Every IT proposal has a "cost of downtime" number in it, and most of those numbers are made up. The honest calculation is harder and more useful, and once you run it, you stop thinking about backup and redundancy as IT expenses and start thinking about them as insurance with a measurable premium.
01The formula that actually works.
Per-hour downtime cost = (hourly payroll of idle staff) + (hourly revenue-generating capacity lost) + (hourly productivity-dependent client impact) + (recovery labor burden + overtime). Sum these, honestly, and you have a per-hour number you can defend.
For most professional-services firms we work with, the honest number comes out between $1,200 and $4,800 per hour of full-office downtime. For a medical practice mid-clinic-day, it can exceed $6,000. For a busy restaurant dinner rush with a POS outage, it can hit $2,500 per hour in comped meals and walk-outs.
02What most calculators leave out.
Stock calculator outputs include "lost revenue" as if every hour is fully bookable. In reality, a law firm with 20 attorneys idle for 3 hours does not lose $20,000 of billable time, they lose whatever fraction of that time was already scheduled for billable work, plus a downstream impact as deadlines compress.
What almost never gets counted: recovery labor cost (your team working overtime to get back to baseline), client goodwill (no one sues you, but next renewal is 40% harder to close), and reputational cost (the google review that mentions "down for a day").
03The insurance framing.
Once you have a per-hour number, the math on backup, DR, and redundancy changes. A cloud-backup service that adds $400/month is a $4,800/year premium that pays for itself the first time it prevents a single four-hour outage.
An HA firewall pair that costs $8,000 more than a single unit is a two-hour-of-downtime hedge. A hot-site DR replica that adds $1,200/month is insurance against a full-day office loss. All three are routinely worth it and often aren't even close.
04The "cheap" alternatives that aren't.
External USB drive as backup: fails when ransomware encrypts the drive (always attached). Nightly tape rotation: 24-hour data loss is no longer an acceptable RPO for any operation that takes payments or notes. "We'll restore from cloud": from which cloud, over what bandwidth, with what RPO and RTO, tested when?
These are the "cheap" backup answers that cost the most when the incident actually happens.
05What "good" looks like.
A practical recovery-objective baseline for SW Florida SMBs: 1-hour RPO (max 1 hour of data loss), 4-hour RTO (back operational within 4 hours), with one immutable off-site copy, annual restore testing, and documented runbook.
That posture, configured correctly, runs $400–$1,200/month for most 25–75-person operations. At a $2,000/hour downtime cost, the payback math is usually "one incident in 3 years."
The BIA is 30-minute interviews with 6–10 people, a one-page dependency map, and a per-system RPO/RTO table your CFO can defend to the board. $3,200 fixed, two weeks, no sales follow-on unless you ask for one.