GEO Insights

Guides, insights, and strategies to dominate AI search.

GuidesAugust 13, 2026

Appearing in an AI Overview: 10 sectors measured

Before trying to appear in an AI Overview, find out whether Google produces one in your sector at all. The rate runs from 89.1 % in online banking to 41.9 % in luxury watchmaking. The format retained changes too: the editorial article dominates everywhere except in restaurants, where the directory listing comes first. The number of sites to work doubles from one sector to another, and so does the observation period needed before judging a result.

7 min
Read article
GuidesAugust 12, 2026

Choosing a GEO agency in Switzerland: six questions

Two GEO offers at the same price can measure very different things: not the same engines, not the same questions, not the same markets. Six questions surface the gap before you sign. They cover the accuracy of the information about your company, what will actually be measured, the role of classic search optimisation, which Swiss sources are targeted, and how long the observation runs. The reference points come from Swiss Atlas, the observatory Repliq runs daily across ten sectors.

10 min
Read article
Case StudiesJune 30, 2026

GEO 2026: how ChatGPT and Perplexity cite you

Your site can rank first on Google and still be absent from ChatGPT or Perplexity answers. That is normal: answer engines do not rank links, they cite a few sources to compose an answer, and each one has its own logic. Repliq recorded the pages that ChatGPT, Google AI and Perplexity cite most in Switzerland, then dissected their content. 4 characteristics come back page after page: a title that echoes the question, a high density of numeric data, a visible update date, an identified author. On top of that comes a major difference between engines: ChatGPT avoids commercial comparison pages but cites reviews (Trustpilot) and official sites, whereas Google and Perplexity readily pick up blogs and comparison sites. This guide names the pages, gives the numbers, and ends with an applicable checklist.

10 min
Read article
InsightsJune 9, 2026

Share of Model: definition, calculation and reliability

Share of Model measures how often a brand appears in answers generated by answer engines (ChatGPT, Google AI Mode, Google AI Overview, Perplexity, Gemini) across a defined set of questions. Its definition is broadly agreed on, its calculation method far less: the choice of denominator, the exclusions and the deduplication rule shift the result by tens of points. Across 8,261 measured answers, the metric proves stable week to week and markets remain fragmented.

15 min
Read article
Case StudiesMay 19, 2026

Swiss hospitality faces a new invisible ranking: the one set by AI

Which Swiss hotels appear in the recommendations of ChatGPT, Gemini and Perplexity, and on what data foundations do these choices rest? This preview summarises the first mapping of AI visibility for the destinations of Geneva, Lucerne, Zermatt and Interlaken. The study analyses each property's share of voice and the weight of authority channels (official websites, Michelin Guide, Wikipedia, consortiums) in algorithmic decisions. The full report, including rankings by stay category and optimisation levers, is available now. Access at the bottom of this page.

13 min
Read article
InsightsMay 18, 2026

Your Brand Now Has an AI Reputation. Are You Measuring It?

AIs do not passively reflect a brand's reputation. They reconstruct it. From customer reviews, articles, forums, comparators, competitor content and sometimes old sources, they produce a synthesis that already influences your customers, candidates, partners and investors. This article walks through five questions to steer this new reputation: what AIs say about you, the sources they use, the queries they trigger, the competitors they associate with your brand, and the alignment between your real strategy and your generated image.

10 min
Read article
Case StudiesMay 7, 2026

From search to recommendation: who answers for your brand in AI?

AI is turning information search into a direct recommendation where users expect arbitration. Our study of 8,800 citations shows that leading Swiss brands today control only 5 to 12% of their own narrative inside these interfaces. This guide unpacks how challengers like rappn and Magic Heidi break through with structured content, and how established brands can reclaim the conversation by becoming reference sources for the algorithms.

9 min
Read article
Blog - Repliq | GEO Guides & AI Visibility | Repliq