Every day, enterprise cameras and sensors capture enormous volumes of information about what is happening in the physical world. Much of that data serves an immediate purpose, such as identifying a security incident or helping protect people and property. But its usefulness doesn’t necessarily end there.
The same video that helps a manufacturer monitor safety can also reveal recurring bottlenecks on a production floor. In retail, cameras deployed for loss prevention can provide insights into customer traffic, store layouts, staffing needs, and product interactions. Across industries, organizations are finding new ways to use existing video infrastructure to better understand how their physical environments operate.
That shift is already underway. The Axis Perspectives Report 2026 found that the share of organizations using video systems for business intelligence nearly doubled between 2024 and 2025, increasing from 20% to 38%.
But the opportunity becomes more complex as these applications grow. Extracting insights from a few cameras at a single location is one thing; doing so across hundreds or thousands of devices, multiple sites, and different business functions is another. Unlocking that value at scale requires organizations to determine not only what their video data can tell them, but where and how that data should be processed, analyzed, and managed.
The advantage of the hybrid edge
Organizations should use a “hybrid edge” approach to powering their video-based business intelligence efforts, says Patrik Pettersson, a strategic adviser at Axis Communications. In this model, analytics at the network edge power real-time detection, processing, and analysis while public cloud platforms perform deeper large-scale pattern analysis. Furthermore, devices and their integration with the cloud platform require harmony, where the edge devices must be capable of working and being managed independently without requiring the cloud connection. This hybrid model reduces bandwidth, improves latency, and fits CIOs’ strategies for cost predictability.
Pettersson advises organizations to work with vendors whose devices are designed for secure and reliable cloud connectivity, offer open platforms for maximum flexibility, and have an established track record with edge-based analytics.
“The smarter the device at the edge and the more it can do, the more it will alleviate costs in the cloud,” Pettersson says. “The harmony between the edge and the cloud is critical for economic cloud scaling for vision intelligence.”
Scaling business intelligence
Leveraging existing camera infrastructure for business intelligence and operational efficiency requires close collaboration between security leaders and the business, Pettersson says. Also, cameras that were originally deployed to meet specific safety and security objectives cannot always support new use cases.
Security directors sometimes initially object to using cameras for business intelligence and operational efficiency, Pettersson notes, but he adds that they often relent once they understand that the relationship can be symbiotic. In turn, business leaders often need to yield to security leaders when leveraging physical security assets.
In some cases, business units may need to invest in dedicated systems or additional devices to achieve the outcomes they’re aiming for without compromising core security needs. “When each party’s interests are met and trust is established, the security director will likely be more open to including their devices in business applications,” Pettersson says.
To start, Pettersson says, organizations should conduct small-scale trials with only a few cameras. By first focusing on a single business department, they can better determine the potential ROI. “Proofs of concept and pilots are a great way to test capabilities in controlled settings and then gradually scale,” he says. “You don’t need to boil the ocean.”
It may take some time to find the right mix of on-premises, edge, and public cloud resources, Pettersson notes. “Often you can start with extremely powerful cloud compute tools that you already know will be cost-prohibitive at scale,” he says. “From there, you tune down and simplify until you get to a balance between cost, accuracy, and performance.”
Finding that balance will become increasingly important as organizations identify new uses for video data. A successful pilot may involve only a handful of cameras and a single business function. Expanding that application across locations, departments, and thousands of devices introduces very different demands on infrastructure and resources.
That makes scalability as much an architectural consideration as a technical one. Processing every piece of data in the cloud may not be practical or economical, just as relying exclusively on the edge can limit opportunities for broader analysis. Organizations need flexibility to determine where processing makes the most sense based on the use case, cost, performance, and insights they hope to gain.
Ultimately, the strategic value of video data isn’t determined by how much of it an organization collects. It comes from the ability to turn that data into useful intelligence and to build an infrastructure capable of doing so as those opportunities grow.