Better Stack (Uptime)

What Better Stack (Uptime) actually is

Better Stack Uptime is the uptime-monitoring module of Better Stack, the observability platform built by the company founded in 2021 by Juraj Masar and Veronika Kolejak. The history matters for anyone searching old bookmarks: the product launched as Better Uptime around 2020, and in 2023 the company folded it and its log-management sibling Logtail into one platform called Better Stack. betteruptime.com still resolves, the docs still say Better Uptime in places, and the billing is now the Better Stack bill. In our view it is one of the more complete monitoring-plus-incident packages in its price class, and unusually, the free tier is a real product rather than a countdown timer.

What the module actually covers: HTTP(s) endpoints with keyword checks, ping, TCP and UDP ports, DNS records, mail servers, SSL certificate and domain-expiration watches, and heartbeats that tell you when a cron job silently stops running. Failures are confirmed from more than one region before an alert fires, which is the single most useful anti-noise feature a monitor can have — a probe with a broken route should not wake your team. When a check does fail, the incident record carries an error screenshot, the server response, traceroute and MTR output, and a second-by-second timeline, so debugging starts before you log in anywhere.

On top of the checks sits incident management: on-call rotations with escalation policies, smart incident merging, and alerting through phone call, SMS, email, Slack, Microsoft Teams, push, Zapier, and webhooks. Status pages (public or private, on your own subdomain) are part of the same product, with maintenance windows that pause alerts and notify subscribers in advance. Deeper signals — logs, metrics, traces, error tracking, real user monitoring, and Playwright-based transaction monitoring that drives a real Chrome instance — live in the wider Better Stack platform on volume-based pricing.

Pricing: what you really pay

The free plan is one of the better offers in this category: 10 monitors, 10 heartbeats, and one status page with 3-minute check intervals, totally free — plus, on the current pricing page, allowances like 100,000 exceptions per month, 5,000 session replays, and 30 GB of metrics. For a personal site, a side project, or a small internal tool, that is genuinely sufficient. Status pages on custom subdomains are available even on the free plan.

Paid pricing is per responder: $29 per license per month billed annually, or $34 month-to-month. A "responder" is a person who receives alerts, and the plan includes unlimited phone-call and SMS alerts for them plus unlimited team members who don't get paged — a billing model we prefer over per-seat pricing that taxes read-only users. Paid plans cut the check interval to 30 seconds and add on-call scheduling, escalation policies, and Playwright transaction checks. Monitors scale in blocks: 10 are included, and 50 more cost $25 per month ($21 per month on annual billing); third-party comparisons put the effective per-monitor rate at about $0.21 per month.

Where the bill can grow unexpectedly is the telemetry side. Metrics are metered (30 GB included, then $0.75 per GB per month monthly or $0.50 on annual), logs and traces bill on ingestion plus retention, real user monitoring bills per GB of web events, and error tracking bills per exception. If you use Better Stack purely for uptime checks, status pages, and alerting, the arithmetic is simple and competitive with StatusCake- or Pingdom-class tools; if you consolidate logs, metrics, traces, and RUM onto it as a Datadog replacement, model the volume first — vendor-commissioned comparisons quote roughly $791 per month for a 100-host deployment with five responders, which is still far below enterprise observability bills but is not pocket change. Enterprise plans (SSO, data residency, dedicated support) are quoted rather than listed.

Performance & infrastructure

Better Stack is a SaaS, so the "infrastructure" that matters to you is the probe network and the alerting path, not rented VMs. Checks run from monitoring locations across multiple continents, and multi-location verification is applied before alerts fire; the specific list of probe regions and their count is not published on the marketing pages, and neither is a map in the docs we found. What is published is the cadence: 30-second checks on paid plans, 3-minute on free. Transaction monitoring executes Playwright scripts in a real Chrome instance, which puts it ahead of simple HTTP monitors for login flows and checkout paths.

The alerting path is the product's strongest operational claim: unlimited voice-call alerts on paid plans, with SMS, Slack, Teams, push, and webhook fallbacks, plus escalation policies so the first responder who doesn't answer gets a second one paged. Incidents can be acknowledged and resolved directly from Slack, status-page updates can be posted while an incident is open, and AI-drafted post-mortems generate the write-up from the timeline once it closes. Integrations cover the expected incident stack — PagerDuty, Opsgenie, Splunk On-Call, Zabbix — and the telemetry side speaks to Datadog, New Relic, Grafana, Prometheus, and the three major clouds.

On track record: Better Stack reports several hundred thousand engineers using the platform, with named customers including Accenture, Time, Decathlon, ESET, Brave, and Drata, and it raised funding from Creandum, Susa Ventures, K5 Global, Credo, Kaya.vc, and Lachy Groom. That is a healthy company for this segment. What it does not publish is an uptime SLA for the monitoring service itself — ironic for an uptime company, though common in the category — so if your contract requires your monitoring vendor to carry an availability commitment, that will be enterprise-negotiated rather than off the shelf.

Where it falls short

The per-responder model has a flip side: if your on-call rotation is large, each responder is a line item, and a 10-person rotation is $290 per month on annual billing before any telemetry. Teams whose alerting is mostly Slack-based can mitigate this by treating fewer people as formal responders, but the pricing does incentivizes thin rotations.

Second, free-tier checks run at 3-minute intervals. For catching an outage before your users do, that is acceptable for a blog and slow for a storefront; most serious use needs the paid 30-second tier. Third, probe transparency is limited: we could not find a published list of check locations or per-location latency reporting comparable to what some competitors show, and no SLA covers the monitoring service itself.

Fourth, the platform has grown from an uptime tool into a full observability stack, and the pricing surface grew with it — responders, monitor blocks, metric gigabytes, log ingestion and retention, exception counts, session replays, web-event gigabytes. That is manageable when you use three of the meters and surprising when you use all of them. Finally, the comparison content on their own community pages is marketing material; we used it for feature detail and cross-checked prices against their pricing page rather than treating its verdicts as neutral.

Who should pick this

Our pick for Better Stack Uptime is any team that wants monitoring, paging, and a status page in one bill without assembling three tools. Small-to-medium SaaS companies fit best: the free tier covers staging and side projects, one or two responders cover an on-call rotation, and the status page covers the customer-communication half of an incident. We would also point solo operators and open-source maintainers at the free plan specifically — 10 monitors and a custom-subdomain status page at zero cost is the best starting package in the monitoring and hosting category we cover.

If you are an enterprise with compliance requirements around vendor SLAs or data residency, go in knowing those are enterprise-tier conversations. If you need per-second or 10-second checks, look at dedicated high-frequency monitors instead. And if you want to compare raw check economics against VPS-based self-hosted probes, our benchmarks page shows what a cheap server costs — self-hosting monitors is cheaper at scale but moves the reliability problem onto you.

Key takeaways

• Uptime checks (HTTP, ping, TCP/UDP, DNS, mail, SSL/domain expiry, cron heartbeats) bundled with on-call incident management and status pages in one product.
• Free plan: 10 monitors, 10 heartbeats, one status page, 3-minute checks; paid plans run 30-second checks with multi-location verification before alerting.
• Paid pricing is per responder — $29/month annually or $34 monthly — with unlimited phone and SMS alerts and unlimited non-paged team members.
• Monitor blocks cost $25 per 50 per month ($21 annual); telemetry (logs, metrics, RUM, errors) is volume-metered and is what drives big bills.
• Strong incident tooling: error screenshots, traceroute/MTR, second-by-second timelines, escalation policies, Playwright transaction checks, Slack-native acknowledge/resolve.
• No published SLA for the monitoring service itself and no public probe-location list — fine for most, a gap for procurement-heavy buyers.

Sources

• https://betterstack.com/uptime
• https://betterstack.com/pricing
• https://betterstack.com/community/comparisons/website-uptime-monitoring-tools/
• https://betterstack.com/community/comparisons/better-stack-vs-statuscake/
• https://www.gappsy.com/tools/better-uptime/
• https://linkedin.com/company/betterstack
• https://startupanz.com/dashboard-alerts-logs-uptime/