Small business owners buy continuous uptime monitors and then rarely look at them. UptimeRobot's free tier is popular; StatusCake, BetterUptime, and Pingdom's paid plans are popular among ops teams that actually get paged when the alerts fire. For an SMB owner who just wants to answer "is my site up and how fast does it respond right now," you don't need continuous monitoring. You need a 30-second diagnostic.
The Uptime Check runs a single server-side probe and reports what any visitor would see right now.
What the probe checks
- HTTP status. 200 is healthy. 301 / 302 shows redirects. 403 / 404 / 500 are the three most common "something is wrong" codes.
- Response time. Server-to-server, one-shot. Under 1 second is fast. 1 to 3 seconds is acceptable but worth investigating. Over 3 seconds points at a hosting tier, a database query, or an unoptimized page.
- TLS handshake. Protocol version, cipher, cert chain authorization, days until cert expiry. A 14-day cert warning lets you catch auto-renewal failures before they break HTTPS.
- Final URL after redirects. Useful when you think your site lives at one URL but your hosting actually redirects to another.
- Content-Type header. Confirms you're serving HTML and not accidentally serving the raw JS bundle or a JSON 404 page.
- Server + Cache-Control headers. Quick check for version-leak (X-Powered-By) and CDN cache policy.
When to use this vs a continuous monitor
Use the single-shot diagnostic when:
- You just updated DNS or switched hosts and want to confirm propagation.
- A customer told you the site was down and you want to see if it's still down.
- You're auditing a client site before a proposal.
- You want a baseline response time before a speed optimization.
Use a continuous monitor when:
- Downtime costs you real money and you need to know about it within minutes.
- You have SLAs with customers.
- You need historical uptime data for compliance or investor updates.
For continuous monitoring, UptimeRobot free tier covers most SMB needs (50 monitors at 5-minute intervals). StatusCake's free tier is similar. BetterUptime and Pingdom are paid with better alerting and status pages.
What the probe can't tell you
A single probe from one server location doesn't represent global reachability. Cloudflare may be healthy for North America but timing out for visitors from South America. A site that returns 200 to a generic server-side probe may return 5xx to logged-in users due to backend failures. Client-side JavaScript errors don't show up in HTTP response codes.
For any of those, a single probe is insufficient. Use the diagnostic as a quick first-pass; use a real monitor for anything you need to trust.
The one thing almost every SMB site fails
Certificate expiry. Cert renewal is supposed to be automatic (Let's Encrypt + certbot, Cloudflare origin certs, hosting-provider-managed certs). It's not always automatic. I've seen sites break HTTPS because:
- The cert-bot cron job stopped running after a server restart.
- The hosting provider changed their cert provider and the old domain was never migrated.
- A CAA record change blocked the renewal authority without anyone updating the CAA.
- The site moved to a new server and the cert didn't follow.
The uptime diagnostic's cert-expiry check flags this 30 days before expiry. Fix it at 30 days; you have time. Fix it at 3 days; you're probably not sleeping that night.
Related reading
- Mega Security Analyzer for full security scan including TLS
- PQC Analyzer for the post-quantum cryptography layer
- SSL Grade for A-F letter grading of TLS config
Fact-check notes and sources
- UptimeRobot, StatusCake, BetterUptime, Pingdom pricing pages as of April 2026.
- Let's Encrypt renewal cycle documentation.
- Google web.dev on response-time thresholds.
This post is informational, not monitoring or security-consulting advice. A one-shot probe is not a substitute for continuous monitoring where uptime is business-critical. Mentions of third-party monitoring services are nominative fair use. No affiliation is implied.
