Statsig Status: How to Check If Statsig Is Down Right Now (2026)
Updated July 2026 · 6 min read · By API Status Check
Quick Answer: Check Statsig Status Now
Statsig is a feature management and experimentation platform used to run feature flags, A/B tests, and product analytics from a single SDK. Unlike some monitoring tools where an outage means a hard failure, Statsig's client SDKs are built to fail open using locally cached configs — which changes what "Statsig is down" actually means for your app.
Two Independent Systems: Config Sync vs Event Logging
- • SDK downloads gate/experiment configs on init
- • Periodically re-syncs in the background
- • Falls back to last cached config if sync fails
- • A degraded sync rarely breaks your app immediately
- • SDK batches exposure and custom events
- • Flushes to Statsig on an interval or app close
- • A logging outage doesn't affect flag evaluation
- • But it can cause gaps in experiment result data
Understanding Statsig's Status Page (status.statsig.com)
Statsig operates status.statsig.com to track their infrastructure components. Because config sync and event logging are decoupled, they're tracked and can degrade independently.
Config Sync / Download API
Serves the latest gate, experiment, and dynamic config definitions to SDKs. If degraded, SDKs continue evaluating against the last successfully cached config rather than failing outright — most users won't notice unless a rollout change fails to propagate.
Event Logging
Accepts exposure logs and custom events from client and server SDKs. Degradation here affects experiment result accuracy and analytics completeness, but does not affect which flag variant a user is served.
Console / Dashboard
The Statsig web app used to create and manage gates, experiments, and view results. A slow or unavailable console doesn't affect already-deployed flags — it only affects your ability to make new changes.
Server-Side Evaluation API
For server SDKs using remote evaluation instead of local config sync, this API evaluates gates on each request. This path has a harder real-time dependency on Statsig's infrastructure than the local-eval client SDKs.
Webhooks & Integrations
Notifies external systems (Slack, PagerDuty, data warehouses) of experiment results and config changes. Failures here don't affect flag serving, only downstream automation.
API
The management API used for CI/CD pipelines, Terraform, and programmatic gate management. Separate from the SDK-facing config sync and evaluation endpoints.
Diagnosing Statsig Flag and Logging Issues
Because Statsig SDKs fail open, most reported "outages" are actually stale-cache behavior, targeting misconfiguration, or SDK initialization timing issues.
Statsig Diagnostic Checklist
# 1. Confirm the SDK finished initializing before evaluating gates
await statsig.initialize(SDK_KEY, user)
# Evaluating before init resolves returns default/cached values
# 2. Check network requests for config sync calls
Filter for requests to api.statsig.com/v1/download_config_specs
# Should return 200 periodically
# 3. Verify the correct client vs server SDK key is used
Client keys and server secret keys are NOT interchangeable
# 4. Check the user object being passed to gate checks
Missing userID or custom properties can silently change targeting results
# 5. Confirm event logging flush requests are succeeding
Filter for requests to api.statsig.com/v1/log_event
Common Statsig Failure Causes That Aren't Outages
- 🔴 Evaluating gates before SDK init resolves: Calling checkGate() before the initialize() promise resolves returns default or bootstrap values, which looks like a flag misfire but is a code-ordering issue.
- 🔴 Mixing up client and server SDK keys: Client SDK keys are scoped for local evaluation with limited data exposure; server secret keys enable full local evaluation. Using the wrong key type causes evaluation errors that look like an outage.
- 🔴 Stale cached config after a rollout change: If a client SDK's background sync interval hasn't fired yet, users may see the previous rollout percentage rather than a change you just published — this resolves itself on the next sync cycle.
- 🔴 Incomplete or missing user object fields: Targeting rules based on custom user properties (plan tier, region, etc.) silently fall through to defaults if those properties aren't populated on the user object you pass in.
- 🔴 Ad-blocker or CSP blocking statsig.com domains: Similar to other client-side SDKs, strict Content-Security-Policy or privacy extensions can block config sync and event logging requests without throwing a visible error.
Monitor your own config sync endpoints from outside your network
If you self-host a Statsig proxy or forwarder, external uptime monitoring catches failures before stale flag configs quietly pile up. Better Stack has a free plan.
Try Better Stack Free →What Statsig Status Colors Mean
Common Statsig Issues and What They Mean
"Gate returns false for a user who should pass"Check the gate's targeting rules and rollout percentage in the Statsig console, and confirm the user object passed to checkGate() includes the fields those rules depend on (userID, custom properties, environment tier). This is far more often a targeting mismatch than an outage."Experiment shows no data or very low exposure count"Verify logEvent/exposure logging isn't being blocked by an ad blocker or CSP rule, and check that the experiment's targeting isn't scoped so narrowly that few users qualify. Also confirm the experiment is actually running (not paused) in the console.""SDK not initialized" error or hang"The initialize() call is waiting on a network request to api.statsig.com that isn't completing. Check network connectivity to Statsig's domains, verify your SDK key is valid, and consider using bootstrapValues or a timeout fallback for critical paths."Server SDK remote evaluation returning errors"Unlike local-evaluation client SDKs, remote evaluation makes a live API call per check. Check status.statsig.com's Server-Side Evaluation API component — this path has a harder real-time dependency and will surface outages more directly than cached client SDKs."Config changes in console not reflected in production"Client SDKs sync on an interval, not instantly. Allow a few minutes for propagation, or force a refresh via the SDK's manual sync method if your integration supports it. If it's been well over the expected sync window, check the Config Sync component for an incident.Frequently Asked Questions
Where is the official Statsig status page?
Statsig's official status page is status.statsig.com. It breaks down status by component: Config Sync/Download API, Event Logging, Console, Server-Side Evaluation API, Webhooks, and the management API.
If Statsig is down, will my feature flags stop working?
Usually not immediately. Statsig client SDKs cache the last successfully synced config and continue serving those values if the network request to fetch new configs fails — this fail-open design means an outage typically causes stale flag values rather than application errors, as long as the SDK initialized successfully at least once before the outage.
Why did my new rollout percentage not take effect right away?
Client SDKs sync configs on a background interval rather than instantly. A change published in the console can take up to that interval (commonly a minute or more, depending on SDK settings) to reach already-running client sessions. Server-side remote evaluation reflects changes on the next API call instead.
How is Statsig different from LaunchDarkly for outage purposes?
Both platforms use a similar local-cache-with-background-sync model for client SDKs, designed to fail open during outages. Statsig additionally bundles product analytics and experimentation into the same event pipeline used for flag exposures, so a logging-path issue can affect experiment result completeness even when flag serving is unaffected.
How do I get alerts when Statsig goes down?
Subscribe to updates at status.statsig.com for official incident notifications. If you run any self-hosted proxy or forwarder in front of Statsig, set up independent external uptime monitoring on that component since Statsig's own status page can't see it.
Alert Pro
14-day free trialStop checking — get alerted instantly
Next time Statsig goes down, you'll know in under 60 seconds — not when your users start complaining.
- Email alerts for Statsig + 9 more APIs
- $0 due today for trial
- Cancel anytime — $9/mo after trial