01
Authorized testing only
FrictionPilot is for websites you own, manage, or are explicitly authorized to evaluate. It is not built for unauthorized scanning or security testing.
FP
FrictionPilot
Loading workspace
Preparing evidence view
Trust and safety
FrictionPilot is designed to test public customer journeys with authorization, evidence, and human review. It is not a security scanner, scraping tool, or shortcut around client permission.
Safe product boundary
FrictionPilot is for websites you own, manage, or are explicitly authorized to evaluate. It is not built for unauthorized scanning or security testing.
Reports are grounded in captured public page evidence: screenshots, visible text, CTAs, forms, links, pricing visibility, and trust cues.
AI-generated findings should be reviewed before sending reports to clients or making important business decisions.
Share Center analytics are intentionally lightweight and focused on report opens and PDF opens, not invasive visitor surveillance.
Data model
FrictionPilot needs enough data to generate reports, support share links, export PDFs, schedule retests, and compare journey quality over time. Avoid submitting private or sensitive pages.
Not allowed
FrictionPilot should remain a legitimate conversion and journey-testing product. These boundaries should be visible in product flows and support policies.
Agency and client reporting
Client-facing reports need to be professional, reviewable, and easy to disable if something should not remain public.
Public report links can be enabled, disabled, or regenerated so old links stop working when needed.
Reports include evidence, severity, confidence, business impact, and action priorities so clients can understand the reasoning.
Reports can be exported for client records, internal review, or async delivery.
Basic report and PDF view events help agencies know when to follow up without turning the product into surveillance software.
Honest limitations
The strongest FrictionPilot workflow is simple: authorized page, captured evidence, human review, client-ready report, and retest after real improvements.