A small law firm in Dallas called us on a Tuesday morning last spring. Their website had been down since Sunday afternoon. They only knew because a client tried to book a consultation and called the office to ask why the form was broken. When we looked, the site had crashed after a plugin update and had been serving a white screen for roughly 42 hours.
The firm was paying $79 a month to a monitoring service. That service dutifully sent five email alerts over the weekend. Nobody at the firm was watching email on Sunday. Nobody was watching email at 6 AM Monday when the second wave of alerts fired. The alerts were doing their job. The job just wasn’t very useful.
This is the pattern we see almost every week. Website outages don’t kill businesses the way people imagine, in a dramatic single event. They kill businesses in slow leaks: a lost consultation here, an abandoned checkout there, a Google ranking that never quite recovers because the crawler hit a 500 error at the wrong moment. Most companies never connect the dots because their monitoring service told them the site “was” down and then “was up,” and nothing else got said.
The gap between monitoring and fixing
Most services marketed as “website monitoring” are actually alerting services. They ping your site every 5 to 15 minutes from a few global data centers. If the response fails, they send an email or SMS. That is the entire product.
What happens after the alert fires is 100% on you.
If it’s 3 PM on a Tuesday and your developer is at her desk, that might be fine. If it’s Saturday night or a holiday morning or an hour when nobody is watching, the alert sits in an inbox until someone eventually notices. In our experience across hundreds of small and mid-market sites, the average time from outage to human awareness is closer to 4 hours than to 4 minutes. On weekends and off-hours it’s often 10 to 20 hours.
Then the actual fixing starts. Someone has to log in, diagnose the problem, restore from backup (if a backup exists and is recent), test that the site is working, and stand up the fix. For most agencies, that’s a 2 to 6 hour window even when they respond quickly.
Add it up and a “small” website outage on a Sunday night regularly turns into 15 to 24 hours of real downtime, even though the monitoring service did exactly what it was paid to do.
What downtime actually costs
Downtime cost varies wildly by business type. A rough working estimate for common categories:
Typical hourly cost of website downtime by business type
Business Type | Cost per Hour Down | Notes |
Small B2B services site (10K visits/mo) | $50 to $200 | Lost leads and consultation bookings |
Mid-market e-commerce ($2M annual revenue) | $500 to $2,000 | Direct transaction loss plus cart abandonment |
High-traffic e-commerce ($20M annual) | $8,000 to $40,000 | Compounds fast; single event can be catastrophic |
SaaS product marketing site | $300 to $1,500 | Trial signups paused, paid campaigns wasted |
Local professional services (law, medical) | $100 to $500 | Consultation form outages, brand damage |
Those are direct revenue impacts. The indirect cost is usually bigger and doesn’t show up on a report: prospects who bounced and never came back, Google crawler visits that hit errors and slowed indexing recovery, negative first impressions on new visitors, staff time spent explaining outages to clients.
The compounding version of the cost, where a search engine notices repeated availability issues and adjusts ranking, is the one most owners never trace back to the outage that started it.
The four disciplines a working “website protection” service actually needs
If the goal is not just monitoring but actual site protection, the service needs to handle four separate disciplines. Most vendors do one or two well and leave the other two as “your problem.”
- Monitoring. Continuous availability checks from multiple geographies. Response-time tracking. SSL certificate expiration monitoring. Alerts that go to a human who is actually watching.
- Restoration. When something breaks, a technician actually restores it. Not “here’s a ticket to open with your hosting provider.” Actual restoration, ideally within minutes, using the latest verified backup.
- Backups. Automated daily or hourly backups stored off the production server, ideally in a different region. Tested restore process (untested backups are stories, not backups). Multiple restore points so a bad update from a week ago can still be rolled back.
- Security patching and updates. WordPress core, plugin, and theme updates applied on a schedule with staging validation. Not “auto-update everything and hope.” Because most outages start with a plugin update pushed at the wrong time.
A service that does 1 but not 2 is an alerting service. A service that does 1, 2, and 3 but not 4 will eventually get breached because unpatched WordPress is the single most common attack vector on small business sites. All four together is what actually keeps a site up over time.
Why most agencies quietly outsource this
Most digital agencies do not run website monitoring in-house. They resell someone else’s tool, mark it up, and pass problems back to their clients when things break. This isn’t malicious. It’s economics. Running true 24/7 restoration requires either an in-house on-call team or an integrated relationship with a service that has one.
The tell that your agency is reselling: when the site goes down, they either can’t fix it themselves (“we’ll open a ticket with the hosting provider”) or the fix takes hours because they have to context-switch out of another project. Either signal means the “monitoring” they sold you is not what you thought it was.
Springbolt’s Website Monitor is built around the four-discipline model above. Monitoring plus restoration plus backups plus security patching, all handled by the same team, all 24/7/365. It exists because the fifth outage-and-rebuild cycle for a client is what convinced us that most sites need an integrated service, not a stack of vendors.
What to check on your current setup this week
Whether you use Springbolt or someone else, three questions worth answering about your current site protection right now:
When was your last successful backup, and has anyone tested restoring from it? A surprising number of “backup” plugins silently stop working. The only proof a backup exists is a successful restore drill. If your team can’t say when the last drill was, treat that as an untested backup.
Who gets called at 3 AM when the site goes down, and what is their response time SLA? If the answer is “an email alert to my dev’s personal inbox,” you have a 6+ hour recovery window baked in. If the answer is “no formal process,” you have a 24+ hour window.
Are WordPress core, all plugins, and the theme updated within the last 30 days, with staging validation? If updates are handled ad hoc or “when we get to it,” the site has an accumulating attack surface and is due for a plugin-related crash within 90 days.
If two of the three answers make you uncomfortable, that’s the gap between what you’re paying for and what you actually have.
The real cost calculation
Owners often push back on paying for a proper website protection service on the basis that “it hasn’t broken yet.” That reasoning is like not carrying home insurance because you haven’t had a fire. The point isn’t the monthly cost. It’s the difference between a 20-minute outage during a plugin update and a 20-hour outage that costs you three signed contracts, a Google ranking drop, and two days of staff explaining to clients why the site was broken.
Most businesses that switch from an alerting service to a true protection service do it after one specific event they’d rather not repeat. The math after that event is obvious. The math before it is easy to postpone.
What good looks like
A properly protected website has these characteristics: outages last minutes, not hours. Backups are current, tested, and stored off-server. Plugin and core updates happen on a schedule with staging validation. Security patches are applied within 24 hours of release. Someone is watching the alerts 24/7 who can actually fix the problem, not just report it.
That’s the standard. Anything less is a hope, not a plan.
The 400+ site pattern
Springbolt currently monitors and protects more than 400 client sites. Across that portfolio, the pattern is consistent: sites on the full four-discipline service average less than 15 minutes of user-visible downtime per year. Sites clients have run under alerting-only services before switching averaged 3 to 8 hours per year, sometimes concentrated in a single bad event.
The math works out to a difference of roughly $30K to $200K per year in preventable revenue loss for a mid-market e-commerce site, depending on traffic and average order value. For a professional services site the number is lower in dollars but similar in customer-relationship impact.
Either way, the case for real protection stopped being about the monthly fee a long time ago.
Frequently asked questions
What is website monitoring and how is it different from website protection?
Website monitoring means checking whether a site is reachable and alerting when it goes down. It does not fix outages. Website protection is a broader service that combines monitoring with active restoration, off-server backups, and scheduled security patching, delivered by a team that actually resolves outages rather than just reporting them.
What does website downtime actually cost a business?
It varies by business type. A small B2B services site loses roughly $50 to $200 per hour down in booking and consultation opportunities. A mid-market e-commerce site at $2M annual revenue loses $500 to $2,000 per hour. High-traffic e-commerce can lose $8,000 to $40,000 per hour. Indirect costs, including SEO ranking impact and lost trust, are typically larger than direct revenue loss.
Why do most monitoring services fail to restore websites when they go down?
Most monitoring services are alerting services. They send an email or SMS when the site becomes unreachable. Actual restoration is not part of the product. The customer or their agency has to log in, diagnose, restore from backup, and test the fix, which typically takes 2 to 6 hours during business hours and much longer nights, weekends, or holidays.
How often should a website be backed up?
For most sites, daily backups are the minimum. Higher-traffic or transactional sites benefit from hourly backups. Backups should be stored off the production server, ideally in a different region, with a tested restore process. An untested backup is a story, not a backup.
What are the most common causes of website outages?
In order of frequency across the 400+ sites Springbolt protects: outdated plugin conflicts (especially after WordPress core updates), unpatched security vulnerabilities that get exploited, expired SSL certificates, DNS misconfigurations after domain changes, hosting provider outages, and traffic spikes that exceed server capacity. Most of these are preventable with routine maintenance.
How long does it take to restore a WordPress site from backup?
For a well-maintained site with recent verified backups, restoration typically takes 15 to 45 minutes. For sites without recent backups or with database corruption, restoration can take 4 to 12 hours or require partial rebuilds. This is the range Springbolt targets when a site under Website Monitor goes down.
What is included in Springbolt’s Website Monitor service?
24/7/365 availability monitoring across multiple geographies, active restoration by an on-call team, automated daily backups stored off-server with tested restore processes, WordPress core and plugin security patching on a scheduled cadence with staging validation, SSL certificate monitoring, and monthly reporting. Trusted by 400+ companies.

