By Jimmy S. | Posted July 8, 2026 | 8 min read
A systems analyst’s take, minus the scary math
Walk into most small businesses in Calgary and the server is the thing nobody thinks about. It runs in a closet, a back room, sometimes literally under a desk. It quietly holds the files, runs the line-of-business app, keeps email moving. And because it just works, it falls off everyone’s radar, right up until the morning it doesn’t.
I want to talk about what actually happens when it goes down. Not the doom-and-gloom version you’ve read a hundred times, but the honest one, including the part where I tell you why the numbers in those other articles are mostly wrong for a business your size.
When a server dies, the symptom is immediate: nobody can get in. Files won’t open, the app won’t load, and the phones-and-email setup half your team depends on stops cooperating. Work doesn’t slow down, it stops, all at once, across whatever departments share that box.
That part is real, and you don’t need a study to confirm it. If your shared drive is down at 9 a.m., your morning is gone. The interesting questions come after.
Here’s where I part ways with most of the IT-blog crowd.
Somewhere in nearly every “cost of downtime” article is a confident figure: $5,600 per minute. It gets repeated like gospel. The problem is that it traces back to a single Gartner blog comment from 2014, pre-cloud, pre-SaaS, and never presented as formal research in the first place. The analytics firm EMA went so far as to call it “an urban legend.”
It’s also an enterprise number. When researchers actually break it out by company size, the picture changes completely. Ponemon’s data-center outage study pegs a small business at roughly $427 a minute, versus $9,000 for a mid-sized data centre. Independent field research by EMA found the “average” per-minute cost is dragged sky-high by giant companies: an org of 1,000–2,500 staff lands around $1,850 a minute, while a 20,000-person enterprise sits north of $25,000.
So when an MSP tells a ten-person Calgary firm it’s losing thousands of dollars a minute, that’s not analysis, it’s a sales pitch borrowing a Fortune 500 statistic. The honest answer is that your number depends on your revenue per hour, your headcount, and how much of your work truly halts when the server’s offline. For most SMBs it’s real money, but it’s hundreds per hour, not the five-figure-per-minute panic you’ve been sold.
That distinction matters, because decisions made on inflated fear tend to be bad decisions.
If you want grounded numbers, skip the vendor blogs and read Statistics Canada. Their 2023 survey of Canadian businesses (released October 2024) is about as neutral as it gets:
Read that last point twice. Fewer businesses are getting hit, but the ones that do are paying more to dig out. That’s the trend line that should concern a Calgary owner, not a per-minute scare figure.
One more useful, unglamorous fact from that survey: 88% of ransomware victims didn’t pay the ransom. Paying was never the expensive part. The recovery was.
The cost nobody budgets for: getting back to normal
The outage window is the part people picture. The recovery is the part that actually hurts.
Even after the server’s back, your team is re-entering what was lost, double-checking what saved, chasing the customer emails that piled up, and reconciling the small messes that happen when people improvise around a dead system. That tail can run for days.
IBM’s 2025 Cost of a Data Breach report– the long-running IBM/Ponemon study- puts hard numbers on the worst version of this. When an outage stems from a breach, the average time to fully identify and contain it is 241 days, and 76% of organizations took more than 100 days to recover completely. The global average breach cost was $4.44 million; in Canada specifically it was $4.84 million, and ransomware-driven incidents averaged $5.08 million.
Those are enterprise-weighted averages too; your exposure is smaller. But the shape of the cost is identical at every size: the recovery dwarfs the outage.
Why server crashes happen (still mostly boring reasons)
The cyberattack angle gets the headlines, but the Uptime Institute, which tracks outages every year, finds that the leading causes of serious failures remain decidedly low-tech: power problems and plain human error, with human error a rising factor. Notably, outage frequency has actually dropped since its 2020 peak, even as the cost per incident climbed roughly 40% since 2019.
In practice, the things that take down an SMB server are the things that were quietly degrading for months:
Almost none of these are sudden. They’re cumulative, and they’re nearly all visible in advance if someone is actually watching.
I’m not going to hand you a numbered listicle of “five pillars.” If you remember nothing else, remember this: recovery speed is the only metric that matters. Everything else is in service of it.
Concretely, for a Calgary SMB that means three honest questions:
Can you restore, and have you proven it? A backup you’ve never test-restored is a hope, not a plan. The businesses that recover in hours instead of weeks are the ones that already know their backups work, because they checked.
Is anyone watching the server, or just reacting to it? The failures above announce themselves early. Basic monitoring turns a planned $400 drive swap into a non-event instead of a Monday-morning catastrophe.
If the main server died today, what’s plan B? Not a polished failover suite, just a real answer. For some shops that’s a second box; for others it’s cloud-hosting the one or two systems you genuinely can’t run without.
That’s it. Tested backups, eyes on the system, and a plan B you could describe in one sentence. The Calgary businesses that handle a crash gracefully aren’t the ones that spent the most, they’re the ones who decided, before the bad morning, that the server in the closet was worth thirty minutes of attention a week.
It’s not glamorous. But neither is explaining to your customers why you’ve gone dark, and you only get to pick one of those conversations.