Known bot, crawler and scanner patterns are separated from real visitor traffic.
Bot filtering and traffic integrity
Separate bots before reporting and trust real user behavior.
Testvoy evaluates known crawler signatures, headless automation, suspicious repetition, unsigned event context, abuse patterns and SRM drift in one traffic-quality pipeline so dirty traffic does not choose the winner.
Clean data, defensible outcomes
An A/B test is only as trustworthy as the traffic entering it.
Bots can distort more than visitor counts. They can affect allocation, conversion rates, funnel steps and decision confidence. Testvoy treats bot filtering as a layered traffic-integrity pipeline from event acceptance to SRM review, not a single list applied after reporting.
Headless-browser and automation-framework patterns are evaluated with context.
Events connect to validation for the correct project and experiment context.
Unexpected variant-allocation drift becomes visible before a decision.
Product capabilities
The core job Testvoy solves with this module.
This section shows what the module helps teams do, which risks it reduces and which decision it makes easier to defend.
Back to product hub01
Evaluate suspicious traffic with more than one signal
The approach adapts layered security principles used by large traffic platforms and evaluates user-agent, crawler and headless signals in context.
- Known bot and crawler signatures
- Headless and automation signals
- Separation of real and suspicious traffic
02
Protect event integrity during ingest
Project and experiment context, repeated events and public-endpoint abuse are checked before the report is built.
- Signed project and experiment context
- Duplicate and low-value repetition checks
- Abuse and rate-limit guardrails
03
Validate traffic allocation before deciding
SRM checks expose drift from expected variant allocation and warn teams before a winner is declared on weak data.
- Expected versus observed allocation
- Reviewable traffic-quality signals
- Clean-report pipeline
Where teams use it
Built for the experiment and optimization workflows teams run every week.
High-volume campaign traffic
Separate rising crawler, repetitive-request and low-quality traffic signals before reporting.
Public event endpoints
Limit repeated, contextually invalid or abusive events from inflating experiment metrics.
A quality gate before declaring a winner
Review the result before rollout or client presentation when SRM or traffic-integrity signals are not clean.
Decision path
The role this module plays in the experiment program.
Evaluate suspicious traffic with more than one signal
The approach adapts layered security principles used by large traffic platforms and evaluates user-agent, crawler and headless signals in context.
Known bot and crawler signatures
Headless and automation signals
Protect event integrity during ingest
Project and experiment context, repeated events and public-endpoint abuse are checked before the report is built.
Control pipeline
Six traffic-quality layers
How it works
Clearer setup, cleaner signals, faster decisions.
Testvoy positions this module to help teams move from experiment idea to report with the same data, context and decision language.
Layered control goes beyond a user-agent list
High-traffic platforms such as Google and Cloudflare use multiple signals rather than one rule. Testvoy adapts that layered thinking to experiment data through known signatures, automation patterns, event context, repetition, abuse guardrails and allocation quality. It does not replace those network platforms; it focuses on the integrity of experiment decisions.
- Known crawler and bot signatures provide the first separation
- Headless and automation signals are read in behavioral context
- Layers complement one another; no single signal becomes absolute truth
Signed event context is where report trust begins
Wrong project keys, wrong experiment context or imitated event flows can silently contaminate metrics. Testvoy separates public surfaces from application authority and validates event context during ingest.
- Project and experiment context can be checked
- Private application keys stay off customer sites
- Duplicate and low-value repetition enter the quality pipeline
SRM completes bot filtering at the decision layer
Even strong traffic filters cannot prevent allocation or tracking defects. Sample Ratio Mismatch compares expected and observed distribution to expose hidden bias before the decision.
- Expected allocation is compared with observed traffic
- A winner is not rushed when drift appears
- Quality signals are interpreted with funnels and statistics
FAQ
Fast answers before the decision.
Does Testvoy replace Cloudflare or Google bot protection?
No. Testvoy is not a network WAF or CDN replacement. It applies layered-security principles to A/B-test events, traffic quality and report integrity.
Which traffic types are checked?
Known crawler and bot signatures, headless automation, suspicious repetition, invalid or unsigned event context, abuse patterns and SRM drift can be evaluated.
Is filtered traffic visible?
The goal is not to silently erase traffic. It is to keep the clean-data pipeline and quality warnings reviewable for the decision.
Which plan includes bot filtering?
Core traffic integrity supports experiment measurement, while advanced bot and quality controls are listed separately in plan comparison.
A layered traffic-quality approach combining crawler, headless automation, signed-event, duplicate, abuse and SRM checks before experiment reporting.
Prevent dirty traffic from inflating decisions