Rothenhall × Napkin · Competitive diagnostic
Where seven creative agencies stand in the answer.
A comprehensive AI-visibility diagnostic, one score and five sub-scores per agency, on the Rothenhall framework. Switch between agencies to see each radar, its findings, the prioritised fixes, and the research behind the numbers.
Seven agencies, one measure.
Every agency was scanned the same way on the same day and scored on the Rothenhall AI Visibility framework. The headline finding: none is blocked from AI, and none appears in the shortlist an assistant builds for “best agency in Dublin”. Great work is not the gap. Being readable and being cited is.
0 / 7
named in the six Dublin agency lists AI assistants cite for that query.
The framework
Five weighted dimensions.
25pts
Machine access
Can the AI crawlers fetch and read the site?
25pts
Entity clarity
Does a model know which company this is?
20pts
Shortlist presence
Is it named when buyers ask?
20pts
On-page extractability
Can a model read and quote the pages?
10pts
Authority signal
Is there real-world proof to reward?
Invisible
The assistants cannot find, read, or identify the company.
Faint
Exists to the models but rarely surfaces. Strong companies with weak machine-visibility sit here.
Present
Appears for some queries that matter, and is read and understood correctly.
Recommended
Named, cited, and characterised well for the queries buyers ask.
Machine access, entity clarity, on-page extractability and heading hygiene are measured directly from each homepage and robots.txt on 14 September 2026. Shortlist presence and authority are assessed from the directory and press lists AI assistants cite for Dublin agencies, checked the same day. Sub-scores roll up to the composite by the weights above. AI answers are non-deterministic; we report tendencies and rates, never a guaranteed placement. A five-run test per buyer question is the standard next step to lock each shortlist rate.
Great work is not the gap. Being readable is.
Every engagement opens with a diagnostic like this one: the score, the five sub-scores, and the specific, reproducible reasons behind each. The findings are yours to keep.