Earned Media: 239% Lift in AI Citations

Why earned media is becoming the decisive visibility channel in the AI era

Generative search systems such as ChatGPT, Perplexity, Google AI Overviews, and Gemini increasingly answer questions themselves instead of just listing links. Brands named as a source in these answers gain visibility. Brands that are missing simply do not exist for a growing share of users. The starting point is sobering: according to an international study, around 90 percent of the brands examined had zero mentions in AI search (Victorious, Q1 2026, 177 brands). The few visible brands share one pattern: they appear in editorial third-party sources, not just on their own website.

This is exactly where earned media comes in. Earned media is editorial attention granted by third parties that a company neither publishes itself (owned media) nor pays for (paid media). It includes coverage in trade media, mentions on industry portals, analyst pieces, forum and community discussions, and citations in studies. Large language models give these sources priority because they are considered independently validated.

The data is clear. In Muck Rack's latest analysis, 84 percent of all AI citations came from earned media (May 2026, more than 25 million links from ChatGPT, Claude, and Gemini across 17 industries). Paid and advertorial content, by contrast, accounted for only 0.3 percent of cited sources. Across several editions of the study, the earned media share ranged between 82 and 89 percent, so it was stable. In practice, this means that if you want AI visibility, you have to earn coverage, not buy it.

The DACH region has an additional accelerator. In Austria, 30 percent of companies with 10 or more employees already used AI technologies in 2025, compared with 20 percent in 2024 (Statistik Austria, ICT survey 2025). AI is therefore also reaching B2B research and supplier evaluation in the domestic market. Visibility in AI answers is no longer a question for the future. It is already a competitive factor.

The citation lift mechanism: how earned distribution raises citation frequency

The central lever of earned media is distributing coverage across many authoritative domains. This effect is called citation lift: the same content, syndicated across a network of editorial third-party portals, is cited far more often in AI answers than the identical story on your own website alone.

Stacker and Scrunch quantified the mechanics behind it in two studies. The first study, from December 2025, analyzed eight articles from different industries and evaluated 944 prompt-platform combinations across five leading LLMs, which corresponded to around 189 unique prompts tested. It examined how citation frequency changes when a story is distributed through third-party news portals in addition to the owned domain. With only the brand's original page, the stories were cited in around 8 percent of answers; including the distributed versions, in around 34 percent. Stacker reports this as a citation lift of 325 percent.

The follow-up study from March 2026 confirmed the effect on a broader basis. Stacker and Scrunch evaluated 87 stories from 30 clients and ran more than 2,600 prompts on eight AI platforms. The stories distributed through Stacker achieved a median citation lift of 239 percent over owned-only publication. 97 percent of the distributed stories received at least one AI citation, compared with 82 percent of owned content. Stacker itself sells this distribution as a service to brands, so both studies are vendor research.

The mechanism can be described in three steps:

• Redundancy across domains: When a statement appears on many independent, trustworthy sites, the likelihood increases that a language model rates it as reliable and anchors it in its answer.
• Co-citation and context: When a brand is repeatedly named in the same topical context, the connection between brand and topic solidifies in the model. That increases the chance of being cited on exactly these topics.
• Crawl and index coverage: More authoritative mentions mean more entry points for the crawlers that feed AI systems. A single owned page can be overlooked; a distributed coverage network hardly ever is.

This yields a clear principle for implementation: a study or a piece of data should reach as many relevant, editorially run media outlets as possible through outreach, instead of just sitting on your own blog. Distribution beats a single placement.

Brand mentions beat backlinks: the new foundation of AI visibility

Classic link building remains relevant for organic Google search. Top-ranking pages still have 3.8 times more backlinks than positions 2 to 10 (Backlinko, analysis of 11.8 million search results). At the same time, the same analysis shows how unevenly links are distributed: around 95 percent of all pages have zero backlinks. For most companies, the pure link race is therefore neither won nor realistically winnable.

For AI visibility, the weighting shifts anyway. An Ahrefs analysis of 75,000 brands found that the three strongest correlations with brand visibility in AI Overviews are all off-site factors: brand web mentions at r=0.664, brand anchors at r=0.527, and brand search volume at r=0.392. Backlinks, by contrast, correlated at only r=0.218. Pure brand mentions therefore correlate roughly three times more strongly than backlinks with presence in AI answers.

The strategic consequence is far-reaching: an unlinked brand mention in a trade article can be more valuable for AI visibility than a placed link. Language models process text and recognize entities; they do not primarily follow hyperlinks. SEO teams used to try to turn every mention into a link. For AI answers, what counts first is that the brand is named in the right context at all. That lowers the barrier in outreach considerably, because editors grant a mention by name more readily than a dofollow link.

Entity SEO as the basis: building the brand as a consistent entity

Earned media only has its full effect when the language model clearly recognizes the mentioned brand as an entity and can distinguish it from similarly named players. Entity SEO creates exactly this foundation. The goal is a consistent, machine-readable brand identity that produces the same picture across all mentions.

Four building blocks form the foundation:

• Structured data: Organization schema with the sameAs attribute links your own domain to authoritative references such as official profiles and directories. This makes the entity's identity explicit.
• Consistent core data: Company name, address, business description, and founding details must be identical across your website, directories, and media profiles. Contradictions weaken entity recognition.
• Build co-citations: Repeatedly naming the brand together with its core topics in independent sources strengthens the topical association. This is entity work that comes directly from earned media.
• Set authority anchors: Entries in established, industry-relevant data sources give the Knowledge Graph stable reference points for the brand.

The 2024 Google API leak supports the view that sitewide authority is a real signal. The leaked documentation comprised 2,596 modules with 14,014 attributes and confirmed, among other things, an attribute called siteAuthority as well as the NavBoost system with click data such as goodClicks and badClicks (Search Engine Land). A brand clearly established as an authoritative entity benefits in both worlds: in classic search and in AI answers.

GEO in practice: preparing content for LLM citations

Generative Engine Optimization (GEO) is the targeted preparation of content so that it gets cited in generative answers. The method is backed by research. The foundational paper by Aggarwal and colleagues shows that GEO can increase visibility in generative answers by up to 40 percent (arXiv, KDD 2024). Effectiveness varies by domain.

Several practical principles follow from this for earned media content and your own website:

• Work with numbers and data: Concrete, sourced statistics are attractive citation building blocks for language models. Your own surveys provide material that is not available elsewhere.
• Enable direct quotes: Clear, self-contained statements are easier to extract and reproduce than convoluted arguments.
• Cite sources transparently: Traceable source citations increase the perceived trustworthiness of content.
• Structure clearly: Meaningful headings, short paragraphs, and unambiguous answers to specific questions make machine extraction easier.

The order matters: GEO optimizes the form of the content, and earned media ensures its authoritative distribution. A perfectly structured text that only lives on your own domain falls short of its potential. Only the combination of citable formatting and broad earned distribution delivers the full effect.

DACH playbook: earned media and digital PR for B2B in Austria

The Austrian and DACH markets require their own approach. Standard international templates rarely work here. What matters are domestic authority sources and a relationship-based approach.

• Austrian Economic Chamber and associations: The WKO and its trade associations offer trade publications, industry newsletters, and directories. A mention there is a strong local authority anchor.
• Austrian business and trade press: Established business media and industry-specific trade titles are the core targets for editorial coverage that carries weight in the domestic market.
• Data and study PR: Your own surveys with an Austrian angle, for example on industry trends or regional market data, are the most effective way to spark editorial interest and deliver citable figures.
• Journalist request platforms: Services such as Qwoted or Featured connect journalists' requests with experts. They are a good way to get into trade articles as a quoted voice.
• LinkedIn: The network distributes expert articles in a professional environment and creates brand mentions that in turn feed into entity building. Since 2023, XING has been repositioning itself from a general professional network to a job network, which makes LinkedIn the more relevant channel for expert articles.

The market shows why this focus on the Google ecosystem still makes sense: in Austria, Google holds 81.87 percent search engine market share, Bing 9.01 percent, and DuckDuckGo 2.75 percent (StatCounter, May 2026). Google AI Overviews and organic search therefore remain the main stages on which earned media work pays off.

Reviews and reputation as an earned media component

Reviews are a distinct, often underestimated form of earned media. They are created by third parties, are not bought, and give language models signals about a brand's quality and reputation. For questions about products and vendors, AI systems often draw on review platforms because these bundle structured third-party opinions.

Three aspects matter in practice. First, volume and recency: a steady stream of recent reviews is more credible than a stale set of old ones. Second, breadth: a presence on several relevant platforms increases the likelihood of being used as a source in different AI contexts. Third, response behavior: replies to reviews generate additional brand-related text and signal engagement.

Reputation signals feed directly into E-E-A-T, that is, experience, expertise, authoritativeness, and trustworthiness. Independent third-party reviews are exactly the external validation these criteria address. They complement editorial coverage and round out a brand's earned media profile.

Legally compliant monitoring and outreach in the DACH region

Brand mention monitoring and PR outreach regularly process personal data, such as the names and contact details of journalists or of the authors of mentions. In Austria and Germany, this is not permitted without a sound legal basis under the GDPR; in Switzerland, the revised Federal Act on Data Protection applies. Under the GDPR, three points are central.

First, the legal basis: for monitoring and processing contact details in a B2B context, legitimate interest under Art. 6(1)(f) GDPR is often an option. It requires a documented balancing of interests between your own PR purpose and the rights of the data subject. Second, processing on your behalf: if you use monitoring tools that process personal data, you need data processing agreements with the providers under Art. 28 GDPR. Third, data subject rights: access and erasure requests must be fulfillable on time, which requires traceable documentation of your data flows.

The data protection legal basis does not replace the consent that telecommunications and competition law require for advertising. In Austria, advertising emails without prior consent are prohibited under § 174(3) TKG 2021, including those sent to businesses. Advertising calls there also require prior consent under para. 1. In Germany, § 7(2) no. 2 UWG requires prior express consent for advertising emails; for advertising calls to businesses, no. 1 requires at least presumed consent. In Switzerland, Art. 3(1)(o) UWG also requires prior consent for mass advertising by email. A cold advertising email is therefore ruled out as a first contact. That leaves LinkedIn messages and postal letters, and in Germany and Switzerland, under narrow conditions, the phone.

In practice, data minimization is the way to go: collect and store only the data you actually need for each outreach. Contact should have a clear professional connection instead of being broad mass outreach. That increases the response rate among DACH newsrooms, which routinely ignore generic requests. A professional connection does not change the consent requirement for advertising emails.

Common mistakes with earned media for AI visibility

• Publishing only on your own domain: If you publish studies and content exclusively on your own blog, you miss out on the citation lift that distributed coverage delivers.
• Relying on paid media: Paid and advertorial content accounts for only 0.3 percent of AI citations. Putting budget into advertorials instead of earned coverage misses the goal.
• Putting links ahead of mentions: For AI visibility, a brand mention in the right context counts for more than the link, since brand mentions correlate roughly three times more strongly than backlinks. Outreach that rigidly insists on dofollow links leaves opportunities on the table.
• Inconsistent entity: Differing company names or descriptions across sources make entity recognition harder and dilute the effect of every mention.
• Not making content citable: Without a clear structure, sourced figures, and extractable statements, even widely distributed content remains hard to cite.
• International templates in the DACH region: Austrian and German newsrooms reliably pass over generic mass requests, which are often in English.

Metrics and measurement: making AI citations visible

Earned media for AI visibility needs its own metrics, because classic link and ranking metrics do not fully capture the effect. Four measurement dimensions are central.

• AI citations and mentions: Does the brand appear in answers from ChatGPT, Perplexity, and Gemini and in Google AI Overviews? You track how often it is named across a defined set of relevant prompts.
• Share of voice: What share of all vendors named on a topic does your own brand account for, compared with competitors?
• Brand mention monitoring: Where, how often, and in what context is the brand mentioned, linked or unlinked? These mentions are the early indicator of future AI citations.
• Citation attribution: Which specific sources and coverage placements show up as cited evidence in AI answers? This reveals which earned media activity actually works.

A recurring measurement rhythm is important because AI answers are volatile: the same question yields different cited sources over time. A defined prompt set, queried regularly and consistently, reveals trends instead of snapshots. Only this continuity lets you assess how individual earned media campaigns contribute to AI visibility.

Further reading

In the AI era, earned media is not a nice-to-have for PR but the central infrastructure for visibility in generative answers. Effective practice connects three levels: a cleanly established brand entity as the foundation, content made citable according to GEO principles, and broad, editorially earned distribution through authoritative third-party sources. In the DACH region, add a focus on domestic trade media, chambers such as WKO and IHK, and associations, backed by privacy-compliant monitoring and relationship-based outreach. Connect this topic with related building blocks such as digital PR, unlinked brand mentions, entity SEO, and measuring AI visibility to turn individual measures into a consistent system.