Everyone knows downtime is costly, but what is the real financial impact? While some organisations may have insight into the impact should a system become unavailable the true costs may be higher than expected.
The recent The Hidden Costs of Downtime report released by Splunk paints a sobering picture. The research, that captured responses from 2,000 executives from Global 2000 companies across 20 countries from APAC, EMEA, North America, and LATAM, found that the average cost of downtime, when the full impact was accounted for, is USD$15,000 per minute.
"It's not just simply the cost of the event," said Marc Caltabiano, the Group Vice President for Australia and New Zealand at Splunk. "It's all the issues around the revenue impact, regulatory fines and ransomware payments. But it's also the human resource cost, the consultancy costs that come afterwards during remediation and the reputational damage."
That number is staggering and growing year on year. In 2024, the lost revenue cost was estimated at USD$49M. In 2026, the figure has almost doubled to $USD$95M.
Caltabiano said the increasing complexity of technical ecosystems is a major contributor to the increased costs.
"We no longer just have our basic systems of record and systems of action. We've got a complex set of autonomous agents. There's a real increase in both internal and external vulnerabilities and the pace at which bad actors are working. And resources are becoming increasingly hard to find."
Although cybersecurity gets a lot of attention when it comes to outages and downtime, there can be many other causes. Hardware failures, interrupted communications links and errant software agents all have significant impact. Recognising an incident and being prepared to act on it requires visibility. And while organisations understand that, Caltabiano says there are still blind spots.
"One of the things I hear from customers is that they're monitoring only some of the systems such as their core systems of record and action. But the surface area is so broad now, and particularly with AI and shadow AI, we need to consider how do you monitor a sufficient part of your estate so you pick up errors and vulnerabilities as early as possible at the right cost without having to retain data that is not as relevant."
With a sprawling technical estate, AI is a powerful enabler in Caltabiano's view. AI accelerates the pace at which we can drive adoption of many capabilities to help monitor and provide more resilience. But he added that it's important to human authority and human review.
"We're going to see more autonomous agents acting within systems. We need to ensure that there's accountability around those agents and we're maintaining resilience. The key to that is capturing the data and looking for the patterns and correlation to ensure that we can identify issues and act as quickly as possible."
Technology is changing at a pace that we haven't experienced in a very, very long time or maybe ever before and organisations are grappling with the investment associated with maintaining business operations while developing new capabilities to stay ahead of competitors. But it remains important to ensure that you're securing your most critical systems. Caltabiano said understanding the importance of those systems and how much information you need to maintain requires assessing risk based on where data sits.
"Often, when we're looking at security vulnerabilities, it's generally a human error that will trigger as big or bigger outage challenge than it would be from a surface attack by a bad actor. This is why end-to-end observability is so important," said Caltabiano.
Part of that internal threat comes from the ungoverned use of AI, or shadow AI. Countering that demands a predictive data driven operation as opposed to a response operation when you have a challenge. That includes automated triage, real-time impact metrics, and identifying whether a piece of hardware which may be slowing down is a predictor of a fault that's about to happen that will bring your environment down.
"Waiting for the phone to ring is no longer good enough," said Caltabiano.
Downtime is inevitable. The complex ecosystems that organisations rely on mean everyone will experience some downtime eventually. Having resilient operations and failover sites is one way of achieving resilience but it can come with a hefty price tag.
Caltabiano said understanding how to avoid that prolonged disruption, putting preventative measures in place and being able to proactively monitor so you can improve mean time to discovery and mean time to recovery can have a significant positive impact. Achieving that relies on a solid data foundation that gives estate-wide visibility.