Downtime gets argued about in percentages, which nobody has an instinct for. Put your own numbers in and it turns into money and minutes instead.
168 if you sell around the clock.
Optional — leave at 0 to ignore.
The same percentages, converted into time you'd be offline — and what that time costs at the rate above.
| Uptime | Downtime / year | / month | / day | Cost / year |
|---|---|---|---|---|
| 99% | 3d 15h 36m | 7h 18m | 14m 24s | $24,066 |
| 99.5% | 1d 19h 48m | 3h 39m | 7m 12s | $12,033 |
| 99.9% | 8h 45m | 43m 48s | 1m 26s | $2,407 |
| 99.95% | 4h 22m | 21m 53s | 43.2s | $1,203 |
| 99.99% | 52m 33s | 4m 22s | 8.6s | $241 |
| 99.999% | 5m 15s | 26.3s | 864ms | $24 |
Cost per year assumes the full allowance is used and that downtime falls during trading hours. Nothing here is sent to us — the numbers stay in your browser.
99.9% uptime reads like a rounding error away from perfect. It allows nearly nine hours offline a year — and because outages arrive all at once rather than in polite two-minute slices, that whole allowance can be spent in a single afternoon.
The other half people miss is when. An hour down at 4am costs almost nothing; the same hour on your busiest morning can cost more than the rest of the month. Averages are a floor, not a forecast.
Revenue is spread across the hours you actually trade, not all 168 in a week — a shop open 40 hours loses roughly four times more per hour than a flat weekly average would suggest. If you add blocked staff, their cost is added on top, because an outage bills you twice: once in sales you didn't make and once in people who can't work.
What it doesn't try to price is the part that shows up later — refunds, support load, and customers who quietly don't come back. Treat the number as conservative.