The conversation I keep having
I’ve had some version of the following conversation with IT directors and CIOs at least once a month for the past year.
The video surveillance system is aging out. The current platform is 7-10 years old, the cameras are reaching end-of-life, and the recording infrastructure is showing its age. An upgrade is coming, and the question is which direction to go.
The cloud vendors are calling. The pitch is consistent: subscription pricing, no on-site infrastructure to maintain, software updates handled automatically, modern user interfaces, AI analytics built in, mobile access, IT burden dramatically reduced. The pitch is genuinely compelling. Many of these platforms are genuinely good products.
But the IT director isn’t ready. The reasons vary. Sometimes it’s bandwidth — the cumulative upload of high-resolution video from dozens or hundreds of cameras strains the internet connection beyond what current circuits can handle, and the upgrade to fiber that would solve it is its own multi-year project. Sometimes it’s data governance — the organization has regulatory, legal, or policy concerns about storing surveillance footage with a third party. Sometimes it’s cost — the subscription math at scale across many cameras runs higher than expected over the contract horizon. Sometimes it’s just operational caution — the organization isn’t ready to depend on the vendor’s cloud infrastructure for life-safety-adjacent functions.
So the IT director defaults to upgrading the on-prem system. Traditional NVR, traditional cameras, traditional architecture. The cloud conversation gets deferred.
Three or four years from now, when the organization is ready for cloud — when bandwidth is solved, governance frameworks are in place, costs are understood — the IT director is going to face the same upgrade decision again. Except now the recently-purchased cameras are sunk capital. The migration to cloud means buying new cameras to replace the ones that were just installed.
This is the trap that’s underwritten a lot of stranded capital in the video surveillance industry over the past few years.
The third option
Here’s what most IT directors don’t know about the camera market right now: not every camera is locked to a specific recording platform. Some camera lines are designed to work with both on-prem NVR infrastructure and cloud-managed platforms, with a migration path that doesn’t require replacing the camera hardware.
Avigilon’s product line is the example I work with most often. Avigilon Unity is the on-prem video management platform — traditional NVR-based architecture, on-site recording, deep enterprise features, the kind of system many IT directors and security directors are familiar with. Avigilon Alta is the cloud-managed platform — subscription-based, hosted infrastructure, modern interface, integrated AI analytics.
The strategic part of Avigilon’s product architecture is that many of the cameras can operate with either platform. A camera deployed today recording to a Unity NVR can later be moved to the Alta cloud platform without replacing the hardware. The transition involves firmware and configuration changes, not a forklift upgrade of the camera fleet.
For an IT director who wants modern cameras now but isn’t ready for cloud, this changes the procurement decision fundamentally. You can deploy on-prem today with cameras you’ll be able to use in your cloud migration whenever it happens. The decision about when to go to cloud becomes a software and operational decision, not a hardware procurement decision.
Why this matters in practice
Let me make the case concretely with the kind of math IT directors actually do.
A mid-sized organization deploying 100 cameras today is making a capital decision in the range of hundreds of thousands of dollars between the cameras, the recording infrastructure, and the installation. If those cameras are platform-locked to on-prem, the eventual cloud migration means a substantial portion of that capital becomes stranded — the cameras either get replaced or get repurposed at lower-priority sites where they age out without ever contributing to the cloud architecture.
If those same 100 cameras can move to cloud without being replaced, the migration cost is dramatically lower. You’re paying for the cloud platform subscription, the configuration work, and probably some incremental network and storage adjustments — but the camera fleet itself carries over. The capital invested today is protected.
For organizations doing this math across hundreds or thousands of cameras across multiple sites, the difference between stranded-capital cloud migration and carry-over cloud migration is material. It can be the difference between cloud being financially feasible and cloud being a phased rollout over many years to amortize the hardware replacement cost.
What to look for in cameras that give you this flexibility
If you’re an IT director or security director scoping a camera procurement and you want to preserve the option to move to cloud later without replacing hardware, here are the specific things to evaluate before signing the purchase order.
Does the camera support multiple recording platforms from the same manufacturer? Some manufacturers segment their on-prem and cloud product lines such that the cameras are designed for one or the other. Others — Avigilon being the example I work with most — design cameras that can operate with either the on-prem VMS or the cloud platform from the same manufacturer.
Is the migration path documented and supported, or is it a workaround? There’s a meaningful difference between a vendor that officially supports moving cameras between their on-prem and cloud platforms, and a vendor where it’s technically possible but unsupported. The supported path means firmware updates, configuration tools, and vendor backing if something goes wrong. The unsupported path means hoping nothing breaks.
Does the camera support ONVIF and standards-based protocols broadly? Even if you’re planning to stay within a single vendor’s ecosystem, ONVIF compliance and RTSP support give you exit options if the vendor’s roadmap changes or you decide to switch platforms later. Cameras that only talk to their own manufacturer’s proprietary protocols constrain your options.
What’s the licensing model on the cloud platform when you migrate? Some vendors require new licenses when cameras move from on-prem to cloud, which can change the financial picture of the migration. Others handle the license transition cleanly. This is worth confirming in writing before the original purchase.
What’s the network and infrastructure migration path? Moving cameras to cloud isn’t just a camera-side decision. The network architecture, the storage configuration, and the integration with other security systems all change. A vendor whose products are designed for the on-prem-to-cloud migration has thought through these pieces; a vendor whose migration is an afterthought has not.
Is this unique to Avigilon?
Honest answer: this kind of cross-platform flexibility isn’t only available from Avigilon, but the depth of integration between Unity and Alta is unusually strong because both platforms come from the same manufacturer with the migration path explicitly designed in.
Other vendors handle the question differently. Some camera manufacturers — Axis, Hanwha, i-PRO, Bosch — produce ONVIF-compliant cameras that can work with many different recording platforms, including various cloud VMS platforms. Cameras from these manufacturers offer real flexibility, though the cloud migration involves moving to a different vendor’s cloud platform rather than a same-vendor transition.
Some manufacturers — Verkada, Rhombus, and certain other cloud-native platforms — produce cameras that are explicitly designed for their own proprietary cloud and don’t operate at all on-prem alternatives. These are committed-to-cloud platforms by design, and the trade-off is that you are stuck with this platform, or you have to scrap the whole system!
The Avigilon Unity-to-Alta path is distinctive because it’s a same-vendor migration with the engineering done on both sides of the transition. That’s worth weighing when evaluating which platform to deploy.
Worth being clear: I’m not arguing Avigilon is the right choice for every organization. I’m arguing that the flexibility to deploy on-prem today and move to cloud later, without replacing hardware, is a real capability worth understanding before you make a 7-10 year camera infrastructure decision.
When this matters most
The on-prem-today, cloud-later option is most valuable for organizations in specific situations:
Organizations facing imminent camera end-of-life who aren’t ready for cloud but don’t want to be locked into on-prem forever. The deploy-on-prem-now, migrate-later path is exactly designed for this scenario.
Organizations with data governance, compliance, or legal concerns about cloud video that may evolve over time. Healthcare organizations, financial services, government, education — sectors where the cloud conversation is genuinely complicated and the answer in five years may be different than the answer today.
Organizations with bandwidth constraints that are being addressed but not yet solved. Fiber expansion, network upgrades, multi-site connectivity improvements — projects that will eventually make cloud video feasible but aren’t done yet.
Multi-site organizations doing phased upgrades. The ability to deploy on-prem at sites being upgraded now, with the flexibility to bring those cameras into a cloud architecture later, supports a coherent multi-year rollout rather than locking each site into whatever architecture was current when it got upgraded.
Organizations doing the financial math on a 7-10 year horizon. Protecting the camera capital across architecture changes is a meaningful TCO consideration that doesn’t always show up in vendor pricing comparisons.
What to ask before procurement
If you’re an IT director, security director, or facility leader scoping a camera procurement in the next 6-12 months, these are the specific questions worth asking — of vendors, of integrators, of yourself:
What’s our realistic cloud video timeline? Not the marketing answer. The actual organizational, technical, financial, and political timeline for moving from on-prem to cloud video. Three years? Five? Seven? Never? The honest answer determines whether the flexibility to migrate matters or doesn’t.
What’s the cost of the camera fleet we’re about to buy, and what does it look like if it has to be replaced when we eventually move to cloud? Run the math both ways — assuming the cameras carry over, and assuming they have to be replaced. The gap between those two numbers is the value of the flexibility.
Which vendors offer real same-vendor on-prem-to-cloud migration paths? Avigilon is one. There may be others. The list is shorter than the cloud video vendor list because most cloud-native platforms aren’t designed to support an on-prem deployment first.
What’s the vendor’s roadmap on both their on-prem and cloud platforms? A vendor with strong investment in both platforms is more likely to support the migration path long-term. A vendor whose on-prem product is in maintenance mode while cloud gets all the investment is signaling where they want customers to end up.
How does this integrate with our broader security stack? Access control, intrusion, fire, mass notification. The camera platform doesn’t sit in isolation, and the integration story has to work in both the current on-prem deployment and the eventual cloud deployment.
How we approach this conversation
Technology Install Partners works with organizations across multiple verticals on video surveillance design, including both on-prem and cloud deployments. The Avigilon Unity-to-Alta migration path is one of the architectures we deploy most often when clients want modern cameras today but aren’t ready to commit to cloud yet.
When we engage with a client on a camera procurement, the conversation isn’t just about which cameras and how many. It’s about what their realistic cloud trajectory looks like, what flexibility they want to preserve, and which architecture protects their capital investment across the changes that are inevitable over a 7-10 year camera lifecycle.
Sometimes the right answer is full cloud now — for organizations where cloud is the explicit destination and the trade-offs are acceptable. Sometimes the right answer is on-prem with a migration path preserved — for organizations that aren’t ready but don’t want to be locked in. Sometimes it’s a hybrid where some sites go cloud and others stay on-prem, with the camera architecture chosen to support either at each site.
If you’re an IT director, security director, or facility leader facing a camera procurement decision and the cloud question is part of what’s making it hard, send a message. The right architecture genuinely depends on the specifics of your organization, and the conversation is worth having before the next procurement decision gets locked in for the next decade.
