Evidence over hype. Notes on AI-answer visibility.
How Logres thinks about recorded AI answers: what was observed, what can be inferred, and what no one should promise.
July 21, 2026
What AI-answer visibility actually is
Buyers increasingly ask AI assistants who to trust. AI-answer visibility is the recorded evidence of what those assistants say — not a promise about what they will say next.
Read note →July 14, 2026
Evidence, not promises: how we measure AI visibility
Pinned questions, stored responses, four scores that are never averaged into one — and why an honest 'unavailable' beats a precise-looking number.
Read note →July 7, 2026
The four drivers behind AI-answer visibility
Publisher Credibility, Topical Authority, Structure & Extractability, and Sentiment & Reputation: how Logres organizes a diagnosis — and what these drivers are not.
Read note →June 30, 2026
What a Baseline Audit actually includes
The Baseline Audit is a paid, versioned diagnosis of how AI assistants represent your business. Here is what it measures, what you receive, and where it deliberately stops.
Read note →June 23, 2026
The claims we refuse to make
Logres sells measurement, diagnosis, and verified implementation. Here is the short list of things you will never hear us claim — and why that protects you.
Read note →June 16, 2026
Why AI answers vary — and what that means for measurement
Ask the same question twice and an AI assistant may answer differently. Variability is exactly why serious AI-answer measurement needs pinned suites and stored responses.
Read note →June 9, 2026
When AI describes your business wrong
Being named is not enough. If the recorded description is outdated or incorrect, it quietly works against you. That is what the Accuracy dimension measures.
Read note →June 2, 2026
Named vs. cited: two different kinds of presence
Being named in an AI answer and being cited as a source are different events with different fixes. Visibility and Authority measure them separately — on purpose.
Read note →May 26, 2026
Delivered is not published is not verified
In remediation work, three words vendors blur together each mean something different. Keeping them distinct is how you know what you actually paid for.
Read note →May 19, 2026
What monthly monitoring can and cannot tell you
A changed score between months is an observation to review, not proof of cause. Here is how to read monitoring honestly — and get real value from it.
Read note →May 12, 2026
The shortlist is the new search result
Buyers increasingly meet your business as one name inside an AI-generated answer. That first impression happens before any click — and most dashboards never see it.
Read note →May 5, 2026
Reading a competitive landscape without inventing numbers
A competitive read answers three questions from recorded evidence: who appears instead of you, on which questions, and leaning on which sources. Anything beyond that is theater.
Read note →