What website downtime actually costs a small business — and what closes the gap.

Most articles about the cost of website downtime quote a four-hour outage at a million dollars. The number is right for enterprises. The number is wrong for a small shop, and the gap in between is what this article is for. A four-hour outage during a small business's busiest order window does not cost a million dollars. It costs the orders that did not happen in those four hours, plus a small amount of trust you cannot invoice back. That is the real number, and the real one is the one that pays for a monitoring subscription.

What a four-hour outage looks like, in dollars

Pick a shop that takes between fifty and two hundred online orders on its busiest day. The orders come in waves; one of those waves lands squarely between three and seven in the afternoon. A four-hour outage that overlaps that wave costs you the orders you would have taken — not all of them, because some customers retry from a phone, and some come back the next day. A reasonable midpoint for a small shop: twenty to forty missed orders, each worth somewhere between twenty and eighty dollars. The total is between four hundred dollars and three thousand dollars of revenue that did not happen.

Underneath that revenue-loss number is a smaller trust number the shop cannot invoice back. The customers who landed on a "site offline" page during the outage do not all come back. They retry the URL the next day, see the site is up, and a small percentage of them blame the shop for being unreliable — not for anything the shop did, just for the shape of the outage they met. The trust loss is hard to measure; it is real. The deeper reason many shops under-buy monitoring is that the trust number is not on the spreadsheet.

For the broader shape of "what runs a small business site", the website monitoring for small business pillar walks through the four checks that cover most of what can fail. This article focuses on the one matrix that pays the monitoring subscription back: what does the same outage cost at different detection latencies.

What the cost moves by, when you find out faster

The four-hundred-to-three-thousand number above assumes you find out about the outage from a customer. The customer's email arrives, you read it, you call your host or developer, the problem gets diagnosed, the fix starts. By the time the diagnosis lands, the four-hour window is already half-over. The math on the lost orders still works out.

The same outage, found in two minutes instead of six hours, costs you about twenty minutes of revenue. Not because anything about the outage changed — the server still went down. The change is in the gap between "server stopped answering" and "a person who can fix it stopped doing what they were doing." Closing that gap is what you are paying for when you buy a monitoring tool, and the closest free analog that surfaces the same kind of detection gap is the free uptime check: one probe, one verdict, no ongoing subscription. It is the right place to start before the subscription becomes the right answer.

The math on detection latency is what the third-party figures on /benchmark/uptime-stats lean on. The tier page publishes the round-number SLA floors, the cadence-vs-tier split, and the MTTR baselines per hosting tier. The numbers in this article are reading off the same fleet, same baseline. The pragmatic figure for a small shop: a one-minute probe cadence on Pro fires 24x faster than a five-minute cadence on Starter, and the revenue difference during a four-hour outage is the orders that fit between minute zero and minute five.

The bought-time calculation

A monitoring tool that costs twelve to nineteen dollars a month and shortens the average detection time from six hours to two minutes has to fire twice a year to break even. Most shops with any kind of online order flow fire it more often than that. The calculation is not a sales pitch — it is the literal sum of two numbers and one rate, the rate being how many order-bearing hours of the year the site is offline.

The arithmetic works the same whether the subscription is month-to-month or one-time. The 'SLA & uptime commitments for SMBs' framing buyers ask about applies here: a 99.9% SLA rounds out to about nine hours of downtime per year, and a small shop on that round-number SLA averages one major outage per quarter. The detection-gap math is what closes the gap between the nine-hour number (the committed SLA) and the four-hour outage in this article. The shop does not still get hit; it gets hit less.

Need a number for YOUR shop? Punch your daily revenue into the downtime cost calculator and see what an outage is costing per month before the hourly figure ever lands in the postmortem.

For the deeper comparison with the modern uptime-plus-incident-management products that advertise an SLA in the catalog itself, the SiteGuardian versus Better Stack comparison lays out the editorial matrix. Most small shops do not need the incident-management tier; some do. The decision matrix at /vs/all ranks the broadest comparator slate against the same detection-latency lens this article defends.

What monitoring does not do

It does not prevent the outage. It does not fix the host. It does not write the postmortem. It only tells you the truth about when the site stopped answering, and it tells you sooner than your customers would. The fix still belongs to whoever owns the deployment, the host, the renewal cadence on the certificate, or the CDN configuration. The monitoring subscription is the watcher — the cheaper tools in the category do exactly this and very little else, and the choice between them is mostly about the size of the alert surface, not the depth of the watch.

It also does not catch the failure modes that hide behind a 200 response — the checkout-broken-but-still-rendering bug, the inoperative contact form, the dead link on a high-traffic page. The four-piece base monitor covers one of those four failure modes. The other three are what the layered stack adds: a separate probe for checkout failure, a separate probe for form failure, and a daily crawl for broken outbound links. The depth of each is its own article; the gap this article covers is what happens to the green-check uptime monitor when the failure is silent.

What the more expensive outages cost

The four-hundred-to-three-thousand figure is the cost on a Tuesday afternoon order wave. The bigger outages — the ones where the certificate expires on a Sunday morning and the visitor sees a red padlock, the ones where the checkout silently fails on Black Friday morning, the ones where the cart looks fine and the payment provider hangs on the redirect — cost an order of magnitude more. The cert-expiry outage is the corner case the monitoring SSL certificate expiry small business piece walks through; the four-hour detection-gap math this article covers applies identically to a Sunday morning cert lapse, just with a trust multiplier on top of the lost-order multiplier.

The detection-gap math also applies to the failure modes the four-piece base monitor misses. A conflict resolution between the base footer monitor and the checkout-failure monitor is what the why uptime-only monitors miss the failures that cost you revenue piece walks through. The cert and the layered-stack pieces are the same content graph the SiteGuardian pricing surfacesits inside — one subscription tier covers all four failure modes, no add-on, no surprise.

The honest pitch

SiteGuardian is one of many monitors that does this. What it does well is the boring part — it is cheap, it is small, it does not grow into a platform that needs its own dashboard. The smallest paid path is a one-time purchase: the lifetime tier, a flat USD purchase that owns the watch on the shop's terms. The month-to-month Starter and Pro tiers are the right fit for the shop whose billing flow prefers a subscription; the lifetime tier is the right fit for the shop that wants to pay once and stop thinking about the monitoring subscription as a recurring line item.

The full pricing surface — what each tier unlocks and how often it bills — lives on the pricing page. The cheapest free path before any of that is the free URL audit: one probe, one verdict, in under thirty seconds. It is the right place to start. The four-piece test this article leans on is the same one the broader cluster of articles on the site grounds every page against, and the detection-gap math this article lands on is what that test pays for.