Website Downtime Cost: How Much Does an Outage Cost?

Website downtime costs most mid-size and large companies more than $300,000 per hour, and the average across all organizations is often put near $5,600 per minute. Your own number depends on revenue per hour, the size of your team pulled into the incident, and the customers who leave and do not come back. The headline figures are useful for scale, but the cost that matters is the one you can calculate for your own site. This guide shows the current benchmarks, the costs people forget to count, and how to work out your real exposure.

How much does website downtime cost?

The most cited industry numbers come from ITIC's 2024 downtime survey and a long-standing Gartner baseline. They measure large organizations, so read them as an upper range rather than a figure for a small store, but the direction is clear: an hour offline is expensive, and it has grown more expensive over time.

One honest caveat on the Gartner number: the $5,600-per-minute figure traces back to a 2014 study, so treat it as a rough, dated benchmark rather than a fresh 2026 measurement. It survives because it is a convenient order-of-magnitude estimate, not because it was re-measured last quarter. The ITIC data is more recent and shows the range has widened, with the largest firms now reporting seven-figure hourly losses.

The costs people forget to count

Lost sales are the obvious line item, and for an ecommerce site or a paid API they are real: no checkout, no revenue, for the length of the outage. But the direct sales hit is usually the smaller part of the total. The costs that quietly add up are harder to see on the day.

• Engineering time. An incident pulls several of your most expensive people off their roadmap to firefight, then to write the postmortem. That time is gone whether or not a customer noticed.
• Support load. Every minute you are down, tickets and calls pile up, and someone has to answer them for hours after service returns.
• Customer churn. A first-time visitor who hits a dead site rarely comes back to try again. For subscription products, a visible outage is a reason a wavering customer finally cancels.
• SLA credits. If you sell an uptime guarantee, downtime can trigger contractual credits or refunds that come straight off the top line.
• Reputation and trust. Outages that show up on social media or a news site cost you deals you will never trace back to the incident.

Add these up and the true cost of an hour offline is usually a multiple of the raw lost-sales figure, which is why the enterprise numbers look so large.

How to calculate your own downtime cost

You do not need a survey to size your exposure. A simple, defensible estimate uses four inputs.

1. Revenue per hour. Take your monthly revenue that depends on the site being up, and divide by the hours you actually trade. A store doing $180,000 a month, open around the clock, earns about $250 an hour on average and far more at peak.
2. Peak multiplier. Downtime rarely waits for a quiet Tuesday. Multiply your average by two to five for the busy hours when an outage is most likely to land.
3. Response cost. Add the fully loaded hourly cost of everyone who gets pulled into an incident, times the hours it takes to detect, fix, and clean up.
4. Churn factor. Estimate the share of affected visitors who do not return, and multiply by their lifetime value.

Run those numbers once and you will usually find that the price of monitoring is a rounding error against the cost of a single missed outage. That is the whole economic case for watching your site closely: detection is cheap, and undetected downtime is not.

Detection time is the number you can actually control

You cannot prevent every outage, but you can control how long one runs before a human knows. The gap between an outage starting and your team being paged is pure, avoidable cost, and it is set almost entirely by two things: how often you check, and how fast the alert reaches someone.

A five-minute check interval means you can be down for nearly five minutes before the first failed check even records, then longer before anyone acts. A one-minute interval from multiple regions, paired with an alert that reaches on-call in seconds, cuts that dead time to a fraction. This is exactly what external uptime monitoring is for: it watches your site the way a customer would, from outside your own network, and routes the alert to Slack, email, SMS, or a webhook the moment something breaks. For teams that carry the pager, monitoring built for on-call turns a silent multi-hour outage into a notification you act on before the first support ticket lands.

Reducing the outages you cause yourself

A large share of downtime is self-inflicted, triggered by a bad deploy, a failed migration, or a config change pushed at the wrong time. Two habits cut that category sharply. First, deploy in a way that does not take the site down to ship: teams that move to zero-downtime deployments remove a whole class of avoidable outages. Second, set a realistic uptime target and measure against it, so you know when your reliability is slipping before customers do. If you are not sure what target to aim for, our guide to a good uptime percentage walks through what each level of nines actually allows in downtime per month.

What to do next

Downtime cost is not really about the scary headline number. It is about the fact that every minute of undetected outage is money leaving the building, and that detection is the cheapest insurance you can buy. Work out your own revenue per hour, add the response and churn costs on top, and compare that to the price of fast, multi-region monitoring with a branded status page your customers can check during an incident. For almost every business that sells something online, the math is not close.