be operating this system for a long time.
Cost one: licensing that grows with the system
Open-platform video management software is not usually sold the way you buy an appliance — once, for the box. It is generally licensed per camera, and offered in tiers, with the larger tiers carrying what larger sites need. We are not printing anyone's numbers here; prices move, and any figure published today misleads you next year. The shape is what matters: the cost of the software scales with the number of cameras, and again with the tier.
Two consequences follow, and they are the ones that surprise owners.
The first is that adding a camera is a software purchase, not just a hardware purchase. In a closed ecosystem, adding a camera to a recorder with spare channels often costs you the camera and the labour. On a licensed platform, it costs the camera, the labour, and a licence. That is not a scandal — it is how the model works, and it is what pays for the software being maintained — but it belongs in the arithmetic before you buy, not on the second phase.
The second is the tier boundary. Feature tiers exist so that a small site is not paying for what a campus needs. It also means a feature you assume is present — a specific integration, a redundancy behaviour, a management capability across multiple servers — may live one tier up. The failure mode is not being overcharged. It is designing around a capability, and finding at commissioning that the capability was in a tier nobody quoted.
There is usually a maintenance or support agreement attached to the licences as well, which is what entitles you to new versions. Decide deliberately whether to carry it, at purchase, and know what it costs annually — the version-compatibility problem below is what that money is buying.
Ask for this in writing: the licence cost per camera at your intended tier, what the next tier up adds, what one additional camera licence costs a year from now, and what the annual maintenance agreement costs and what happens if it lapses.
Cost two: somebody owns a server
This is the cost that gets left off the quote most often, because it is not something anyone sells you.
An open-platform VMS is software, and software runs on a server — a real computer in your building or in a data centre you rent, not a sealed appliance. It has an operating system that receives security patches, a maintenance window in which those patches get applied, storage that fills, drives that wear out, and a hardware lifespan after which it must be replaced. It also needs backing up in a way that includes its configuration, not just its video, because the configuration is the part that takes weeks to rebuild.
So the question that decides whether an open platform suits you is rarely technical. It is: who does that job, and are they named?
In a business with existing IT capability — an internal person, or a managed service provider who already patches your servers — the answer is easy and the cost is marginal. In a business without one, this is a new operating responsibility that did not exist before, and it does not go away. Either it becomes a line item in a support agreement with your integrator, or it becomes nobody's job, which is the outcome we see most often. An unpatched general-purpose server is a bigger target than a sealed recorder, and video systems are among the most commonly exploited devices on a business network. Cloud vs on-premise video recording walks the same maintenance question through both models; the open-platform decision sits on top of that one, not instead of it.
Ask for this in writing: who applies operating-system patches to the video server and how often, what the server replacement horizon is, and whether the system configuration is backed up separately from the footage.
Cost three: version compatibility becomes a thing you manage
In a closed ecosystem, compatibility is the manufacturer's problem: they ship the recorder and the cameras, test them together, and the update that arrives has been through that testing.
On an open platform you have at least three moving parts that version independently — the video management software, the firmware on cameras from several manufacturers, and any analytics or integrations bolted alongside — each with its own release schedule and support lifecycle. Most of the time this is uneventful. Occasionally it is not: a camera firmware update changes a behaviour the VMS relied on, a VMS upgrade drops support for an older device generation, or an integration you depend on lags a major version behind and pins the whole system in place.
None of that is an argument against open platforms. It is an argument for treating upgrades as a planned activity with a compatibility check in front of it, rather than a button someone presses on a Friday. Vendors publish supported-device and compatibility information; the discipline is reading it before upgrading rather than after something stops recording.
Ask for this in writing: which party is responsible for testing an upgrade before it is applied, and what the rollback plan is if a camera stops recording after one.
Cost four: "supports ONVIF" is a baseline, not parity
This is the misunderstanding that does the most quiet damage to open-platform projects.
Interoperability between a camera and a VMS from different manufacturers usually runs over ONVIF, which is real, works, and is the reason the open-platform model exists at all. But a conformance claim guarantees a baseline of interoperability — a defined set of behaviours, described by a profile — and a baseline is not the same as every feature of that camera being available in that software.
The gap is usually not in the video. It is in everything around the video: a manufacturer's own analytics classifications, a specialty imaging mode, event types only that manufacturer generates, configuration only that manufacturer's own tool exposes. Those can arrive in the VMS through a manufacturer-specific driver rather than through ONVIF — and whether that driver exists, for your camera model, on your VMS version, is a question with a factual answer that somebody should get before you buy the camera.
The full version of this argument — including why a conformance claim is a manufacturer's own declaration, and why the profile letter is the part that carries meaning — is in what ONVIF profiles actually mean when buying cameras. Read that one first if you have not. The extension here is commercial rather than technical: on an open platform, the ONVIF baseline protects your ability to keep recording from a camera whose manufacturer you have fallen out with. It is a floor under the estate, not a promise that the camera you were shown in a demo will behave identically under different software.
Ask for this in writing: for each camera model, whether the features you were shown are available through the VMS's manufacturer-specific driver or through ONVIF alone, and what specifically is lost if it is ONVIF alone.
Four conditions. The more you have, the stronger the case:
Multiple sites. One management layer over several buildings is where the platform model earns its licence cost, and where per-site recorder sprawl becomes a real operational burden.
A mixed estate you did not choose. A building acquired with someone else's cameras in it, three inherited generations of equipment, one location whose environment demands a specialty camera nobody else makes. An open platform absorbs a mixed estate; a closed one asks you to normalize it, at your expense.
A long horizon. You own the building, or the lease runs long, and you expect this system to outlive the people who installed it. A long horizon is what guarantees at least one forced decision — an end-of-support, a price change, a manufacturer you no longer wish to buy from — and that is when the exit option pays.
Existing IT capability. Someone already patches servers here. If you have this, cost two is nearly free. If you do not, it is the whole decision.
Who is better served by something simpler
If you run one building, with a modest camera count, a stable layout, and nobody whose job includes server maintenance, an open platform will likely cost you more and give back less than a well-chosen matched system with a support agreement. The freedom to mix manufacturers is worth little to a site that will never mix manufacturers, the exit option is worth little on a horizon shorter than the equipment's life, and a general-purpose server nobody owns is worse than a sealed appliance nobody owns.
That is not an argument for lock-in — it is an argument for buying the lock-in knowingly, and for spending the effort where it pays at your scale: sensible camera selection, cabling that will carry the next generation, retention decided on purpose, and someone named as responsible for keeping the system alive.
The honest test is one question: what happens to this system in year six? If the answer involves multiple sites, mixed equipment, or a manufacturer relationship you are not certain will survive that long, buy the platform and budget its real costs. If the answer is "the same shop replaces it," a simpler system, specified carefully and maintained properly, is the better purchase — and the difference can go into cameras and cabling that will still be right when the software changes.
For the related decisions this one touches, see what a unified platform actually unifies and edge, server, or cloud analytics.
