How much does one hour of downtime really cost (with a calculator)

When a manager asks "how much did that 47-minute outage cost us?", the standard answer is "I dunno, some sales". It's the wrong answer — and it's why monitoring investment always sits at the bottom of the priority list.

You can compute it. You can put it in a spreadsheet. And the number almost always shocks people.

The base formula

Start simple:

Direct cost = (Revenue / business hour) × (Hours down) × (Impact factor)

• Revenue / business hour = monthly revenue ÷ real business hours (not 24×30).
• Impact factor = how much of the traffic in that window would actually convert. E-commerce at lunchtime on a Wednesday: 0.7 to 1.0. B2B at 4 AM: closer to 0.1.

Example: a store doing $120k/month, open 8–22, 6 days a week ≈ 336 useful hours/month = $357/hour. One hour down on a Tuesday at 3 PM (impact 0.9) = $321 of straight revenue gone.

But that's the number everyone already calculates. The real problem lives in the costs no one adds up.

The 4 invisible costs

1. Wasted CAC

You paid Google Ads, Meta Ads, an influencer, SEO. The user clicked. Landed on your domain. Saw an error. Left. You paid for the click and you don't have the lead.

Wasted CAC = (Paid sessions lost) × (Cost per session)

A store spending $6k/month on paid media with 60k paid sessions/month = $0.10 per session. One hour down at peak (say 400 paid sessions/h) = $40 set on fire. And your Meta Ads dashboard will show "campaign performing well" — because the click happened.

2. SEO penalty

Googlebot has a crawl budget. If it hits your homepage and gets 5xx, it reduces visit frequency. If the window is long enough for it to retry and get 5xx again, pages start dropping out of the index. Recovery can take weeks.

You can't estimate this exactly without historical data, but: every position lost on a high-value long-tail keyword can cost dozens to hundreds of visits, month after month.

3. Support overload

Every minute down generates tickets. Each ticket costs $2–$6 depending on channel (chat, email, phone). One hour down across a base of 10k active customers easily generates 80–200 tickets — not from the outage itself, but from side effects ("my order disappeared", "I paid and got no confirmation", "the app keeps crashing").

4. Reputation and churn

Hardest to measure, most expensive long-term. A customer who tries your product and hits an error three times in a row enters the red churn zone. In B2B SaaS, losing a customer means 12 to 36 months of subscription gone — a single bad hour can cost more than a year of pro-tier monitoring.

Simple calculator

Drop this in a spreadsheet:

For a well-placed hour of downtime (peak, business day), the real total is usually 3 to 5× the raw "lost revenue" number. That's why a serious company doesn't haggle over monitoring — the ROI already pays back in the first incident avoided.

The cost of finding out late

If the customer is the one telling you the site is down, your window is the time until the first angry tweet — on average 8–15 minutes for a retail brand, 30–60 for B2B. Multiply by the cost/hour you just calculated.

Good monitoring drops that number to ~60 seconds. That's the difference between two scenarios:

And that's the detection win alone, before any resolution time.

Where Sentinela fits

We check your site every 60 seconds from multiple regions, alert via email/webhook/Telegram in seconds, and surface MTTD and MTTR in a dashboard so you can hold your team or your provider to an SLA. We won't fix the incident for you — but the stopwatch will be ticking on your side.

For the bill on the other side, see uptime isn't availability and is free site monitoring worth it? — understanding the cost of not monitoring is the first step to sizing how much to invest.

The simple rule

Calculate it once. Write the number down. Tape it to the product team's wall. Next time someone tries to cut $40 from monitoring, you have the argument ready.