“We can’t really afford downtime” is a feeling. “A down day costs us about $14,000” is a number — and the difference between the two is the difference between IT decisions made by argument and IT decisions made by arithmetic. The number takes about thirty minutes to build, it’s the single most clarifying figure in all of business technology, and here’s the worksheet.

The four buckets of a down hour

1. Labor that keeps costing while producing less. Payroll runs during outages; output doesn’t. Take your fully loaded hourly payroll for affected staff and multiply by an honest productivity-loss factor — rarely 100% (people find something to do) but rarely under 50% when core systems are down. Twenty people averaging $38/hour loaded, working at one-third capacity, is ~$500/hour of paid non-production. Note that this bucket alone usually shocks people, and it’s the floor.

2. Revenue that doesn’t move. This one is business-specific, which is why generic downtime statistics are useless to you. A practice with a full schedule loses appointments that mostly don’t rebook. A contractor misses a bid deadline — an all-or-nothing loss. A shop can’t invoice or take orders; e-commerce stops mid-transaction. Estimate per hour on a busy day, because outages don’t schedule themselves for slow ones. Some revenue defers rather than disappears — be honest about which, but note that “deferred” revenue collected late still costs cash-flow and goodwill.

3. The recovery itself. Emergency IT labor (at emergency rates, if you’re on break/fix), overtime to catch up, re-entering the day’s work by hand, expedited hardware. Recovery routinely costs more than the outage’s first bucket — and it lands after the outage, when everyone thinks it’s over.

4. The slow bleed. Clients who experienced your outage recalibrate their confidence — the one who was on deadline quietly diversifies vendors; the prospect who called a dead line drew a conclusion. Unmeasurable precisely, real universally. A defensible convention: add 10–25% on top of buckets 1–3, higher if your brand is reliability.

Run the math twice: one hour (buckets 1–2, mostly) and one full day (all four). Write both numbers down. For a typical 20–30 person business, the day figure lands in five digits — which most owners find both surprising and, on reflection, obviously right.

What the number is for

This is where thirty minutes of arithmetic starts paying rent, because the downtime number converts almost every IT debate into a comparison:

  • Backup and recovery tiers: a BCDR setup that cuts recovery from two days to two hours costs $X/month; your number says what those saved hours are worth. The decision becomes math. Usually easy math.
  • Managed vs. reactive support: prevention’s whole value proposition is downtime that never occurs. If monitoring and maintenance prevent even one down-day a year, your number tells you what that’s worth against the fee — the comparison that makes managed pricing legible.
  • Redundancy questions: a failover internet line costs $100-something a month and mitigates a specific, countable outage type. Your hourly number answers it in one line — this is the calculation that turns “should we get a backup connection?” from a preference into a quote.
  • Aging hardware: the server past its lifespan isn’t “still working fine” — it’s carrying a rising probability of a day priced at your number. Refresh decisions stop being about whether the old machine still boots.
  • The incident-response shutdown decision: when containment means turning systems off, the leader deciding needs to weigh hours against risk with a dollar value on the hours — mid-crisis is the worst time to derive it.

It’s also, bluntly, the number that makes conversations with us (or any provider) honest in both directions: it justifies real protections and it exposes over-selling. A $3k/day business doesn’t need a $2k/month resilience stack, and a provider who knows your number can say so.

Keep it current, keep it visible

Recalculate annually alongside the IT budget — headcount and revenue move, and the number moves with them. Put it on the one-page plan. Share it with whoever manages your IT: it’s the context that makes every recommendation they bring you evaluable in ten seconds. And expect the number itself to change behavior — in our experience, the meeting where a leadership team first sees its down-day figure is the meeting where the backup testing schedule stops being negotiable.

The bottom line

Thirty minutes, four buckets, two numbers: what an hour costs, what a day costs. It’s the exchange rate between downtime and dollars — and once it exists, technology decisions that used to run on adjectives (“critical,” “probably fine,” “too expensive”) run on arithmetic instead. Few numbers in your business work harder.