- The average Global 2000 company faces a $15,000 cost per minute after an incident or outage, Splunk study finds
- Customers are often the first to notice an incident, causing major reputational damage
- With many falsely identifying attacks as IT issues, greater observability is needed
New data from Splunk has claimed unplanned downtime now costs Global 2000 companies around $600 billion every year, which marks a 50% increase over the past two years.
Splunk reported the average G2000 company faces a per-minute cost of $15,000 when an outage occurs, which translates to an average annual revenue loss of $95 million.
But the costs extend far beyond just revenue, with the average firm seeing a 3.4% drop in stock prices. Regulatory fines also average the not-so-insignificant sum of $51 million, the company revealed.
The hidden costs of downtime
Severe cyberattacks continue to rise, with high-profile incidents like those of M&S and Jaguar Land Rover in 2025 dominating the headlines, but it's not just the frequency that's rising. It's also costs, with the average ransomware payout nearly tripling since 2024 to $40 million.
One of the more unquantifiable outcomes is a loss in brand reputation, with half (47%) of tech leaders revealing that customers are among the first to notice service disruptions. Four in five (81%) believe this results in customer loss.
Then there's the human resources needed to rectify issues – one in five marketers say it takes them an entire quarter to get back to their previous state.
Time to resolution is another issue, with a third (36%) of security leaders reporting that downtime is often wrongly attributed to an IT issue rather than a security breach, severely slowing identification and remediation times.
"Downtime is inevitable," SVP and GM Kamal Hathi said, but "prolonged disruption is not."
Hathi believes that "align[ing] technology with business outcomes, empower[ing] people with context, and design[ing] systems that bend, but do not break, under pressure" often the best results, indicating a greater need for observability and context.
