Start With the Business Activity Affected
An outage’s cost depends on when it happens, which activity stops, what customers do later and what recovery actually costs. Start with sales expected during that window, or use a clearly stated monthly-average fallback. Record the source of each assumption, then rerun low, base and high scenarios.
Subscription revenue does not automatically disappear in proportion to outage hours. For SaaS or internal systems, identify the actual affected sales, credits, penalties and additional recovery costs. Enter zero where there is no supported loss. Use the 3-2-1 backup guide to distinguish having a backup from being able to restore the service.
Formulas and Assumptions
Method reviewed 15 October 2026. All amounts are EUR, excluding tax and financing. The input values are planning assumptions supplied by the reader, not established industry averages.
- Expected hourly sales = your outage-window hourly estimate, or monthly affected sales / relevant hours in that month. The monthly fallback assumes uniform demand across those hours; its default 720 hours describes a 30-day month.
- Sales exposed = expected hourly sales × outage duration.
- Sales not recovered later = sales exposed × affected share × (1 − recovered share).
- Lost contribution = sales not recovered later × contribution margin.
- Incident cost = lost contribution + incremental recovery spending + credits or penalties + other distinct direct costs.
- Annual scenario cost = incident cost × expected incidents per year.
Contribution margin subtracts the costs avoided when a sale is lost. Do not add those avoided costs back as damage. Recovery spending should cover additional expense, such as an emergency contractor or paid overtime; avoid counting ordinary payroll or a cost already included elsewhere twice.
Allocated productivity exposure = affected staff × your loaded hourly cost × outage duration × productivity reduction. It is displayed separately and is not added to incident cost. Reduced staff output, payroll and lost sales can describe overlapping effects.
The optional investment comparison uses your proposed outage duration and frequency, keeping the other revenue and per-incident cost inputs unchanged:
- Avoided annual cost = baseline annual scenario cost − proposed residual annual scenario cost.
- Net annual benefit = avoided annual cost − additional annual investment.
- ROI = net annual benefit / positive additional annual investment × 100. With zero added investment, ROI is undefined.
The proposed scenario can still incur incidents. A negative net benefit or ROI is a valid result. The calculation does not automatically add churn, lifetime value or a reputation surcharge, or assume that buying a service prevents all losses. It does not model correlated incidents, changing demand, multi-year discounting or uncertainty distributions.
Worked Example With Every Input
This is an illustrative scenario, not an observed typical business:
Sales exposed are 50,000 / 720 × 2 = €138.8889. After 25% are recovered, sales not recovered are €104.1667. At 40% contribution margin, lost contribution is €41.6667. Adding €200 recovery spending gives €241.67 per incident and €966.67 per year before rounding intermediate values.
Allocated staff productivity is €175 per incident, shown separately. In the proposed scenario, annual cost is €441.67; avoided annual cost is €525.00. Subtracting the €24,000 investment gives −€23,475.00 net annual benefit, or −97.81% ROI. These modeled avoided costs do not justify that investment on their own.
Load the worked example fills all of these inputs into the form. Calculations retain precision internally and round only for display. The PDF report includes the inputs, assumptions, results and method link.
Sources and How They Are Used
These sources support the method; they do not prescribe this example’s sales, staff rate, incident frequency or improvement assumptions. Use your analytics, incident records, contracts and actual employment or supplier costs for those values.
For operational next steps, see what to automate in IT operations and how to evaluate server management companies.