The questions vendors don't answer
The marketing pages of every uptime monitoring tool look roughly the same. They all promise reliability. They all show a dashboard with green checks. They all list the same integrations. None of them tell you what actually matters when something breaks at 3 AM.
This guide is the version we wished existed when we were evaluating tools ourselves. The questions below are the ones we found buried three pages deep in docs, hidden in pricing footnotes, or that we only learned the answer to after switching providers.
Check cadence: the underrated dimension
The interval at which a tool checks your monitors is the single most important spec. It defines the floor on your time-to-detection.
If you check every 5 minutes, your average detection delay is 2.5 minutes (and your worst case is just under 5). If you check every 30 seconds, your average is 15 seconds. Across a year of operation, that gap is the difference between catching outages early and customers tweeting at you first.
What to look for:
- Most "Pro" tier plans give you 1-minute checks. Some give 5.
- 30-second checks are increasingly the standard at the top tiers.
- Sub-30-second checks rarely matter unless you\'re running a financial product where seconds count.
Watch for vendors that show "1-minute" in marketing but only deliver it on annual billing of higher tiers. Read the pricing table footnotes carefully.
Multi-region confirmation
Single-region monitoring is a trap. A check that fails from one region tells you that one region can\'t reach your site. That\'s often a regional ISP problem — not your problem — and shouldn\'t wake anyone up.
Good tools confirm a failure from two or three regions before declaring an incident. This single feature reduces alert noise by an order of magnitude. Without it, every Cloudflare hiccup, every regional BGP wobble, every ISP route flap becomes an outage page.
Questions to ask:
- How many regions does the tool check from?
- How many must agree before declaring an incident?
- Is multi-region available on all plans, or just the top tier?
The status page question
Public status pages have become table stakes for any business that takes uptime seriously. They reduce inbound support volume during incidents, build trust by showing you\'re not hiding outages, and create a paper trail that helps with enterprise sales.
What we\'ve learned watching the market:
- The biggest standalone status-page vendor charges $99/month minimum.
- Most uptime tools now bundle status pages, but quality varies wildly.
- Custom domains (status.yourdomain.com) are often paywalled to higher tiers.
- Subscriber notifications (email/SMS to customers) are sometimes a separate add-on.
Decide upfront whether you need status pages, then check how each vendor prices them — not just whether they\'re "included."
Alert routing and on-call
The alerting story is where good monitoring tools differentiate themselves from mediocre ones. The basics — email, Slack, webhook — are universal. What matters is the layer above.
Look for:
- Per-monitor routing. Your marketing site outage shouldn\'t page the same person as your payments outage.
- Quiet hours. Per-recipient time windows when only critical alerts go through.
- Escalation policies. If the primary on-call doesn\'t ack within X minutes, page the secondary.
- Bundling. When 50 monitors fail at once, you should get 1 incident, not 50 alerts.
- Auto-resolve. When the monitor recovers, the alert should clear automatically.
If you have an on-call rotation, having these built into the monitoring tool means you can avoid bolting on PagerDuty or Opsgenie ($25-40/user/month) on top of your monitoring bill. That math adds up fast for a five-person team.
Extensibility: webhooks and API
No monitoring tool will have every integration you need. Webhooks and a real API are the escape hatches that let you bridge gaps yourself.
Test the webhooks during evaluation:
- Are payloads documented and stable?
- Is HMAC signing supported so you can verify the request came from the vendor?
- Are there retry semantics if your webhook endpoint is down?
- Are auto-resolve events sent (not just incident-open)?
For the API, check whether you can fully manage monitors programmatically (create, update, delete, fetch history). Some vendors offer read-only APIs on lower tiers and full APIs only on enterprise.
Pricing traps to watch for
Things that look good on the pricing page but bite later:
- "Unlimited monitors" with slow check intervals. 1000 monitors at 5-minute intervals is far less useful than 100 at 30 seconds.
- Per-user pricing. Some vendors price per team member. A 10-person team can suddenly cost more than the monitoring itself.
- Status page paywalls. Custom domain, subscriber notifications, branded design — often locked behind upgrades.
- SMS surcharges. Watch for per-message charges that stack up during multi-day incidents.
- API rate limits. Generous-sounding limits that turn out to be 60 requests per hour.
A practical evaluation checklist
If you only do one structured evaluation step, do this: take three vendors and walk through this checklist for each.
- What\'s the minimum check interval at the plan I\'d actually buy?
- Multi-region confirmation: included? How many regions? How many must agree?
- Are status pages included? Custom domain? Subscriber notifications?
- Does it cover all eight check types I might need (HTTP, ping, port, SSL, keyword, page-load, DNS, cron heartbeat)?
- Alert channels included vs paywalled at my plan tier?
- Built-in on-call rotation, or do I need to add PagerDuty?
- Webhook quality: HMAC, retries, auto-resolve, documented payloads?
- Full API or read-only?
- Price per added user (if any)?
- What\'s the actual monthly bill at my expected scale, after add-ons?
The vendor that wins this checklist for your specific situation is rarely the one with the flashiest marketing site. It\'s usually the one that includes the things you\'ll actually use without nickel-and-diming around the edges.
