What research firms say downtime actually costs.

Before you build your own model, it helps to know the reference points the industry already uses — independently surveyed, not vendor-marketing numbers.

average cost of downtime IT downtime statistics downtime cost per minute downtime cost benchmarks by industry
Scroll for the benchmarks
$5,600 Per-minute average, Gartner (2014 baseline, still widely cited)
97% Of large enterprises: one hour costs over $100K (ITIC, 2024)

The numbers everyone quotes, with their actual source.

These figures get repeated constantly without attribution. Here's where each one actually comes from.

$5,600 / minute Gartner's widely cited 2014 baseline average cost of downtime — roughly $336,000 per hour, across all company sizes and industries. It's the single most-quoted downtime statistic in the industry, and it's over a decade old.
$2,300–$9,000 / minute The range reported the same year by Avaya, reflecting how much company size and industry vertical move the average in either direction — a reminder that any single-number average hides very wide variation.
~$9,000 / minute Ponemon Institute's 2016 update to the same style of average, suggesting downtime costs had risen materially in the two years after Gartner's figure.
97% over $100K / hour ITIC's 2024 survey of over 1,000 firms found 97% of large enterprises (1,000+ employees) report that a single hour of downtime costs their company more than $100,000.
41% over $1M / hour The same ITIC survey found 41% of enterprises report hourly downtime costs between $1 million and over $5 million — the tail end of the distribution that a single average obscures entirely.

External benchmarks are a starting point, not your number.

Because published surveys report broad averages, this site's industry calculators use their own illustrative starting defaults instead — calibrated to the qualitative risk profile of each sector, not a specific survey. Use them as a sharper starting point than a single global average, then replace every input with your own data.

Downtime benchmarks, answered.

Questions that come up when citing an industry statistic instead of a company-specific number.

Why is the most-cited statistic over a decade old? Gartner's 2014 figure was widely republished and became the de facto industry reference, even though more recent surveys like ITIC's now report materially higher costs — always check the publication date behind any downtime statistic you cite.
Should I use a benchmark instead of building my own model? Use a benchmark to sanity-check your own number, not to replace it — a global average can't reflect your specific revenue-per-hour, incident frequency, or SLA exposure the way a few minutes with the calculator can.
Why do estimates vary so much between sources? Different surveys sample different company sizes and industries, and "cost of downtime" itself gets defined inconsistently — some include only direct revenue loss, others fold in productivity, compliance, and reputational cost.
Where do real incident cost figures come from? See the case studies section for sourced, company-specific cost figures from seven real outages, as an alternative to survey averages.

Skip the average. Model your own number.

Enter your own systems, revenue, and incident history for an estimate specific to your business.

Mode

Accent