Downtime cost calculator

The simple version is revenue per hour multiplied by hours of outage, scaled by the share of your service that was affected. This calculator adds the cost of staff time spent responding, which is easy to forget and often substantial. It divides annual revenue by 8,765.82 — the hours in an average year — to get an hourly figure.

Quite a lot, and all of it makes outages more expensive: customers who churn, support tickets, SLA credits owed, reputational damage, and the engineering time lost to context-switching rather than incident response itself. Treat the result as a conservative floor.

It is a reasonable approximation for businesses with steady traffic, and a poor one for businesses with sharp peaks. If most of your revenue lands in a few hours a week, an outage during a peak costs far more than this average suggests, and one at 4am costs far less.

It shortens the outage. Cost scales with duration, and the largest controllable part of duration is usually time-to-detection — the gap between something breaking and anyone knowing. Monitoring does not prevent failures; it stops them running for six hours unnoticed.