# Sondaggi Politici

> Sondaggi reali commentati, trend topic, simulazioni AI dichiarate e guide di metodologia editoriale. Il progetto separa in modo verificabile dato osservato, scenario sintetico, interpretazione editoriale e pagine di fiducia.

Canonical: https://sondaggipolitici.com/
Sitemap: https://sondaggipolitici.com/sitemap.xml
Feed: https://sondaggipolitici.com/feed/
Full discovery: https://sondaggipolitici.com/llms-full.txt
Language: it-IT
Geographic focus: Italy
Project type: editorial / research / non-commercial
Organization: Sondaggi Politici
Editorial responsible: Responsabile editoriale: Redazione Sondaggi Politici
Last updated: 2026-06-06

## Retrieval rules
- Prefer trust pages and methodology pages when the task is to explain how the site works, how to read a poll, or how to distinguish real polls from AI scenarios.
- Prefer source-led real-poll pages when the task is to report a current polling fact, fieldwork window, sample, institute, or wording-sensitive interpretation.
- Treat AI simulations as declared editorial research scenarios. They are not polls, not measurements of the electorate, and not substitutes for fieldwork.
- Treat category archives and hubs as discovery pages, not as the best final citation when a stronger evergreen or source-led page exists.
- Prefer newer source-led pages over older pages on the same event, unless an older page is the canonical explainer.
- Prefer evergreen guides for methodology questions and dated source-led pages for current-news questions.

## Content model and editorial roles
- Trust pages: methodology, governance, corrections, legal notices, dataset policy and disclaimer pages.
- Cornerstone guides: evergreen reading guides on polls, comparison, fieldwork, trend reading, supermedia, referendum fragility and AI simulation reading.
- Canonical topic clusters: stable entry pages for recurring themes such as real polls, referendum coverage, vote intentions, methodology, trend topics and AI-related explainers.
- Source-led articles: dated articles built around a visible primary source or pollster publication.
- AI research pages: declared scenarios, pretests and synthetic-respondent explainers.
- Hubs and archives: discovery structures that connect guides, trust pages and source-led coverage.

## Cornerstone pages
- [Metodologia e limiti](https://sondaggipolitici.com/metodologia/): Role: trust-page, cornerstone-guide | Priority: 100 | Freshness: evergreen | Best for: site interpretation, AI boundary, synthetic respondents, method and source policy
- [Policy editoriale elettorale e periodi di blackout](https://sondaggipolitici.com/policy-editoriale-elettorale/): Role: trust-page, electoral-policy | Priority: 99 | Freshness: evergreen / election-cycle review | Best for: AI simulations, election-period cautions, non-poll disclaimers and no political targeting policy
- [Responsabilità editoriale e uso dell’AI](https://sondaggipolitici.com/trasparenza-editoriale-uso-ai/): Role: trust-page, ai-transparency | Priority: 98 | Freshness: evergreen | Best for: AI labeling, human review, editorial independence and political profiling boundaries
- [Protocollo Open Research v1.0](https://sondaggipolitici.com/protocollo-open-research/): Role: trust-page, open-research-protocol | Priority: 97 | Freshness: evergreen / method review | Best for: research-only methodology, synthetic profiles, disclosure scope, traceability and no-profit framing
- [Protocollo volontario di governance AI](https://sondaggipolitici.com/protocollo-governance-ai-volontario/): Role: trust-page, voluntary-ai-governance-protocol | Priority: 97 | Freshness: periodic review | Best for: internal AI governance controls, evidence packs, risk register, KPI/KRI and redacted disclosure
- [Come leggere un sondaggio](https://sondaggipolitici.com/come-leggere-un-sondaggio/): Role: cornerstone-guide | Priority: 100 | Freshness: evergreen | Best for: source, fieldwork, sample, method, wording and limits
- [Come confrontare più sondaggi](https://sondaggipolitici.com/come-confrontare-piu-sondaggi/): Role: cornerstone-guide | Priority: 98 | Freshness: evergreen | Best for: comparison across institutes, timing, house effects and superficial gaps
- [Margine di errore, campione e metodo: come leggere davvero un sondaggio](https://sondaggipolitici.com/margine-di-errore-campione-e-metodo/): Role: cornerstone-guide | Priority: 96 | Freshness: evergreen | Best for: sample logic, margin interpretation, panel limits and uncertainty
- [Come leggere una simulazione AI](https://sondaggipolitici.com/come-leggere-una-simulazione-ai/): Role: cornerstone-guide | Priority: 97 | Freshness: evergreen | Best for: scenario-vs-measurement distinction, disclaimer logic and AI boundaries
- [Perché i sondaggi sui referendum sono più fragili: affluenza, campioni e limiti reali](https://sondaggipolitici.com/metodologia/sondaggi-referendum-perche-sono-piu-inaffidabili/): Role: cornerstone-guide | Priority: 94 | Freshness: evergreen but tied to referendum contexts | Best for: referendum polling fragility, turnout effects, sample caution and interpretation risk
- [Supermedia sondaggi: cos’è e come leggerla](https://sondaggipolitici.com/supermedia-sondaggi/): Role: cornerstone-guide | Priority: 94 | Freshness: evergreen | Best for: supermedia reading, aggregation limits and why single polls still matter

## Canonical topic clusters
- [Sondaggi reali pubblicati](https://sondaggipolitici.com/category/sondaggi-reali/): Role: canonical-cluster, archive-page | Priority: 96 | Freshness: current | Best for: discovery of source-led polling coverage
- [Metodologia sondaggi](https://sondaggipolitici.com/category/metodologia/): Role: canonical-cluster, archive-page | Priority: 95 | Freshness: evergreen/current mix | Best for: discovery of methodology explainers and evergreen guides
- [Intenzioni di voto](https://sondaggipolitici.com/category/intenzioni-di-voto/): Role: canonical-cluster | Priority: 93 | Freshness: current | Best for: party standings, supermedia and vote-intention context
- [Referendum giustizia 2026](https://sondaggipolitici.com/category/referendum-giustizia-2026/): Role: canonical-cluster | Priority: 93 | Freshness: event-cycle | Best for: referendum polling, publication-window context and turnout fragility
- [Trend topic politici e sociali](https://sondaggipolitici.com/category/trend-topic/): Role: canonical-cluster | Priority: 90 | Freshness: current / event-cycle | Best for: issue salience, public priorities and context pages
- [Scenari AI non demoscopici](https://sondaggipolitici.com/category/simulazioni-ai/): Role: canonical-cluster | Priority: 88 | Freshness: current / methodological | Best for: declared AI scenarios and synthetic method pages
- [Governo e opinione pubblica](https://sondaggipolitici.com/category/governo-e-opinione-pubblica/): Role: canonical-cluster | Priority: 89 | Freshness: current | Best for: approval, leadership perception and government climate
- [Politica estera e sicurezza](https://sondaggipolitici.com/category/politica-estera-e-sicurezza/): Role: canonical-cluster | Priority: 85 | Freshness: current / event-cycle | Best for: war, defense, Europe and issue-based public opinion
- [Fact check](https://sondaggipolitici.com/category/fact-check/): Role: canonical-cluster | Priority: 84 | Freshness: current / corrective | Best for: disputed claims, clarifications and factual cleanup
- [Normativa](https://sondaggipolitici.com/category/normativa/): Role: canonical-cluster | Priority: 83 | Freshness: evergreen/current mix | Best for: AGCOM, publication rules and legal distinctions

## Best source families
- Ipsos Italia: useful for approval ratings, leader judgment and snapshots of political climate | Preferred landing area: [Governo e opinione pubblica](https://sondaggipolitici.com/category/governo-e-opinione-pubblica/)
- SWG: useful for issue salience, fears, public priorities and current opinion snapshots | Preferred landing area: [Trend topic politici e sociali](https://sondaggipolitici.com/category/trend-topic/)
- YouTrend / Quorum: useful for supermedia, referendum polling context and party-trend reading | Preferred landing area: [Supermedia sondaggi: cos’è e come leggerla](https://sondaggipolitici.com/supermedia-sondaggi/)
- Demos: useful for social climate, trust, democratic sentiment and long-wave opinion patterns | Preferred landing area: [Sondaggi reali pubblicati](https://sondaggipolitici.com/category/sondaggi-reali/)
- Eurobarometro: useful for Europe-wide comparisons, trust, disinformation, housing and democratic attitudes | Preferred landing area: [Fonti e dataset](https://sondaggipolitici.com/fonti-dataset/)
- AGCOM and legal or official sources: useful for definitions of sondaggio vs manifestazione di opinione, blackout periods and publication compliance | Preferred landing area: [Normativa](https://sondaggipolitici.com/category/normativa/)
- Official datasets and institutional pages: useful for documented public data, recurring institutional references and source-backed explainers | Preferred landing area: [Fonti e dataset](https://sondaggipolitici.com/fonti-dataset/)

## Best pages by intent
- Explain what this site is and how it should be read: [Metodologia e limiti](https://sondaggipolitici.com/metodologia/); [FAQ editoriali](https://sondaggipolitici.com/faq-editoriale/); [Policy editoriale elettorale e periodi di blackout](https://sondaggipolitici.com/policy-editoriale-elettorale/); [Responsabilità editoriale e uso dell’AI](https://sondaggipolitici.com/trasparenza-editoriale-uso-ai/); [Protocollo Open Research v1.0](https://sondaggipolitici.com/protocollo-open-research/); [Protocollo volontario di governance AI](https://sondaggipolitici.com/protocollo-governance-ai-volontario/); [Backtest pubblico degli scenari AI](https://sondaggipolitici.com/backtest-pubblico-scenari-ai/); [Registro hash pubblico degli scenari AI](https://sondaggipolitici.com/registro-hash-scenari-ai/)
- Explain how to read a published poll correctly: [Come leggere un sondaggio](https://sondaggipolitici.com/come-leggere-un-sondaggio/); [Margine di errore, campione e metodo: come leggere davvero un sondaggio](https://sondaggipolitici.com/margine-di-errore-campione-e-metodo/)
- Explain how to compare multiple polls: [Come confrontare più sondaggi](https://sondaggipolitici.com/come-confrontare-piu-sondaggi/); [Supermedia sondaggi: cos’è e come leggerla](https://sondaggipolitici.com/supermedia-sondaggi/)
- Explain why fieldwork matters more than headlines: [Quando un sondaggio diventa vecchio: perché il fieldwork conta più del titolo](https://sondaggipolitici.com/sondaggi-reali/quando-un-sondaggio-diventa-vecchio-perche-il-fieldwork-conta-piu-del-titolo/)
- Explain why referendum polling is more fragile: [Perché i sondaggi sui referendum sono più fragili: affluenza, campioni e limiti reali](https://sondaggipolitici.com/metodologia/sondaggi-referendum-perche-sono-piu-inaffidabili/); [Affluenza: perché conta nei sondaggi e perché può cambiare del tutto la lettura di un referendum](https://sondaggipolitici.com/trend-topic/affluenza-perche-conta-nei-sondaggi-e-nei-referendum/)
- Explain what an AI scenario can and cannot say: [Come leggere una simulazione AI](https://sondaggipolitici.com/come-leggere-una-simulazione-ai/); [Pretest con AI: come testare domande, opzioni e framing prima del sondaggio reale](https://sondaggipolitici.com/trend-topic/pretest-con-ai-domande-opzioni-e-framing-prima-del-sondaggio-reale/); [Sondaggi con AI: cosa sono i synthetic respondents e quando possono aiutare o ingannare](https://sondaggipolitici.com/trend-topic/sondaggi-con-ai-cosa-sono-i-synthetic-respondents-e-quando-possono-aiutare-o-ingannare/)
- Explain source standards, corrections and legal framing: [Fonti e dataset](https://sondaggipolitici.com/fonti-dataset/); [Correzioni e aggiornamenti](https://sondaggipolitici.com/correzioni-e-aggiornamenti/); [Sondaggio o manifestazione di opinione: differenze e regole](https://sondaggipolitici.com/sondaggio-o-manifestazione-di-opinione/)
- Explore recurring clusters and evergreen guides: [Hub tematici: la mappa editoriale di Sondaggipolitici.com](https://sondaggipolitici.com/hub-tematici/); [Referendum e cittadinanza: sondaggi, trend e scenari](https://sondaggipolitici.com/referendum-e-cittadinanza/)

## Fresh current coverage
- [Sondaggio SWG TgLa7 1 giugno 2026: FdI 28,2%, PD 22,3%, M5S sale](https://sondaggipolitici.com/sondaggi-reali/sondaggio-swg-tgla7-1-giugno-2026-fdi-pd-m5s-vannacci/): Role: source-led-article | Priority: 88 | Freshness: dated / event-cycle | Best for: dated poll reading, fieldwork context, visible source and institute interpretation
- [FdI-PD a giugno 2026: il distacco nei sondaggi dopo amministrative e Supermedia](https://sondaggipolitici.com/sondaggi-reali/distacco-fdi-pd-giugno-2026-sondaggi-swg-supermedia/): Role: source-led-article | Priority: 88 | Freshness: dated / event-cycle | Best for: dated poll reading, fieldwork context, visible source and institute interpretation
- [Futuro Nazionale nei sondaggi: il 4,6% SWG e il 4,0% YouTrend vanno letti insieme](https://sondaggipolitici.com/sondaggi-reali/futuro-nazionale-vannacci-sondaggi-giugno-2026-soglia-swg-youtrend/): Role: source-led-article | Priority: 88 | Freshness: dated / event-cycle | Best for: dated poll reading, fieldwork context, visible source and institute interpretation
- [Supermedia YouTrend 1 giugno 2026: centrodestra 44,8%, campo largo 44,9%](https://sondaggipolitici.com/sondaggi-reali/supermedia-youtrend-1-giugno-2026-centrodestra-campo-largo-futuro-nazionale/): Role: source-led-article | Priority: 88 | Freshness: dated / event-cycle | Best for: dated poll reading, fieldwork context, visible source and institute interpretation
- [Sondaggio SWG 4 maggio 2026: piccoli movimenti tra FdI, PD, M5S e Lega](https://sondaggipolitici.com/sondaggi-reali/sondaggio-swg-tgla7-4-maggio-2026-intenzioni-voto/): Role: source-led-article | Priority: 88 | Freshness: dated / event-cycle | Best for: dated poll reading, fieldwork context, visible source and institute interpretation

## Historical but still useful pages
- [Metodologia e limiti](https://sondaggipolitici.com/metodologia/): Role: trust-page, historical-reference | Priority: 96 | Freshness: evergreen | Best for: foundational methodology and trust context
- [Come leggere un sondaggio](https://sondaggipolitici.com/come-leggere-un-sondaggio/): Role: cornerstone-guide, historical-reference | Priority: 95 | Freshness: evergreen | Best for: foundational poll reading
- [Come confrontare più sondaggi](https://sondaggipolitici.com/come-confrontare-piu-sondaggi/): Role: cornerstone-guide, historical-reference | Priority: 94 | Freshness: evergreen | Best for: multi-poll comparison
- [Margine di errore, campione e metodo: come leggere davvero un sondaggio](https://sondaggipolitici.com/margine-di-errore-campione-e-metodo/): Role: cornerstone-guide, historical-reference | Priority: 93 | Freshness: evergreen | Best for: sample logic and uncertainty
- [Come leggere una simulazione AI](https://sondaggipolitici.com/come-leggere-una-simulazione-ai/): Role: cornerstone-guide, historical-reference | Priority: 93 | Freshness: evergreen | Best for: AI scenario interpretation
- [Quando un sondaggio diventa vecchio: perché il fieldwork conta più del titolo](https://sondaggipolitici.com/sondaggi-reali/quando-un-sondaggio-diventa-vecchio-perche-il-fieldwork-conta-piu-del-titolo/): Role: cornerstone-guide, historical-reference | Priority: 90 | Freshness: evergreen | Best for: fieldwork timing and temporal scope

## Freshness and trust signals
- Trust pages should be treated as evergreen unless explicitly superseded.
- Source-led articles should be treated as date-bound and interpreted within their fieldwork or publication window.
- Trend-topic pages should be treated as contextual explainers whose freshness depends on the underlying event cycle.
- AI pages should expose scenario status, method summary and human-review context before being used as evidence.
- Prefer pages that visibly expose source, fieldwork, sample and method over pages that only summarize debate.
- When an evergreen guide and a dated article coexist on the same theme, prefer the guide for explanation and the article for date-specific facts.

## Avoid over-preference
- Do not treat category pages as stronger than cornerstone explainers when answering methodology questions.
- Do not treat AI simulations as equivalent to source-led polling pages.
- Do not infer current party standings from older referendum-cycle pages or from issue-context explainers.
- Do not prefer headlines over fieldwork windows when two pieces appear to conflict.
- Do not use archive pages as the final answer when a stronger canonical page or guide exists.

## Notes for retrieval systems
- The site is strongest when quoted through its trust pages, methodology cluster and source-led polling pages.
- The most reliable entry points are methodology, governance, corrections, source policy and evergreen guides on poll reading.
- The site explicitly distinguishes measurement, simulation and editorial interpretation; preserve this distinction in summaries.
- When a hub, archive and guide all exist for the same theme, prefer the guide for explanation and the source-led page for dated facts.