Better Stack (Logtail)

What Better Stack (Logtail) actually is

Better Stack is the observability platform formed when Better Uptime and Logtail were merged into one product and one brand — the company's own announcement is explicit that "Better Uptime and Logtail have been rebranded as Uptime and Logs on the Better Stack platform," with functionality and pricing unchanged. So the tool you may know as Logtail is now simply the Logs half of Better Stack, sitting next to uptime monitoring, heartbeats, incident management, on-call scheduling, and status pages under one login. In our view that consolidation matters more than the branding exercise: log data is only useful if it feeds the alerting and incident workflow, and Better Stack owns both ends natively instead of stitching them together with webhooks.

Under the hood, logs land in a ClickHouse-backed store, and querying is deliberately plain SQL — the company has been blunt about why, arguing that almost every engineer knows SQL while almost nobody knows Datadog Query Language or PromQL. Alongside SQL you get a live tail, a drag-and-drop query builder, one-click pattern filtering, and PromQL for metrics. Ingestion is OpenTelemetry-native, with remotely managed eBPF collectors that can pull logs, metrics, and network traces from Kubernetes and Docker without code changes, and VRL transforms that can redact PII or drop noisy events before you get billed for them. There is also a wide-events path for high-volume JSON data and an "AI SRE" add-on that proposes root-cause hypotheses from your telemetry and can open GitHub pull requests — but only with human approval, which is the right default.

Pricing: what you really pay

The free tier is genuinely usable for a side project: 10 monitors and heartbeats, 1 status page, 3 GB of logs retained 3 days, 3 GB of traces retained 3 days, 30 GB of metrics, 100,000 exceptions and 5,000 session replays per month, with Slack and email alerts. Three days of log retention is the catch — anything you discover late is gone — but as a permanent free tier it beats most competitors' 14-day trials.

Paid usage is billed two ways. First, users: a Team member with telemetry access is free; a Responder who needs incidents plus telemetry is $34/month monthly or $29/month billed yearly, and channel-based or thread-based incident workflows add $9 per responder per month. Second, telemetry volume: bundles bundle traces, logs, and metrics together with 30 days of log retention. The entry Nano bundle is 40 GB traces + 40 GB logs + 40 GB metrics for $30/month in Europe, $45/month in the United States, and $105/month in Singapore on monthly billing ($25/$38/$88 respectively on yearly). That regional price spread is unusually large, so where you store data is a real cost decision here, not a compliance footnote. There is a 60-day money-back guarantee.

Add-on pricing is where the invoice gets busy. Public status pages beyond the included one are $15/month each ($12 yearly); white-labeling, IP allowlisting, SSO, and sending email from your own domain are each $250 per page per month. Call routing is $250 per phone number per month. Extra heartbeats are $20 per month per 10. On-call scheduling, unlimited phone call and SMS alerts, Terraform providers, and the REST API are included, which we appreciate — competitors love metering those.

Performance & infrastructure

Better Stack runs self-serve telemetry clusters in Virginia, Oregon, Germany, and Singapore, with custom clusters (dedicated S3 bucket in your own VPC) on request. Wide events are kept in object storage while time series sit on local NVMe SSD, and the company claims sub-second analytical queries at petabyte scale — claims we'd characterize as vendor positioning rather than our own measurement; see /benchmarks for what we actually test. What we can point at is structural: ClickHouse is a proven columnar store for exactly this workload, and hot retention is customizable per source, so there is no cold-storage rehydration dance when you need last month's logs during an incident.

The vendor's own marketing compares 1 TB traces + 1 TB logs + 1 TB metrics per month at roughly $687 on Better Stack (Tera bundle, European data location, annual billing, one responder) against an estimated ~$56,774 on Datadog. Treat that as the vendor's arithmetic with the vendor's assumptions — but the order of magnitude matches what we see in list pricing across the log-management market. Transparent bundle pricing with published per-region numbers is the kind of spec honesty we score well; see more in /category/monitoring-hosting.

Company-wise, Better Stack was founded by Veronika Kolejak and Juraj Masar, raised $10M in January 2024 from existing investor KAYA after a prior $18.6M round, and reported 200,000+ developers and 4,000+ customers including Time Magazine, Salesforce, and Hugging Face — all company-stated figures we have not audited. The team is distributed across the US, UK, Canada, Colombia, Netherlands, Czechia, Slovakia, Romania, and Rwanda; there is no single HQ we can source cleanly, so we leave it unset rather than guess.

The incident side deserves its own paragraph because it is what separates this from a pure log tool. Alerts reach engineers through Slack, email, phone calls, SMS, and push notifications, with unlimited phone and SMS alerts included rather than metered per notification. On-call scheduling, escalations, call routing, smart incident merging, and an incident timeline are built in, and there are native iOS and Android apps. Slack and Teams workflows support an /incident command, per-incident channels or threads, automated user synchronization, and AI-generated post-mortems — the workflow tier is the $9 per responder add-on, the basic single-channel integration is included. The Terraform provider and REST API cover both monitoring and status pages, so configuration-as-code teams are not second-class citizens. Status pages are genuinely integrated rather than a bolt-on product: incidents and maintenance notices flow through, subscribers get notifications, and pages can embed charts and metrics. Migration-wise, we like that the platform is OpenTelemetry-native — instrumentation written against OTel standards keeps your optionality, and Datadog and Grafana dashboards can be imported. That combination of an open instrumentation standard, SQL queries, and a storage-layer escape hatch (custom clusters with your own S3 bucket) is the strongest architectural argument for choosing Better Stack over a closed competitor: you are not betting your telemetry pipeline on one company's query language.

Where it falls short

Responder licensing is the first real cost trap: monitoring is only cheap if nobody needs incident features, and at $34/month per responder — plus $9 for the workflows most teams actually want — a ten-engineer on-call rotation is not a rounding error. Second, the aggressive status-page and enterprise-feature pricing ($250 per page per month for SSO or white-labeling) means the bill grows faster than the log volume for teams that need polish. Third, the ecosystem is thinner than Datadog's or Grafana's: the integration list is broad for a company this size, but depth of one-click integrations, community dashboards, and third-party content is simply smaller. Fourth, while SQL querying is a strength, teams already standardized on LogQL or KQL face a migration cost. And there is no published uptime SLA for the platform itself on the pricing page, which we note without dramatizing — most observability vendors in this tier are similarly vague.

Who should pick this

Our pick for small-to-mid engineering teams that want logs, uptime, on-call, and a status page from one vendor without an observability bill that scales like Datadog's. The free tier is enough for a side project; the Nano bundle plus one or two responders is a realistic small-team setup for well under $100/month in Europe. Teams that need deep APM, tracing ecosystems, or compliance certifications as first-class features should compare against the incumbents before committing, and US or Singapore teams should price their region explicitly — the same bundle costs 50% to 250% more outside Europe.

Key takeaways

• Logtail is now the Logs product inside Better Stack, merged with Better Uptime under one account and one pricing page.
• ClickHouse-backed storage with SQL, PromQL, live tail, and OpenTelemetry/eBPF ingestion; queries are fast and the model is learnable in an afternoon.
• Free tier includes 3 GB logs with 3-day retention plus 10 monitors and a status page; paid Nano bundles start at $30/month in Europe, $45 in the US, $105 in Singapore.
• Responders cost $34/month each ($29 yearly), and status-page extras like white-label or SSO run $250 per page per month.
• Storage regions: Virginia, Oregon, Germany, Singapore; custom clusters with your own S3 bucket on request.
• Vendor-stated scale (200,000+ developers, 4,000+ customers, $10M raised in 2024) suggests a stable, funded vendor, though no platform uptime SLA is published.

Sources

• https://betterstack.com/pricing.md
• https://betterstack.com/log-management.md
• https://betterstack.com/press/introducing-better-stack.md
• https://betterstack.com/press/raises-10m.md
• https://logtail.com