The honesty caveat
We run an uptime monitoring service. We obviously think MyUptimeBot is a good choice. But this post isn’t a sales pitch — it’s the actual evaluation framework we’d use if we were the customer instead of the vendor. Some of the criteria below are ones we score well on; some we don’t. We’ll note both.
The things that actually matter
After running and using many monitors, the criteria that separate “fine” from “actually good” come down to seven things:
1. Multi-region rechecking before alerting
This is the single most important feature, and it’s the one easiest to miss when evaluating because every vendor says “we have multiple regions.”
The thing to ask: when a check fails from one region, do you immediately recheck from another region before alerting me, or do you alert on the first failure?
A vendor with 50 monitoring regions that pages you on any single-region failure will train you to ignore the pager within a week. A vendor with 3 regions that requires 2-of-3 to fail before paging gives you actionable alerts.
If a vendor’s marketing emphasizes the number of regions instead of the behavior between them, that’s a tell.
(MyUptimeBot does multi-region recheck by default on every plan. We won’t claim others do; ask them.)
2. Free tier that’s actually usable
Most monitoring services have a free tier. The distinguishing question: can you actually monitor a real site on it, or is it a 14-day trial in disguise?
What “actually usable” means:
- At least 1 monitor, indefinite duration
- Check intervals under 5 minutes
- Email or push notifications included
- SSL expiry monitoring included
- No “credit card to enable” tricks
A useful free tier is how vendors prove their product works. A locked-down free tier is how vendors push you to upgrade before you’ve decided you want to.
(Our free tier: 1 monitor, 3-minute checks, email + push, SSL monitoring, no card. UptimeRobot’s: similar. Better Stack’s: time-limited. Pingdom’s: nonexistent.)
3. Check frequency tiers
Faster checks cost more — this is fine and expected. What matters is whether the pricing makes sense for your use case.
Common patterns:
- Free / lowest tier: 3–5 minute checks
- Mid tier ($10–$30/mo): 60-second checks
- Pro tier ($30–$100/mo): 30–45 second checks
- Enterprise: 15–30 second checks
If a vendor charges enterprise prices for 60-second checks, they’re overpriced. If they charge consumer prices for 15-second checks across hundreds of monitors, they’re either undermonitoring or losing money — both worth being suspicious of.
4. Notification channels and routing
Email, push, SMS, webhook are table stakes. What separates good from mediocre:
- Per-monitor channel routing. Your marketing site alerts on email; your checkout alerts on SMS + push. Vendors that force all monitors to use the same notification setup are less useful.
- Notification reliability. SMS that arrives 10 minutes late is worthless. Push notifications that get throttled by mobile OS battery savers are worthless. Ask for delivery time stats.
- Quiet hours / scheduling. Some monitors only need to alert during business hours. Vendors that support this save real on-call sanity.
(We do per-monitor channel routing. Quiet hours are coming.)
5. SSL certificate monitoring included
SSL expiry is the single most common preventable outage. A vendor that doesn’t include SSL monitoring as part of basic uptime checking is missing the highest-impact feature.
Things to verify:
- Does SSL monitoring come with all plans, or is it a paid add-on? (It should be all plans.)
- How early do alerts fire? 30 days out is standard; 60 days is better for slow renewal processes.
- Does it work for self-signed and internal certs? (Less common, but matters for enterprise.)
6. Honest historical reporting
Every vendor shows “uptime over the past 30 days.” What separates good from bad:
- Granularity. Can you see minute-by-minute, or just daily summaries?
- Filtering. Can you exclude scheduled maintenance windows? (You should.)
- Export. Can you pull the raw data into your own systems?
- Honesty. If a check failed and self-recovered, does the dashboard show the dip, or does it round to “all good”?
A vendor whose reporting hides their own monitoring failures isn’t trustworthy. Look for one that shows the noise — that’s how you know they’re not gaming the numbers.
7. Status page that’s actually online during outages
Most monitoring vendors run their own infrastructure on a single cloud. When that cloud has issues, the monitoring service goes down too — and you lose visibility at exactly the moment you need it.
The test: where does the vendor host their own status page? If it’s on the same cloud as their monitoring, it’ll go down when they go down. If it’s on a separate provider (CloudFlare Pages, GitHub Pages, a different cloud), they’ve thought about this.
You can’t fully escape the “monitor the monitor” problem, but you can pick a vendor that takes it seriously.
The things that look important but mostly aren’t
A few features that show up in marketing copy but are usually less important than they sound:
“X regions in N countries!” What matters is whether the alerting logic uses those regions intelligently. 50 regions with single-region alerting is worse than 3 regions with quorum alerting.
“AI-powered anomaly detection.” Mostly marketing for “we have basic statistical thresholds.” If a vendor leads with AI, ask what’s actually being detected and how it differs from a simple threshold.
“Real user monitoring included!” RUM is useful but it’s a different product from uptime monitoring (see our synthetic vs RUM post). A vendor that bundles them is fine; one that only offers RUM and calls it uptime monitoring isn’t.
“Used by [big company].” Usually means a single team at that company once tried the product. Doesn’t tell you much about your use case.
Long feature lists. A monitoring service is either good at the core function or it isn’t. Twenty features that are mediocre are worse than five that are excellent.
The price-to-value sanity check
A rough framework for whether you’re paying the right amount:
- $0–$5/mo: One website you care about, basic checks, email alerts. Free or near-free should cover it.
- $10–$30/mo: Several sites, faster checks, mobile app, SMS included. Most small businesses live here.
- $30–$100/mo: 50–100 monitors, fastest checks, team features, multiple status pages. Growing businesses.
- $100–$500/mo: Enterprise tier — unlimited monitors, integrations, SLA, dedicated support. Reasonable.
- $500+/mo: You’re paying for the support relationship and compliance certifications more than the monitoring itself.
If you’re paying more than $30/mo and you only monitor one site, you’re overpaying. If you’re paying $30/mo and monitoring 200 sites, you’re probably under-tooled.
The actual comparison shortlist
If you’re evaluating right now, the services that consistently come up in honest discussions:
- UptimeRobot — generous free tier, mature, sometimes flaky on alert delivery. Good for personal use.
- Better Stack (formerly Better Uptime) — modern interface, status page included, more expensive than peers. Good for teams.
- Pingdom — long-established, owned by SolarWinds, premium pricing, premium features. Good for enterprise.
- StatusCake — good middle-tier pricing, somewhat dated UI, solid feature set.
- HetrixTools — cheap, lots of features, less polished.
- MyUptimeBot — that’s us. Free tier, mobile-app-focused, mascot-driven brand, smaller than the names above. Honest about what we’re good at and what we’re not.
For self-hosted: Uptime Kuma is the standard, free and excellent if you have somewhere to run it. Statping, Hund, and Vrok are alternatives.
How to actually decide
After you’ve narrowed to 2–3 candidates:
- Sign up for the free tier of each. Spend 10 minutes setting up the same site on each.
- Compare the setup friction. How long to first working monitor? How obvious is the configuration?
- Let them run for a week. See whose alerts arrive when, how reliably, and how noisy.
- Check the mobile experience. If you’ll be on-call, this matters more than the desktop UI.
- Try one deliberate failure. Block your site from the monitor (firewall rule, take it down briefly). See how fast each alerts, and through which channels.
Most monitoring services look identical until you actually use them. A week of side-by-side trial costs you nothing and tells you more than any feature comparison can.
The most important thing, plainly
The biggest mistake isn’t picking the wrong monitoring service. It’s not picking one at all.
If you’re stuck choosing, sign up for whatever has the best free tier today, set up your most important site, and revisit the decision in 6 months. The cost of indecision is the customers who quietly leave the next time something breaks. The cost of picking “good enough” and moving on is zero.
Start free with us, or pick one of the others. The thing that matters is having something running.
