The outages you never hear about
When people talk about “downtime,” they picture a dramatic event — your site is unreachable, the support inbox fills up, customers call to complain. That kind of outage is bad but at least visible. You know it happened. You can respond.
The outages that actually hurt small businesses are quieter:
- The contact form has been broken for three weeks because a plugin update changed the recipient email
- The checkout page returns a 500 error for users on Safari but works fine for the owner using Chrome
- The site takes 12 seconds to load from mobile networks, and bounces 60% of visitors before they see the offer
- The booking calendar shows availability that doesn’t actually exist because the integration broke
- The “Buy now” button silently does nothing on iOS — JavaScript error, no purchase logged
None of these are outages in the technical sense. The server is up. The HTTP status is 200. Your uptime monitor (if you have one) is green.
But the business is broken. And it’s been broken for days, weeks, sometimes months, before anyone notices.
A typical pattern
Here’s a story we hear a lot, in various forms:
A small business owner notices contact form inquiries have dropped lately. Maybe it’s seasonal? Maybe marketing is off? She doesn’t act on it for two weeks.
Then someone she knows mentions they tried to email through the form last month and never got a response. She tries the form herself. The submission button works. She gets the success message. Nothing arrives in her inbox.
She digs in. The form plugin was updated three weeks ago. The update reset the recipient email field to the plugin developer’s default — a generic
noreply@address that never reached her.Three weeks of inquiries, gone.
The technical version of this happens constantly. The plugin update, the JavaScript error, the misconfigured CDN, the expired API integration — they each affect a slice of users in a way the owner doesn’t experience themselves.
Why owners don’t notice
A few reasons silent outages are so hard to catch:
1. The owner doesn’t use the site like a customer. You log into the admin. You navigate from bookmarks. You’re on your office Wi-Fi using the browser you’ve used for years. The customer is on a phone, on cellular, hitting the homepage from a Google search. Your experience and theirs diverges.
2. Most users don’t complain. Industry studies vary, but a useful rule of thumb: for every customer who reports a problem, somewhere between 10 and 100 customers experienced the same problem and just left. By the time complaints are loud, the damage is already done.
3. The breakage is partial. Total outages get noticed because everyone fails. Partial outages only affect, say, Safari users on iOS 17 — which is real customers but invisible if the owner uses Chrome on a laptop.
4. There’s no signal in the data. Analytics show traffic; they don’t always show conversion. A 50% drop in checkout completion shows up in revenue weeks later, not in a real-time dashboard.
The math on what it costs
Imagine a small e-commerce site that does $15,000 in revenue a month. Average order value is $50. That’s 300 orders/month, 10 a day.
Now imagine the “Buy now” button breaks for 30% of visitors — say, anyone on iOS Safari, which is a typical mobile share. Same traffic, 30% can’t actually buy.
Revenue drops to $10,500/month. The site looks healthy. Traffic is steady. The owner notices “things are slower than they were” but attributes it to season, competition, or normal variance. Two months pass. $9,000 in lost revenue before someone realizes.
The dollars matter. So does the trust. Customers who try to buy and can’t usually don’t try again later. They go to a competitor. The lost revenue isn’t recovered the day you fix the bug.
What a “monitor” can and can’t catch
Traditional uptime monitoring catches the obvious outages — site returns 500, certificate expired, DNS broken. It doesn’t catch most of the silent failures above. The site returns 200 with a broken form, and the monitor is none the wiser.
To catch silent outages you need different signals:
Content change monitoring. Specify a keyword that should appear on a key page. If it disappears, alert. This catches the kind of failure where a page “loads” but the content is wrong — a redesign that broke navigation, a CMS update that ate the headline, defacement.
Form submission monitoring. Periodically submit a test entry to your contact form and verify it arrives. This catches the “form looks fine but emails don’t deliver” failure mode directly.
Funnel monitoring. For e-commerce, scripted browser checks that walk through “add to cart → checkout → confirm” and verify each step works. Catches conversion path breakage that uptime checks miss.
Real-user analytics with errors. Tools like Sentry, LogRocket, or Datadog RUM track JavaScript errors per session. If a third of your iOS users are hitting an uncaught exception, you’ll see it in the error count even if uptime monitoring is green.
These are layered with basic uptime monitoring, not replacements for it. Each layer catches different failures.
Practical first steps for small businesses
If you’re running a small business website without any of this, here’s an honest priority list:
1. Set up basic uptime monitoring on the homepage. Free, takes 5 minutes, catches the obvious outages. MyUptimeBot’s free plan covers this; so do most competitors.
2. Add content change monitoring on the homepage and one or two key pages. Define a keyword — your business name, a call-to-action, a price — that must always appear. If it disappears, you’ll get an alert. Available on most paid tiers, including ours.
3. Set up a monthly “verify the basics” routine. Once a month, on a different device than your usual one, submit a test inquiry through the contact form. Add an item to cart and try to check out. Click the major navigation links. Most operators discover something broken every month or two.
4. Add error tracking (Sentry has a free tier) to catch JavaScript errors that don’t break the page but break specific features for specific users.
5. Watch the leading indicators in your analytics. Bounce rate spikes, conversion rate drops, “average time on page” dropping to near-zero on specific pages — these are signals worth investigating, not noise.
None of this is expensive. The total cost for the above on a small business is somewhere between $0 and $30 a month. The cost of not doing it is the silent outage that’s already eating your revenue and you don’t know about it yet.
A useful exercise
If you’ve read this far, here’s a 15-minute exercise worth doing today:
- Open your website on a phone you don’t normally use (or borrow one). Ideally one running iOS if you usually use Android, or vice versa.
- Find your business on Google. Click through as a customer would.
- Try to do the thing your customers come to do — submit a form, book a service, buy a product.
- Note every place something looks wrong, takes too long, or doesn’t work as expected.
You will almost certainly find something. The question isn’t whether your site has rough edges — every site does. The question is whether you’d rather find them before customers do, or after.
The principle
The most expensive outages aren’t the dramatic ones. They’re the quiet ones that affect 10% of your customers for weeks before anyone notices. Monitoring exists to catch the quiet ones at the same speed you’d catch a loud one.
Uptime monitoring is the floor. Content checks, error tracking, and periodic manual verification are the layers above. Each layer catches a different kind of silent failure. The combined cost is small; the combined value is the difference between “we lost a customer once” and “we lost a third of our customers and didn’t realize for two months.”
If you’re running a small business website without monitoring, set up something basic today. It doesn’t have to be us. It just has to be something.
