Strainer · AI Search Ranking

engine

What are we measuring, and for whom?

Name the company, describe the buyer. We put live questions to a frontier model exactly as that buyer would ask them, then read back who it recommends.

Company under review

Company name

Website

Category and competitor set are inferred when you run.

Market cut

Industry you are selling into

Geography

Language

Customer size

Ideal customer profile

Specific prompts to track — one per line, optional

Website analytics

Drop in a screenshot of your referrers/sources view — GA4, Vercel, Plausible, anything. We read the numbers straight off the image. Optional, but it's what turns a ranking into a pipeline number.

Drop a screenshot, click, or paste PNG or JPEG · up to 4 images · the referrers table is the one that matters

Why the referrers table. Assistants don't browse the web, they read a search index — Claude reads Brave, ChatGPT read Bing. So search.brave.com and bing.com referrals are the fingerprint of assistant-mediated discovery, and chatgpt.com / claude.ai / perplexity.ai referrals are agent traffic outright. We use them to estimate what share of your sessions are already agent-driven.