Measurement completeness100%
160 / 160 planned answers were retained.
Vispertise™ audit demo
The Hotel Chelsea
New York · Hotel · Sep 17, 2026, 04:09
Vispertise™
overall score 84%
overall score 84%
Score calculation
How the 84% is built
The overall score combines seven measured signals: blind discovery performance, competitive position, source strength and visual readiness. Each signal contributes with its fixed weight; missing pillars are never redistributed.
Weighted contribution to the main score
15.7
+ 13.5
+ 14.4
+ 8.3
+ 13.0
+ 12.4
+ 7.1
=
84%
-
Mention visibility (MV)
78.7/100How often and how prominently the place appears in evaluated AI answers.
20% weight 15.7 points -
Recommendation strength (RS)
90/100How strongly the answers recommend the place when it is mentioned.
15% weight 13.5 points -
Claims: 124 correct, 5 incorrect, 136 unverifiable · Answers: 20 · Questions: 5. Source collection is partial. The score describes retained, verified evidence and does not establish complete website coverage. Unverifiable claims are reported separately.
Factual accuracy (ER)
96/100Accuracy = correct / (correct + incorrect). Verifiability = (correct + incorrect) / all eligible claims. Unverifiable claims are not counted as incorrect. Accuracy is relative to the available references, not independent fact certification. Percentages are equally weighted by persona, then provider; the summed counts separately describe the evidence base. Repeated claims are not necessarily distinct facts.
15% weight 14.4 points -
Question coverage (QC)
83.3/100How broadly the place appears across the tested question families.
10% weight 8.3 points -
Competitive position (CP)
86.4/100How the place performs against profile-matched alternatives in visibility, recommendations and share of appearances.
15% weight 13.0 points -
Source authority (SA)
82.8/100How often mentions are supported by credible, attributable public sources.
15% weight 12.4 points -
Visual readiness (VI)
71/100How well the retained image evidence supports accurate, decision-useful AI understanding.
10% weight 7.1 points
The headline percentage is rounded to a whole number.
The Hotel ChelseaOwned website and public research
Profile behind the audit
Who and what the AI benchmark evaluated
A The Hotel Chelsea egy igazolt, történelmi karakterű New York-i szálloda a Chelsea negyedben. Kínálata a körülbelül 19 négyzetméteres kompakt szobától a 158 négyzetméteres, két hálós lakosztályig terjed, továbbá összenyitható és akadálymentes szobákat is közöl. A helyszín több éttermet, lobbybárt, helyi vendégek számára is elérhető penthouse spát, szállóvendégeknek huszonnégy órás fitnesztermet és legalább hat név szerint felsorolt rendezvényteret kínál. A közzétett rendezvénykapacitások 10 és 100 fő között mozognak, de két éttermi állókapacitásnál hivatalos forrásellentmondás látható, a termek alapterületei és több elrendezéstípus külön kapacitásai pedig nem szerepelnek a felülvizsgált forrásokban.
This evidence-grounded profile determined the relevant audiences, locations and decision situations used to build the audit.
Automatic category profile
Grounded in public evidence
- Accommodation100.0%
- Reviewed sources
- 60
- Real decision situations
- 32
Services and offers
Shown: 11 / 11
- Szálláshely-szolgáltatás
- Akadálymentes vendégszobák
- Rendezvényhelyszín-bérlés
- Privát étkezés és étterem-kivásárlás
- Café Chelsea étkezés
- El Quijote étkezés
- Teruko japán étkezés
- Lobby Bar
- Spa- és wellnesskezelések
- Fitneszterem
- Parkolási kedvezmény
Geographic relevance
Profiled entity locationChelsea, New York, Egyesült Államok
Published service coverageHelyszíni szolgáltatások a Hotel Chelsea New York-i épületében
Other operating locationsChelsea, New York, Egyesült Államok
Language availability
- Published content: Angol
Possible audience reach — diagnostic only2
- A New York-i szálloda valószínűleg az Egyesült Államok más részeiről érkező utazókat is eléri, de ez nem közzétett szolgáltatási terület.
- A történelmi New York-i szálloda nemzetközi látogatókat is vonzhat, de a felülvizsgált források nem neveznek meg nemzetközi értékesítési piacot.
Profile source basis:
Owned website and public research
Official website · 60
Questions the reviewed sources do not yet answerThese are evidence gaps, not assumed negative facts. They are kept diagnostic and do not reduce the visibility score.5
- Conflicting informationA hivatalos rendezvényoldal 50 fős állókapacitást, a Café Chelsea oldala pedig az összekapcsolt terekre legfeljebb 60 fős koktélelrendezést közöl.
- Conflicting informationA hivatalos rendezvényoldal 60 fős állókapacitást, az El Quijote saját oldala 70 fős állókapacitást közöl a teljes kivásárlásra.
- Not found in reviewed sourcesA felülvizsgált hivatalos források nem közölnek négyzetméterben vagy négyzetlábban megadott alapterületet a rendezvénytermekhez.
- Not found in reviewed sourcesA felülvizsgált források nem közlik a csoportos szobafoglalások minimális vagy maximális szobaszámát.
- Not found in reviewed sourcesA felülvizsgált források nem közölnek a fizikai New York-i helyszínen túli, név szerint meghatározott szolgáltatási területet.
Relevant audiences8
From evidence to action
How this audit was built
Follow the six-stage path. Select a stage to see what was processed, what was examined and which retained output it produced.
01 / 06
Evidence-grounded scope
The submitted place is resolved against public evidence. Its identity, category, services, geography and decision situations determine who and what the benchmark may test.
Processing and examination units
- 60reviewed sources
- 8relevant audiences
Retained output
A retained entity, category and audience profile; uncertain claims remain marked as gaps rather than facts.
02 / 06
Audiences and questions
Relevant audiences are turned into repeatable decision situations and question families. The plan fixes the tested scope before answer collection begins.
Processing and examination units
- 5modelled personas
- 14question families
- 40tested questions
Retained output
A retained question plan that connects each question to an audience, language and decision context.
03 / 06
AI benchmark and evidence
Every planned question is measured across the configured AI systems and repetitions. Each answer is retained as a separate examination unit before interpretation.
Processing and examination units
- 4measured AI systems
- 40tested questions
- 160evaluated answers
- 1measurement repetitions
Retained output
A traceable answer set with question, language and system context for the Evidence explorer.
04 / 06
Seven signals and deep analyses
Answers, sources, matched alternatives and accepted images are evaluated separately. The place graph adds relationship context; it does not create an extra hidden score.
Processing and examination units
- 7weighted score signals
- 59graph entities and relationships
- 18publishable image evidence items
Retained output
Verified signal scores and, when every required measurement is available, their fixed-weight contributions and the rounded Vispertise™ overall score.
Each signal score × its fixed weight = contribution in points. The seven contributions are added; the headline percentage is the rounded total.
- MVMention visibility 78.7 × 20% = 15.7
- RSRecommendation strength 90 × 15% = 13.5
- ERFactual accuracy 96 × 15% = 14.4
- QCQuestion coverage 83.3 × 10% = 8.3
- CPCompetitive position 86.4 × 15% = 13.0
- SASource authority 82.8 × 15% = 12.4
- VIVisual readiness 71 × 10% = 7.1
Weighted total84.5%
05 / 06
AI query fanout feedback
Where returned evidence permits, we connect observed AI search queries to the original question, linked sources and owned-content coverage. These observations are incomplete: they do not reveal hidden reasoning, change scores or prove why a place was recommended.
Processing and examination units
Retained output
Retained search feedback and content-coverage findings. Supported gaps can inform implementation tasks; uncertain cases remain checks.
No verifiable search feedback is retained for this audit. This does not mean that no searches occurred.
Inspect questions, answers and search feedback
06 / 06
Customer outputs and implementation
The retained measurements are translated into readable analyses, inspectable evidence and evidence-bound implementation work without changing the original audit history.
Processing and examination units
- 4Retained output
- 958implementation tasks
Retained output
Overview, analysis views, Evidence explorer, prioritized tasks and available delivery packages.
Analysis views
Visibility, competition, source, visual and relationship views explain where each result comes from.
Open output →
Evidence explorer
The exact retained questions and answers can be searched behind findings and tasks.
Open output →
Implementation
Evidence-bound work items turn the strongest justified improvements into assignable action.
Open output →
Score and summary
The overview combines the seven weighted signals without inventing a second score.
Open output →