Better Stack Review

Better Stack is what you get when an uptime monitor and a log viewer merge and keep the pricing page simple.

The company started as Better Uptime and Logtail, folded both into one product, and now sells three things on a single bill: uptime monitoring with status pages, incident management with on-call schedules, and a telemetry backend for logs, traces and metrics.

My opinion up front: for a team under about 20 engineers it is the most predictable observability bill on the market, and that predictability is the product.

This review is for developers and small platform teams deciding whether Better Stack can replace a stack of UptimeRobot plus PagerDuty plus a log tool, or whether it can stand in for Datadog.

I go through what the product includes, how the uptime and telemetry halves work day to day, what each line item costs for the US and Europe data regions, and the real limits: no classic per-host APM, a slow free check interval and a region surcharge that most comparison articles miss.

Five alternatives are priced at the end so you can check the numbers against your own stack.

Better Stack is a hosted monitoring and observability platform built from two earlier products, Better Uptime and Logtail, that now share one account, one pricing page and one incident timeline.

It is sold to developers rather than to enterprise procurement: you sign up with a credit card, the free plan is labelled for personal projects, and every paid line item is a published number.

The Uptime product checks websites, APIs, ping, SSL and domain expiry, DNS, SMTP, POP3 and IMAP, and runs Playwright scripts as transaction monitors. It also hosts status pages on your own subdomain, and the incident side handles on-call schedules, escalation rules, phone calls and SMS.

The Telemetry product stores logs, traces and metrics. It is OpenTelemetry-native, accepts Vector, syslog and Prometheus remote write, and lets you query with SQL or PromQL from drag-and-drop dashboards or a live tail.

Retention is set per source, and the vendor's own pitch is that nothing goes to cold storage: every log stays queryable for its retention window. The same backend powers the real user monitoring add-ons, error tracking with 100,000 free exceptions a month and session replay with 5,000 free replays.

Day one is the uptime half. You add a URL, choose a check type and a check interval, pick alert channels (Slack, Teams, e-mail, push, SMS, phone), and point a status page at your domain. Heartbeat monitors cover cron jobs: the job pings a URL on each run and Better Stack alerts when the ping stops.

The free plan checks every 3 minutes; paid monitors go down to 30-second checks, which is the number that matters if your SLA is measured in minutes.

On-call sits on the same incidents. Each person who should be paged needs a Responder license, which unlocks phone and SMS alerts, escalation policies based on time and who is on duty, and one-tap acknowledgement.

People who only need to read dashboards are Team members and cost nothing, so a 12-person team with three people on rotation pays for three seats, not twelve.

The telemetry half is a pipeline. You ship logs with Vector, an OpenTelemetry collector or a platform integration, pick the data region (Virginia, Oregon, Germany or Singapore), and query in SQL or PromQL.

Bundles are sized in GB per month across the three signals, with 30-day retention on the paid tiers, and anything above the bundle is metered per GB ingested and retained.

The region you pick changes the bill: the US-region bundles cost 50% more than the Europe-region ones, and Singapore more again, so the cheap price in most reviews is the European one.

Better Stack publishes every price, which already separates it from half the observability market. The figures below were read on Better Stack's pricing page in October 2026; the competitor figures come from Datadog pricing, Grafana Cloud pricing, UptimeRobot pricing, Checkly pricing and New Relic pricing, read the same month.

Three line items make up a Better Stack bill. Monitors beyond the free 10 come in packs of 50 billed yearly, or $25 a month on monthly billing; extra heartbeats come in packs of 10 at $17 yearly ($20 monthly). Each person who gets paged needs a Responder license, $34 a month on monthly billing and less on yearly; Team members who only read dashboards are free. Telemetry is a bundle sized by GB: in the US region, Nano is $38 a month, Micro $150, Mega $315 and Tera $630 billed yearly, and 20% more on monthly billing. The Europe region is cheaper for the same allowances (Nano $25, Micro $100, Mega $210, Tera $420 yearly), and Singapore is dearer. Overage is metered at $0.10 to $0.35 per GB ingested and $0.05 to $0.18 per GB retained, depending on region.

Cost scenario, computed from those rates: a five-person team with 40 HTTP monitors, two people on the on-call rotation and about 40 GB a month of logs in the US region pays for one monitor pack, two responders ($58) and the Nano bundle, which is $117 a month billed yearly, or $1,404 a year. The same team on monthly billing pays $138. Move the data to Germany and the yearly figure drops to $104 a month. For comparison, Datadog's Infrastructure Pro plan alone is $75 a month for five hosts billed annually, before logs (billed per GB ingested and per million indexed events) or APM at $31 per host, and Grafana Cloud's Pro plan starts at a $19 platform fee plus usage.

Devshot data: of the 210 monitoring tools Toolradar tracks as of October 2026, 64% are free, freemium or open source, and the median cheapest paid plan across the 51 tools that publish one is $19 a month. Better Stack's monitor pack sits near that median, and its on-call license is where it moves above it.

Better Stack fits teams of two to roughly twenty engineers who want one bill for uptime, paging and logs. The per-responder model rewards teams where only a few people take the pager, the status pages are good enough to face customers, and a Nano or Micro bundle covers the log volume of most SaaS products with a handful of services. Teams in Europe get the extra benefit of the cheaper data region.

It is a weaker fit for three kinds of buyer. Teams that need per-host APM with service maps and continuous profiling will find the traces product thinner than Datadog or New Relic.

Teams with more than a few hundred GB of logs a month should price Grafana Cloud and the overage rates carefully, because the Tera bundle ends at 700 GB per signal. And anyone who only needs uptime checks and nothing else gets more free monitors from UptimeRobot, 50 against 10, at a 5-minute interval.