CapEx to OpEx: How to Calculate Your True Surveillance Storage Cost

Moving surveillance storage to the cloud is not a straightforward decision because of the many conflicting priorities involved. Sacrificing on-prem control for lower costs and more predictable billing is a big move for workloads currently hosted fully on prem. And when data is moved to a hybrid cloud environment or to a large hyperscaler solution, the costs can be high, yet the benefits and credits thrown at you from large providers make the underlying value hard to surmise in a distinct TCO statement.

Every time someone pulls footage, the egress fees, API call charges, and retrieval costs add up. By month two, the bill looks nothing like the original estimate.

The reality is that hyperscale billing models weren't designed for surveillance workloads. Instead of eliminating cost volatility, the move from CapEx to OpEx, on prem for the cloud, traded one form of it for another.

If you're evaluating that transition, currently mid-migration, or reconsidering a contract you're already in, the math is more complicated than most vendors let on. And that gap between projected cost and actual cost doesn't announce itself in advance.

Why did the switch to cloud storage make sense?

The CapEx model has real limitations: it's expensive, inflexible, and the costs don't stop at purchase. Buying storage hardware means a large, upfront capital hit, a depreciation schedule, and full ownership of maintenance, failure, and refresh cycles. When a drive fails or a system ages out, the cost lands on your team.

Cloud storage allocates that spend to OpEx, a monthly bill based on what you use rather than what you own. Finance teams prefer this because it improves cash flow and frees up capital that would otherwise be tied up in depreciating hardware. The vendor owns the infrastructure, the maintenance, and the refresh burden. In theory, you scale up or down based on actual need rather than forecasting three years of capacity and buying for peak.

The challenge is that OpEx savings assume the monthly bill is predictable and scales linearly with usage. Hyperscale billing doesn't behave that way; for surveillance workloads specifically, the difference between that assumption and reality is where budgets get into trouble.

Why is the hyperscale billing model a problem for surveillance?

Hyperscale cloud storage wasn't designed with surveillance in mind. The pricing models behind AWS, Azure, and Google Cloud were built around the assumption that your data sits mostly at rest (you write a lot, you read occasionally). Cold and archival tiers with hyperscalers are cheap precisely because providers expect infrequent data access.

Surveillance doesn't work that way. Footage gets pulled constantly for many reasons, including investigations, legal holds, audits, incident reviews, and insurance claims. Your security operations team doesn’t access data once a quarter; they access it daily. Every time they pull footage, the bill grows in ways the original cloud storage estimate didn't reflect.

What does pay-as-you-go cloud storage actually cost?

Storage pricing is only one line item of many. On most hyperscale platforms, you're also paying for:

• Egress: Every time footage leaves the cloud, like to an investigator's workstation, a VMS client, or an external reviewer, you're paying for that data transfer. For surveillance teams pulling footage regularly, egress alone can dwarf the base storage cost.

• API calls: Every time your system creates, modifies, requests, lists, or retrieves an object, that's a billable API call. VMS platforms generate these constantly in the background, often without operators realizing it.

• Retrieval: Cold and archival tiers charge a separate per-gigabyte fee to access stored data. The lower the storage tier, the higher the retrieval cost, and the longer you wait for the data.

That headline price you saw was only part of the bill. The rest shows up when your team actually uses the system. According to the 2026 Wasabi Global Cloud Storage Index, these fees and related charges account for 50% of the average organization's cloud storage bill. Not overhead. Not edge cases. Half the invoice.

How do compliance mandates affect your storage decision?

Tightening mandates across CJIS, HIPAA, FERPA, and the Clery Act have made retention and access requirements more specific and the consequences of getting storage decisions wrong more significant. That pressure is accelerating cloud adoption before many teams have had time to properly evaluate the cost model.

The problem is that those same mandates also make the decision of cloud storage provider harder to reverse. Once you've configured retention policies around a specific provider's architecture, set the retention windows, mapped the access controls, and integrated the VMS, switching becomes a migration project. That means getting cloud storage wrong is a budget problem you may end up managing for years.

How much surveillance data are you actually storing?

Higher-resolution cameras, longer retention mandates, and analytics workloads have driven up data volumes fast, and most teams are working from storage estimates that are already out of date.

The variables that determine your actual footprint go beyond camera count: resolution, frames per second, hours of recording per day, retention period, compression type, and video quality all factor in. Change any one of them and the numbers shift significantly.

What should surveillance storage actually deliver?

A well-structured surveillance storage setup should deliver all five of these factors. None of them should be considered optional:

• Predictable cost. Flat-rate or close to it. If you can't forecast next month's bill, the model isn't working for you.

• No egress or retrieval fees. This one directly affects operational workflows, not just your budget. If your team hesitates to pull footage because of cost, they can’t do their jobs.

• Native VMS compatibility. The storage layer should be invisible to investigators and operators. If it requires workarounds, it will also eventually require troubleshooting.

• Configurable retention controls. Make sure it’s audit-ready, aligned to your applicable mandates, and adjustable without a support ticket.

• Capacity that scales without a procurement cycle. You were trying to get away from hardware refresh cycles. Your cloud storage shouldn't just recreate that in a different form.

Secure by default. Immutability, robust IAM functionality, and protection for stringent long-term mandates is table stakes for surveillance footage to protect against accidental and malicious deletion attempts.

How do you calculate your true surveillance storage cost?

Before you finalize your next cloud storage contract, work through these five steps:

1. Audit your actual storage footprint. Look at camera count, resolution specs, and retention requirements. Pull real numbers from your current VMS to see how you’re using it now.

2. Map your true OpEx cost. Add up monthly storage charge, plus egress fees, API call fees, and retrieval costs. If you don't have line-item visibility into all four, you don't really know what you’ll be charged. Get the actual invoice, not the estimate from the provider.

3. Run a 12- and 36-month TCO projection. Include projected data growth from camera upgrades, analytics expansion, and any retention extensions you're likely to face. The compounding effect of volume growth plus per-access fees adds up fast over three years.

4. Verify VMS compatibility before migration. Native integration means a smooth transition. Workarounds add operational overhead and create failure points you'll have to address eventually.

5. Plan for scale. AI workloads, higher-resolution feeds, and analytics pipelines will expand your storage requirements. Build a projection before you commit to a pricing model.

What did your cloud storage estimate leave out?

The math has always been there. Most vendors just didn't show it to you.

The shift to cloud storage meant lower overhead, no refresh cycles, and infrastructure that scales without a procurement fight. What didn't get disclosed upfront was how the billing model behaves once a surveillance team actually starts using it.

Egress fees, API charges, retrieval costs: none of these are hidden in the fine print by accident. They're just not what vendors lead with. And for workloads that generate and access data the way surveillance does, they're not incidental.

The CapEx versus OpEx decision still matters. So does the move to cloud. The part that gets skipped is whether the specific billing model you're signing maps to the workload you're actually running. That's the question worth asking before you commit.