How to Calculate Video Surveillance ROIs

ROI calculations are powerful but can be distorted. While they hold the promise of identifying objective value, they can often obscure the truth.

The goal of this review is to help the security manager better understand supplier ROI calculations and allow the manager to modify or adjust for accurate and realistic results. Integrators and manufacturers could also benefit from applying these principles.

Good ROI calculation require understanding operational details more than they do math or money. Once you understand the operational details, the math and money are simple.

Here are the 4 principles in preparing a ROI calculation:

• Understand the alternative to this proposed investment
• Understand the full cost
• Understand the technological deficiencies of this investment
• Verify that operational assumptions are correct

The most basic trick to play in ROI analysis is to choose an alternative that is clearly bad but not relevant to your case. Most vendor ROIs do this. One topical example is with NVRs. Frequently, NVRs make claims that they drive ROI by enabling centralized monitoring or integrating with applications like POS or access control. While certainly true, from an ROI perspective, this is irrelevant because DVRs do the same things. It does not make sense for a security manager to compare an NVR to a VCR or to nothing because almost everyone has a DVR or would consider a DVR as an alternative to a NVR. To make a business case for the NVR, it needs to be compared to a DVR.

For instance, if an NVR cost "$10,000" and a DVR cost "$8,000", the investment for purpose of calculating the ROI would be $2,000 (the premium for the NVR over the DVR). At the same time, the NVR could only claim returns on abilities that it uniquely has over the DVR, thereby eliminating from consideration aspects such as centralized monitoring and application integration. If you do not take this approach and simply calculate an ROI of an NVR versus a VCR, you could be wasting money by paying extra for a NVR when a DVR could have delivered the same value.

NOTE: I think NVRs often generate more value than DVRs so this is not a criticism of NVRs. This is a critique of the process often used to justify NVR purchasing decisions.

Megapixel camera suppliers often advocate camera elimination but this can sometimes distort ROI calculations. For instance, a recent whitepaper examined a scenario where 13 analog cameras could be replaced by (2) 3 Megapixel cameras for covering a 100 foot wide outdoor area. The paper concluded that the megapixel camera solution was actually cheaper. This assumption is misleading because the alternative here is really using 2 or 3 analog cameras. That is what most security managers use today and with that as the alternative the cost of the megapixel camera scenario is significantly higher than analog cameras.

NOTE: The megapixel cameras in this scenario may deliver much higher ROI by being able to solve previously unsolvable cases due to their greater quality. I am not objecting to the design, simply the method on how the financial justification was being made.

The security manager and megapixel vendor should concentrate on demonstrating the increased return delivered specifically by the enhanced image quality. Specifically, only cases solved with a megapixel camera that could not be solved by an alternative analog camera should be factored in the ROI for megapixel cameras. If identifying a license plate was critical in solving a a case, the megapixel camera should get credit for it. But if the case could be solved by identifying that the car was a white Civic, an analog camera would be equally capable and the megapixel camera should not get credit.

This distinction is routinely blurred but if you are to truly determine an accurate ROI, this is a critical factor.

Principle #2: Understand the full cost

Often, vendor supplied ROIs leave out indirect costs. These become hidden costs that can drag your true ROI down significantly.

One of the hidden costs of video analytics is the need for monitoring. Depending on the level of false alerts, you may need to dedicate resources to assess and verify the alerts. This cost could become quite significant. You may be able to get the technology to work as advertise but you may need to dedicate extra operational resources to bring it to that level. Make sure you understand what if any indirect costs are needed and factor this in.

Megapixel cameras are another example of indirect costs. With megapixel cameras, it is not only the increased camera cost but the increased cost of the storage and bandwidth. Almost all megapixel cameras in production use much more inefficient compression than analog cameras. Also, if you truly want enhanced resolution in megapixel cameras, this will further increase storage costs (and often network costs).

Again, these both may be justifiable but a fair analysis most include any additional cost for them.

Principle #3: Technological Deficiencies

When a vendor provides you an ROI, usually it assumes that the technology works as advertised. With new technology that sometimes turns out not to be the case. Also, sometimes, the technology works but not in the circumstances you need it in.

This is one of the key issues with video analytics. It is easy to say that perimeter violation has the potential to reduce losses significantly. However, it depends on how well it works. If it turns out that your facilities have a lot of snow, the system may not work properly during those times. This can reduce the potential loss reduction projected. Similarly, you may want to use a megapixel camera to capture license plates and faces in a very dark area at night. Many megapixel cameras work poorly with low light conditions. If you were projecting to solve cases during this time, this may not actually work.

Similarly, the system may turn out to be too hard to use so that your operators fail to solve as many cases as the technology might potentially deliver.

Carefully review what the vendor's projections are and make sure that any technological deficiencies are reflected in the ROI calculation.

Principle #4: Operational Assumptions

Suppliers can only make best guesses as to the operational realities of a security manager. Often those guesses are very optimistic or simply do not match your organization's situation. Examples of these assumptions include loss per incident, number of incidents per month, number of incidences that this system will solve.

First, you need to ask and understand what these operational assumptions are in a vendor provided ROI. Compare that to your actual metrics and re-adjust to determine appropriate levels. How much time does the system really save you? How many incidents per year can you really solve with the new system that you could not with old?

It's probably going to differ from the vendor assumptions, so be ready to adjust the ROI calculations.

The challenge in all financial models is the assumptions made. By using these 4 principles, you can better assess and determine the right assumptions to make. Identify hidden costs and problems that a theoretical ROI may ignore and keep your suppliers honest.

Untangle common ROI confusions and distortions and you will be rewarded with an accurate ROI providing clarity on genuine business value.