Innovation in Action Means Proving It in the Field

The show floor at GSX 2026 in Atlanta was full of new capabilities: AI analytics, drones, automated monitoring, new approaches to access control. Yet the most useful conversations of the week had less to do with what the technology could do and more to do with whether it would keep working once it was installed.

Security Info Watch captured the tone in its show coverage. Axon's Mike Shore summed up the customer's question as "how do I deploy all this stuff and get it operational?" Motorola Solutions' Danny Gagnon observed that "detection is being commoditized," and Everon CEO Don Young noted that customers prioritize "consistent and reliable service" over technical sophistication (Bosch, 2026). SDM's editor heard a similar theme: most companies are weighing AI for its economic impact rather than adopting it for its own sake (Hodgson, 2026).

For ISVs, that shift raises a practical question about what innovation looks like to the buyer. Our view is this: when features start to look alike, what sets a solution apart is proof that it keeps performing after it ships, and that proof depends on seeing the system long after the ISV has usually lost sight of it.

When Features Converge, Performance Becomes the Product

Every strong detection feature eventually gets matched. Models improve, competitors catch up, and what once set a platform apart becomes the expected baseline. The Security Industry Association's 2026 Security Megatrends point the same way, with trends such as "The Security Hardware Layer Is Reinvented" and "End-to-End Solutions and One-Logo Approaches" (Security Industry Association, 2025).

As features converge, buyers compare something harder to copy: how the complete system behaves over months and years. Does the analytics platform still hit its accuracy target after a winter of glare and weather? Does the recording server keep up after camera counts double? Those answers depend on the whole stack, and the customer credits or blames the software vendor for all of it.

As we explored last month, ISVs often own the promise without controlling the platform expected to fulfill it. That gap tends to widen once the system goes live, because the ISV's visibility usually ends at the ship date.

The Proof Gap

We call this the proof gap: the distance between what was validated before deployment and what the customer's system is actually doing today.

Validation answers whether a configuration performs under expected workloads. It can't answer what happens after a firmware update, a network change, a camera knocked out of alignment, or a storage array filling faster than planned. In video environments, many of these problems are quiet. An image slowly loses focus. A server runs hotter under a heavier analytics load. A switch starts dropping frames at peak hours. Nothing fails outright, so nothing raises an alarm.

The analytics keep running on degraded input, and accuracy drifts. Often the first signal the ISV receives is a customer complaint, and by then the conversation has turned to who is at fault. Support teams start by reconstructing a history they never saw. That is where resolution time stretches, and GSX coverage noted that organizations increasingly measure success through metrics like mean time to resolution (Bosch, 2026).

What an Invisible Problem Costs

One engagement we described in February shows how expensive the proof gap can become. BCD supported a large ISV on a high-visibility project for a global Fortune 50 customer. A new build used GPUs that had been validated elsewhere, though never at the density this deployment required. After rollout, degraded performance appeared only at higher GPU density, caused by untested interactions between the ISV's software, GPU drivers, and supporting toolkits.

Resolving it took nearly two months. Approximately $40 million in revenue and forecasted opportunities were affected, and relationships were strained across the customer, the integrator, the ISV, and BCD.

In February, we treated this as a validation lesson, and earlier testing in the BCD Innovation Lab would likely have caught the issue. There is a second lesson. Once the problem reached the field, each week spent diagnosing it extended the strain across what we described in July as the trust chain. The longer a problem stays hard to see and isolate, the more of that chain it reaches. Closing the proof gap at both ends, before deployment and after it, limits how far a problem can travel.

Validation Before, Visibility After

That thinking sits behind BCD's partnership with EyeOTmonitor, announced ahead of ISC West in March. EyeOTmonitor is a cloud-based platform that discovers networked devices and builds an interactive map of cameras, servers, switches, and VMS. Paired with BCD's validated hardware platforms, it monitors server performance and resource use, camera health and connectivity, network performance and bandwidth dependencies, and VMS integration. Its ImageAssure analytics look for blur, obstruction, glare, and scene drift, the slow image problems that rarely trigger an alert on their own (BCD, 2026).

"Purpose-built infrastructure is the foundation of any high-performance video environment, but performance has to be provable, not assumed," said Daniel Gewargis, Chief Technology Officer at BCD. "EyeOTmonitor gives our customers the visibility to continuously validate system performance and protect their investment."

For an ISV, the value shows up in the conversation it changes. When a customer reports that analytics accuracy has slipped, the team can begin with a record of what the hardware, network, and cameras have been doing. Engineering sees whether the platform is still operating inside the conditions it was validated for. Support can separate a software issue from an infrastructure one early in the diagnostic process. Sales can point to performance evidence at renewal time.

EyeOTmonitor is the first of several partnerships designed to extend what a validated platform can prove over its life, with more to announce in the months ahead.

Isn't Monitoring the Customer's Job?

A fair objection is that monitoring belongs to the end user or the integrator. Many customers already run network monitoring, and ISVs have good reason to avoid owning responsibilities outside their software.

That division works when the tools can see what matters to the application. General IT monitoring reports whether a device is online. It rarely reports whether the video feeding an analytics model is still good enough to produce accurate results, or whether a server has drifted outside the conditions it was validated for. Those are the signals tied to the ISV's promise, and they are the ones most likely to go unwatched.

Taking on that visibility has costs of its own: more data, decisions about who acts on an alert, and clear agreements with integrators about roles. The ISVs handling this well tend to treat post-deployment visibility as part of the platform they offer, with defined boundaries, so the question of who watches the system is settled before the first complaint arrives.

Innovation That Holds Up

GSX showed an industry with plenty of new capability that is growing more selective about what earns a place in production. That selectivity favors solutions that can show their performance over time, on the hardware and in the conditions customers actually run.

At BCD, we see innovation in action as the pairing of validated platforms before deployment with continuous visibility after it, so an ISV's promise can be checked against evidence at every stage of the system's life.

Let's Compare Notes

If your team is thinking about how to keep sight of systems after they ship, or where the proof gap sits in your own solution, we're glad to talk it through.

Hardware. Handled.

References

BCD. (2026, March 11). BCD and EyeOTmonitor announce strategic partnership at ISC West 2026. https://www.bcdvideo.com/news/bcd-and-eyeotmonitor-announce-strategic-partnership-at-isc-west-2026/

Bosch, R. (2026, September 16). GSX 2026 conversations put practical value ahead of technology hype. Security Info Watch. https://www.securityinfowatch.com/security-executives/article/55405445/gsx-2026-conversations-put-practical-value-ahead-of-technology-hype

Hodgson, K. (2026, September 21). GSX 2026 impressions. SDM Magazine. https://www.sdmmag.com/articles/105828-gsx-2026-impressions

Security Industry Association. (2025, December 18). Security Industry Association announces the 2026 Security Megatrends. https://www.securityindustry.org/2025/12/18/security-industry-association-announces-the-2026-security-megatrends/

The show floor at GSX 2026 in Atlanta was full of new capabilities: AI analytics, drones, automated monitoring, new approaches to access control. Yet the most useful conversations of the week had less to do with what the technology could do and more to do with whether it would keep working once it was installed.

Security Info Watch captured the tone in its show coverage. Axon’s Mike Shore summed up the customer’s question as “how do I deploy all this stuff and get it operational?” Motorola Solutions’ Danny Gagnon observed that “detection is being commoditized,” and Everon CEO Don Young noted that customers prioritize “consistent and reliable service” over technical sophistication (Bosch, 2026). SDM’s editor heard a similar theme: most companies are weighing AI for its economic impact rather than adopting it for its own sake (Hodgson, 2026).

For ISVs, that shift raises a practical question about what innovation looks like to the buyer. Our view is this: when features start to look alike, what sets a solution apart is proof that it keeps performing after it ships, and that proof depends on seeing the system long after the ISV has usually lost sight of it.

When Features Converge, Performance Becomes the Product

Every strong detection feature eventually gets matched. Models improve, competitors catch up, and what once set a platform apart becomes the expected baseline. The Security Industry Association’s 2026 Security Megatrends point the same way, with trends such as “The Security Hardware Layer Is Reinvented” and “End-to-End Solutions and One-Logo Approaches” (Security Industry Association, 2025).

As features converge, buyers compare something harder to copy: how the complete system behaves over months and years. Does the analytics platform still hit its accuracy target after a winter of glare and weather? Does the recording server keep up after camera counts double? Those answers depend on the whole stack, and the customer credits or blames the software vendor for all of it.

As we explored last month, ISVs often own the promise without controlling the platform expected to fulfill it. That gap tends to widen once the system goes live, because the ISV’s visibility usually ends at the ship date.

The Proof Gap

We call this the proof gap: the distance between what was validated before deployment and what the customer’s system is actually doing today.

Validation answers whether a configuration performs under expected workloads. It can’t answer what happens after a firmware update, a network change, a camera knocked out of alignment, or a storage array filling faster than planned. In video environments, many of these problems are quiet. An image slowly loses focus. A server runs hotter under a heavier analytics load. A switch starts dropping frames at peak hours. Nothing fails outright, so nothing raises an alarm.

The analytics keep running on degraded input, and accuracy drifts. Often the first signal the ISV receives is a customer complaint, and by then the conversation has turned to who is at fault. Support teams start by reconstructing a history they never saw. That is where resolution time stretches, and GSX coverage noted that organizations increasingly measure success through metrics like mean time to resolution (Bosch, 2026).

What an Invisible Problem Costs

One engagement we described in February shows how expensive the proof gap can become. BCD supported a large ISV on a high-visibility project for a global Fortune 50 customer. A new build used GPUs that had been validated elsewhere, though never at the density this deployment required. After rollout, degraded performance appeared only at higher GPU density, caused by untested interactions between the ISV’s software, GPU drivers, and supporting toolkits.

Resolving it took nearly two months. Approximately $40 million in revenue and forecasted opportunities were affected, and relationships were strained across the customer, the integrator, the ISV, and BCD.

In February, we treated this as a validation lesson, and earlier testing in the BCD Innovation Lab would likely have caught the issue. There is a second lesson. Once the problem reached the field, each week spent diagnosing it extended the strain across what we described in July as the trust chain. The longer a problem stays hard to see and isolate, the more of that chain it reaches. Closing the proof gap at both ends, before deployment and after it, limits how far a problem can travel.

Validation Before, Visibility After

That thinking sits behind BCD’s partnership with EyeOTmonitor, announced ahead of ISC West in March. EyeOTmonitor is a cloud-based platform that discovers networked devices and builds an interactive map of cameras, servers, switches, and VMS. Paired with BCD’s validated hardware platforms, it monitors server performance and resource use, camera health and connectivity, network performance and bandwidth dependencies, and VMS integration. Its ImageAssure analytics look for blur, obstruction, glare, and scene drift, the slow image problems that rarely trigger an alert on their own (BCD, 2026).

“Purpose-built infrastructure is the foundation of any high-performance video environment, but performance has to be provable, not assumed,” said Daniel Gewargis, Chief Technology Officer at BCD. “EyeOTmonitor gives our customers the visibility to continuously validate system performance and protect their investment.”

For an ISV, the value shows up in the conversation it changes. When a customer reports that analytics accuracy has slipped, the team can begin with a record of what the hardware, network, and cameras have been doing. Engineering sees whether the platform is still operating inside the conditions it was validated for. Support can separate a software issue from an infrastructure one early in the diagnostic process. Sales can point to performance evidence at renewal time.

EyeOTmonitor is the first of several partnerships designed to extend what a validated platform can prove over its life, with more to announce in the months ahead.

Isn’t Monitoring the Customer’s Job?

A fair objection is that monitoring belongs to the end user or the integrator. Many customers already run network monitoring, and ISVs have good reason to avoid owning responsibilities outside their software.

That division works when the tools can see what matters to the application. General IT monitoring reports whether a device is online. It rarely reports whether the video feeding an analytics model is still good enough to produce accurate results, or whether a server has drifted outside the conditions it was validated for. Those are the signals tied to the ISV’s promise, and they are the ones most likely to go unwatched.

Taking on that visibility has costs of its own: more data, decisions about who acts on an alert, and clear agreements with integrators about roles. The ISVs handling this well tend to treat post-deployment visibility as part of the platform they offer, with defined boundaries, so the question of who watches the system is settled before the first complaint arrives.

Innovation That Holds Up

GSX showed an industry with plenty of new capability that is growing more selective about what earns a place in production. That selectivity favors solutions that can show their performance over time, on the hardware and in the conditions customers actually run.

At BCD, we see innovation in action as the pairing of validated platforms before deployment with continuous visibility after it, so an ISV’s promise can be checked against evidence at every stage of the system’s life.

Let’s Compare Notes

If your team is thinking about how to keep sight of systems after they ship, or where the proof gap sits in your own solution, we’re glad to talk it through.

Hardware. Handled.

References

BCD. (2026, March 11). BCD and EyeOTmonitor announce strategic partnership at ISC West 2026. https://www.bcdvideo.com/news/bcd-and-eyeotmonitor-announce-strategic-partnership-at-isc-west-2026/

Bosch, R. (2026, September 16). GSX 2026 conversations put practical value ahead of technology hype. Security Info Watch. https://www.securityinfowatch.com/security-executives/article/55405445/gsx-2026-conversations-put-practical-value-ahead-of-technology-hype

Hodgson, K. (2026, September 21). GSX 2026 impressions. SDM Magazine. https://www.sdmmag.com/articles/105828-gsx-2026-impressions

Security Industry Association. (2025, December 18). Security Industry Association announces the 2026 Security Megatrends. https://www.securityindustry.org/2025/12/18/security-industry-association-announces-the-2026-security-megatrends/

• SHARE: