When systems go down, the costs don't arrive as one line item. They leak out of the business in a dozen directions at once: payroll spent on people who can't work, deals that stall, deadlines that slip, customers who quietly try a competitor, and the overtime spent catching up afterward.
Because the cost is scattered, most leadership teams underestimate it — and then underinvest in the unglamorous work that prevents it. A simple estimate changes that conversation.
A back-of-the-envelope model
Start with labor. Take your fully loaded payroll for the people affected by the outage and prorate it per hour. If 50 people averaging $35/hour fully loaded lose most of their productivity for an hour, that's over $1,500 in payroll alone — before anything else goes wrong.
Then layer in what's true for your business: revenue that depends on systems being up (orders, billable time, closings, appointments), deadline exposure (filing dates, ship dates, court dates), and recovery drag — the hours spent re-entering data and untangling the backlog after systems return.
Finally, consider frequency. An environment that loses an hour here and an afternoon there, monthly, quietly costs far more per year than a single dramatic outage — it just never triggers the meeting.
What this means for prevention
Once you have even a rough hourly figure, decisions get easier. Redundant internet for a fraction of one outage's cost stops being a luxury. Proactive monitoring and maintenance — the core of managed IT — stops looking like overhead and starts looking like insurance with a measurable deductible.
The goal isn't zero risk; that's not purchasable. The goal is matching your investment in reliability to what unreliability actually costs you — a number most businesses have never written down.