Free calculator
What does a broken page cost before anyone notices?
Minutes broken × your spend rate × how often it happens. No counterfactual revenue in it and no model behind it — move the sliders and the page does the multiplication in front of you.
A year of breaks, noticed in 4h
$8,333
6 × 240 min × $5.79
The same year, caught here
$521
6 × 15 min × $5.79
Stays in the account
$7,813
94% of it
At $250,000 a month, 100% of the account bills at $5.79 a minute while it is broken. Two of the assumptions in that pull against each other and we would rather name both than call the answer conservative: dividing a month evenly by its minutes understates it, because almost nobody spends evenly at four in the morning and the daytime rate is the one billing while a page is down; assuming a break takes out every campaign you run overstates it — INC-2291 was 3 campaigns, not an account. Move that last slider until it matches what one break actually reaches; the sum is only worth as much as that number is. Nothing here estimates revenue you would have earned: every dollar in it bought nothing by definition, because a destination that cannot convert converts nobody, which is why it lands on acquisition cost as pure numerator.
The one figure we supply
Fifteen minutes, and where it comes from
Everything on the left of that calculation is yours. The only number we put in is the right-hand column, and it is the real one: 10 minutes to detect — two checks five minutes apart, because one failure is a blip and two in a row is an outage — plus 5 for a person to act on an alert that arrives with the action already written and priced.
For the worked version on a real incident: INC-2291 ran for 213 minutes at $58.59 a minute and cost $12,480. Detection took 6 minutes. The other 207 were a person asleep, which is the part a faster dashboard does not fix.
Common questions
Costing downtime, answered plainly
How do you calculate the cost of downtime for paid advertising?
Multiply your spend rate by the minutes the destination was broken, then by how often that happens in a year. The spend rate is your monthly spend divided by the minutes in a month (43,200). There is deliberately no estimate of lost revenue in this: spend that reaches a page which cannot convert bought nobody by definition, so it is counted as pure waste rather than as a guess about sales that might have happened.
Why does this calculator not include lost revenue?
Because lost revenue is a counterfactual and waste is a fact. We know exactly what left the account while the page was broken; we would have to invent a conversion rate to claim what it would have earned. A number a reader can check beats a bigger number they cannot.
Is dividing monthly spend evenly across the month accurate?
No, and two things pull against each other, so it is worth naming both rather than only the flattering one. Dividing evenly UNDERSTATES it: almost nobody spends at a flat rate at four in the morning, and the daytime rate — the one billing while a page is broken during business hours — is higher than the average. But the sum also assumes a break takes out everything you run, which OVERSTATES it, and by more: INC-2291 was three campaigns, not an account. That is what the share control is for. Set it to what one break actually reaches and the answer is neither a floor nor a ceiling but arithmetic.
How fast can a broken landing page realistically be caught?
Two checks five minutes apart, so about 10 minutes to raise it, plus roughly 5 more for somebody to act on an alert that arrives with the action already written — 15 minutes in total. Requiring two consecutive failures matters: one failed check is a blip, and alerting on blips is how a team learns to mute the channel.
A calculator is a strange thing for a monitoring company to lead with, since it tells some readers their number is too small to bother about. That is the point: the same arithmetic that makes this worth buying above a certain spend says plainly that below it you should not, and the pricing page runs it in that direction too. If the figure on the right looks large, the next question is not what it costs to fix — it is whether the page your ads point at is broken right now, which takes about three seconds to find out.