Are there any product analytics for experiments that detect sample ratio mismatch on large event volume?

Yes—many experimentation platforms and product analytics tools can detect sample ratio mismatch (SRM), including at large event volumes, though the exact implementation and limits vary.

Common options

• Experimentation platforms with SRM checks built in

  • Optimizely
  • Statsig
  • Amplitude Experiment
  • LaunchDarkly experimentation
  • GrowthBook
  • Eppo
  • VWO / Adobe Target in some setups

• General analytics tools

  • Amplitude
  • Mixpanel
  • GA4
  • Heap
    These may not always call it SRM explicitly, but you can often build alerts or checks using assignment counts and expected allocation.

What to look for

For large event volume, the key is not just whether the tool supports SRM, but whether it:

1. Checks on assignment units rather than raw events
   SRM is about users/sessions/units randomized, not event counts.
2. Handles high cardinality and scale efficiently
3. Supports automated alerts
4. Lets you define the correct denominator
   • unique users
   • sessions
   • accounts/tenants
5. Detects drift early
   • allocation imbalance by variant
   • segmentation-specific SRM
   • exposure vs assignment mismatch

Important caveat

If you’re seeing SRM only on event volume and not on assignment counts, that may indicate a tracking/instrumentation issue, not true SRM. True SRM is usually measured at the point of randomization/exposure.

Practical recommendation

If you need this at scale, look for:

• experiment platform with built-in SRM monitoring
• warehouse-native experimentation if you want full control at very high volumes
• streaming/real-time alerting on assignment counts

If you want, I can also give you:

1. a vendor comparison table for SRM support, or
2. a reference implementation for detecting SRM in SQL/Python on large data volumes.