Search Truth

Third-party rank and traffic estimates are models of the web. Google Search Console is what Google actually reported for your verified property. Those two will disagree sometimes. That is normal, not a mystery you should ignore.

Search Truth puts estimated ranks beside real Search Console performance, including brand versus non-brand splits, so you do not build strategy on fantasy numbers for your own site. It requires GSC connected. It is part of domain intelligence, not a claim that every estimate everywhere is useless. Feature overview: Search Truth (/features/search-truth).

Why estimates and Search Console disagree

Different corpora. Different locations. Personalization. How brand queries get captured. Different click models. Classic estimate tools (think the big third-party suites people compare to Semrush or Ahrefs) often show opportunity-shaped numbers that help size markets and competitors. GSC shows observed demand for properties you verified.

None of that makes third-party data "always false." It makes it a different instrument. Search Truth exists so you see both on one desk instead of arguing from two screenshots.

What Search Truth shows

Side-by-side estimated ranks versus real GSC performance for the property. Brand and non-brand splits so a pretty branded CTR does not hide a weak non-brand story. Drift signals when the gap between estimate and observed gets worth a second look. Domain alerts can include Search Truth drift so hygiene does not depend on you remembering to open the page every Monday.

What you need connected

Google Search Console on the project. GSC reads are free. Setup and performance context live across projects (/docs/projects) and analytics (/docs/analytics). Domain intelligence refresh keeps the join up to date after setup. Pasting a domain alone is not magic auto-ingest of Search Truth without the connections and refresh path in place.

Keyword cannibalization (related)

Cannibalization work looks at queries where multiple of your own pages compete. When you are investigating a pair, you can go deeper with content-similarity style compare. Search Truth and cannibalization meet when the estimate, the ranking URL, and the GSC-performing URL tell different stories: wrong URL winning, split demand, or an estimate that never matched how Google actually served your site.

How the agent uses Search Truth

A natural question is exactly the title of this page: "Why don't my tool estimates match Search Console?" The agent should open Search Truth and GSC, explain the gap in plain language, and only use numbers the connection returned. Same analysis can surface through MCP (/docs/mcp). Agent judgment rules: the agent (/docs/the-agent).

How this changes prioritization

For your own site, prefer GSC-backed striking distance and CTR work when you choose what to fix next. Follow-ons: opportunities (/docs/opportunities) and CTR gaps (/docs/ctr-gaps). Use third-party data for competitor gaps and markets you cannot verify in GSC (/docs/competitor-research). Do Next can reflect verified opportunity so the queue is not pure estimate theater (/docs/opportunities, and Do Next in the Visibility Engine loop on /docs/geo).

Costs

GSC reads are free. Domain intelligence and third-party pulls may be paid, with cost shown first. See credits (/docs/credits). No unique Search Truth sticker price on this page.