AI search is changing discovery. The measurement should stay honest.
Reffed was built to make AI-search readiness understandable without turning uncertain signals into fake precision.
Most businesses need a starting point.
AI assistants, AI search experiences and retrieval systems are becoming another way customers discover products and services. But the category is full of dashboards, jargon and claims that blur the difference between a website signal and an actual model response.
Reffed starts with something concrete: the public website. It inspects evidence that can be observed, explains what is missing, and turns the gaps into a short list of things to improve.
A few rules we try to keep.
Evidence before confidence
Website observations, model-generated estimates and live provider responses are labeled separately. Unknown evidence stays unknown.
Short action lists
The useful outcome is not a giant dashboard. It is understanding the most important gaps and what to work on next.
Repeatable measurement
Reffed Watch reruns the same readiness framework so a business can compare its own site over time without pretending every score change proves causation.
Simple commercial terms
The audit is free. Watch is $59/month or $590/year. Billing is managed through Stripe and the subscription can be canceled without a sales call.
Audit first. Automation second.
The free audit and Reffed Watch work with any public website. Watch+Fix is a WordPress beta: Reffed prepares supported fixes, you approve each one, and you can undo any change.
We would rather ship a narrower product that can be verified than advertise a broad automation stack before it is commercially ready.
Operated from Canada.
Reffed is built for customers worldwide. For support, email support@reffed.ai.
Read the public methodology for the scoring and evidence framework.