Can AI Reverse Security’s Declining Margins?

Repurposing and expanding video surveillance infrastructure may be the key to long-term success

Matt Powell headshot
Matt Powell is managing director for North America at ISS.

The last several years have been a bit of a roller coaster ride for the security industry. On one hand, the uncertainty created by the global spread of COVID-19 was a bit of a boon initially to the market, creating demand for solutions, such as remote video monitoring, that organizations could use to keep tabs on what were largely vacant facilities at the time. Another benefit brought about by the pandemic was that it allowed integrators to get into buildings to work without having to worry about disrupting the day-to-day activities of employees or tenants.

However, as weeks turned into months and months dragged on into years, the global supply chain issues brought about by on-again, off-again lockdowns created significant project backlogs and inventory challenges and even affected the ability of integrators to provide basic levels of service. If these impacts were not bad enough, the pandemic also significantly contributed to skyrocketing inflation rates across all manner of products, including security solutions.

In fact, according to the Security Business State of the Security Industry 2022 report, 60% of integrators surveyed reported being affected in some way by rising interest rates and inflation, with responses ranging from being unable to secure loans and funding to delaying a potential sale of the business to having to cut back on wages, raises and benefits for employees. Additionally, 46% of integrators said that every manufacturing partner they worked with raised their equipment prices in 2022, while another 34% said that 75-99% of their vendors increased prices.

These challenges have exacerbated one of the industry’s biggest challenges – declining margins on security hardware. Surveillance cameras, in particular.

The Race to the Bottom

It may seem counterintuitive, but some of the greatest innovations in video surveillance technology over the past few decades – which have simultaneously served to propel adoption across nearly every market imaginable – have now been incorporated by most manufacturers into their own product lines, thereby making cameras a commodity. The proverbial IP tipping point, in which sales of network cameras exceed analog, seems to have come and gone without anyone taking much notice.

No one really talks about IP vs. analog that much anymore because IP has become the de facto standard. Similarly, high-resolution and multisensor cameras are now widely available from a majority of brands. The meteoric rise in the 2010s of China-based video surveillance manufacturers that provided parallel or even superior functionality to many industry stalwarts also served to drive prices lower.

This combination of increased competition and a relative lack of product differentiation has resulted in what seems like ever-shrinking margins for integrators as the manufacturing community has embarked on a “race to the bottom” where the winner is whoever can provide the best camera at the cheapest possible price.

The consequences of this new business paradigm are finally coming home to roost for many companies that thought they could ride the hardware gravy train over the long term. The aforementioned State of the Security Industry report, for example, found that while nearly three quarters of integrators reported revenues being “significantly” or “slightly” up in 2022 compared to 2021, less than 60% of respondents polled said that their gross profit margins increased similarly over the same time period.  

So, rather than acting in the best interest of their clients and building a solution that is truly “best-of-breed,” a significant number of integrators have resorted to discount shopping among their vendor and distribution partners as a way to squeeze an extra point or two of margin out of their project pipeline.

 This is not only dangerously myopic for integrators as they look to build long-lasting relationships with end users, but it has also incentivized vendors to look into ways of bypassing the channel to sell direct, as they too have experienced a significant shift in the profitability of their businesses. Many integrators have decided to hedge their bets against this ongoing trend by exploring new technologies and service offerings that will enable them to charge a premium above and beyond simple hardware sales.

Living on the Edge

One area of heavy emphasis within the video surveillance market for a while now has been the development of cheaper, more powerful processors embedded within the cameras themselves as a way to reduce the infrastructure footprint. If all of the image processing and prerequisite capabilities needed for a deployment can be done inside the camera, then the need for things like specialized servers, cabling and other products that quickly add up on a video install should be greatly reduced. Or so the thinking goes.

The problem with this notion is that edge processing is currently not at the point where it can consistently and reliably deliver the same performance that onsite solutions do, nor will it have the capability to do so anytime soon. The ability to process high-resolution images alongside other data-intensive applications like video analytics inside of cameras is still years away from becoming a realistic possibility. Therefore, any potential cost savings a business would realize by leveraging a surveillance network with edge-based processing would quickly be wiped out by the performance issues and the inability of operators to get the data they need.

In fact, during an event at ISC West 2023, security technologist and SIA Public Safety Working Group Chair Steve Surfaro cautioned against the market looking to the edge as a way to run artificial intelligence (AI), analytics and other tools.

“Right now, most solution providers in the industry are looking to do edge AI. Absolutely ridiculous right now,” Surfaro said. “You should be focusing on offloading the CPU cycles off of your VMS platform and using an AI processor.”

The “as-a-Service” Trap

Another popular way the industry has looked to offset declining margins is through the use of video surveillance as a service, access control as a service and other hosted/managed offerings. This enables businesses to swap large, upfront capital expenditures for a monthly operational cost, similar to the residential market where dealers provide alarm systems and sensors for free in exchange for creating a recurring monthly revenue (RMR) generating account.

This has largely been painted as a win-win for both integrators and end users. RMR enables integrators to reap the benefits of steady and reliable income streams, as opposed to having to rely on one-off projects with long sales cycles that are becoming increasingly affected by a wide range of economic factors. In addition to not having to invest in video or other security hardware, end users can also rest easy from a maintenance perspective, knowing that all of their updates will be done automatically by the service provider.

With what seem like so many benefits for all involved, where’s the rub? The issues are actually myriad for both integrators and end users. With regard to the integrator, they are largely outsourcing the customer relationship to a third-party provider – be it a cloud solution vendor or a video monitoring firm – who can then take the customer direct or possibly harm the integrator’s reputation should they provide a poor level of service. And the end user, since they do not own the equipment outright, is forced to use whatever hardware has been specified for the project and cannot scale with the same level of flexibility that comes with system ownership.

AI and Video’s Third Wave

Yet another area that integrators are turning to as a way to increase their profit margins is through the adoption of AI-powered video analytics, which hold enormous potential for businesses even beyond traditional security applications and use cases. Looking at video surveillance innovation in terms of waves, with the first being analog cameras and DVRs and the second being IP cameras and video management software, the industry appears to now be in the third wave, with AI-driven software algorithms and neural network training methodologies. While the first analytics that hit the market in the mid-2000s largely overpromised and underdelivered, today’s Al solutions are starting to pay dividends as a force multiplier for end users and a moneymaker for integrators.

Not only can analytics be tailored to help organizations mitigate specific types of risks depending on their operating environment, they also represent a potential windfall for the security industry as a whole. The 2023 SDM 100, for instance, quoted several integrators who see significant opportunities for their businesses in providing AI analytics to the market.

“The market was generally strong, but not much different than 2021,” according to one alarm company. “CCTV was strongest and analytics were key.”

Another alarm company reported, “In its commercial and national account business units, [the company] experienced growth in most verticals in 2022 as employers made efforts to bring employees back to the office safely. These employers were also seeking new ways to efficiently manage multi-site locations and overcome labor shortages, evidenced by new investments into AI and cloud-based systems.”

A majority of integrators (56%) who took part in the State of the Security Industry survey reported having a strong interest in adding AI-based technologies to their offerings. And though some think of AI as only being a video surveillance tool, the technology is being increasingly used in access control, since facial recognition can be leveraged as a secondary or even primary method of authentication in certain circumstances.

Despite the varied applicability, increased product differentiation, and higher profit margins that AI and video analytics bring to the table, many integrators remain fixated on simply selling more cameras as the best way to grow their business and overall footprint in the market. Perhaps unsurprisingly, the vast majority of integrators (77%) in the State of the Security Industry survey reported that IP cameras were their top performing video surveillance product offering in the previous year, with all other product categories – including video management software (45%), video storage and playback (39%) and cloud/hosted video (31%) – far behind.

The number of camera options that integrators have to choose from continues to dwindle, which will further amplify the margin crunch. According to a 2023 report from market research firm Novaira Insights, demand for both video surveillance hardware and software remains strong as the global market, excluding China, grew by more than 13.2% in 2022; however, with each passing year, the industry’s largest vendors continue to increase their portion of the video surveillance pie. The research firm noted that the combined market share of the world’s 10 largest vendors grew from 61% in 2020 to 64% in 2022.

As consolidation intensifies in the years to come, there will likely be fewer and fewer options for integrators to mark up. Resorting to bargain hunting is not a business model for long-term, sustained success, nor is investing in largely unproven technology and service offerings.

AI and the current generation of video analytics provide integrators with a key advantage over many of these other alternatives: pre-existing infrastructure. The proliferation of IP and high-definition cameras means that analytics can usually be added to a system without the need for additional hardware, providing integrators with the ability to resell them without having to invest time and manpower resources in ripping and replacing the current technology stack.

Artificial intelligence is fundamentally transforming the way companies do business and it will revolutionize how they think about and approach security and safety. The industry will be anything but “business as usual” for the foreseeable future, and AI is a big reason why.