Brand Signals and Entity SEO: Knowledge Graph and AI Authority

Why brand signals and entity SEO are decisive for B2B in the DACH region

Search engines now capture meanings as well as character strings. Google itself describes its knowledge network, the Knowledge Graph, as a system that has amassed over 500 billion facts about five billion entities. An entity is a uniquely identifiable thing: a company, a person, a product, a place. This very shift from keywords to entities is the core of entity SEO. If you sell B2B services in the DACH region, you compete for search terms and, at the same time, on whether your brand is recognized as a reliable, clearly defined entity.

The impact extends beyond classic search. In Austria, Google handles around 81.87 percent and Bing around 9.01 percent of search queries (as of May 2026), while 8.69 million people are online, at an online penetration of 95.3 percent (as of January 2025). On top of that come AI answer systems such as ChatGPT, Perplexity, and Google AI Overviews, which either cite brands or don't. Brand signals and a clean entity representation increasingly decide whether a company is visible in both worlds: in the blue list of links and in the generated answer.

Basics: entities, the Knowledge Graph, and the shift from strings to things

An entity is a real or conceptual object with a unique identity and relationships to other entities. In the Knowledge Graph, a Viennese machinery manufacturer is not stored as a sequence of letters but as a node with attributes (founding year, location, industry, management) and edges to other nodes (suppliers, chamber of commerce, products). Google describes this inventory as a system that pulls together facts from the web and from open and licensed databases, bringing together over 500 billion facts about five billion entities.

In practice, this changes the logic of optimization. Instead of tuning a page to a single keyword, the task is to make one dominant entity per URL clear and to describe it in a semantically consistent way. The principle is things, not strings: the search engine should recognize what a page is about even if the exact search term does not appear in the text. Entity SEO thus connects three levels: the machine-readable description of the brand (structured data), the connection to established knowledge sources (sameAs, Wikidata), and the topical depth of the content, which gives the entity meaning and context.

What the Google API leak reveals about brand signals

In May 2024, thousands of internal documents from Google's Content Warehouse API became public. They confirm that Google processes far more nuanced brand signals than it officially acknowledged for years. Several attributes named in the leak are relevant to entity SEO, even though their exact weighting is not apparent from the documents.

• siteAuthority: A documented, domain-wide authority signal. It refutes the earlier claim that Google does not use a site-wide authority value and suggests that the perceived strength of a brand is relevant to rankings.
• NavBoost and click signals: A system that evaluates user interactions such as clicks and selections in the search results. Brand awareness that leads to targeted clicks on your own brand can have a positive effect here.
• Brand mentions and links: Attributes that assess link quality, source pages, and context. They support the assumption that, alongside the number of backlinks, the way the brand is embedded in trustworthy sources also counts.

The lesson for B2B companies in the DACH region: brand signals are not a soft add-on but an input that Google measurably processes. A strong, consistently named brand that appears in relevant sources and is actively searched for sends exactly the signals the leak describes as closely tied to rankings.

Winning and managing a knowledge panel: entity home and sameAs

A knowledge panel is the visible result of an entity that is established in the Knowledge Graph. You cannot order one directly. It appears when Google can connect enough consistent facts about an entity from trustworthy sources. A strategic build-out rests on a few proven components:

• Define an entity home: A unique, canonical home for the entity, usually the About Us or company page. It is the reference point that all other signals point to.
• Maintain a Wikidata entry: Wikidata is an open, machine-readable knowledge base that Google uses as a source. A maintained, well-referenced entry makes the entity uniquely identifiable for machines. A full Wikipedia article, by contrast, is subject to strict notability criteria that many SMEs do not meet.
• Google Business Profile: For companies with a physical location, a central anchor of trust with a verified address, category, and contact details.
• sameAs links: The sameAs property in schema markup declares the brand's official profiles (website, LinkedIn, XING, Wikidata, industry directories) as belonging together. This way, the company itself confirms that these scattered sources refer to the same entity.

The common denominator is consistency. The fewer contradictions there are in name, address, industry, and description across all sources, the easier it is for Google to connect them to a single, trustworthy entity.

Entity salience and semantic presence

Entity salience describes how central an entity is to the content of a page. A URL built around one clearly dominant entity sends a stronger signal than a page that mixes several topics. In practice, this means defining one main entity per URL and aligning the text consistently with it.

Three mechanisms strengthen semantic presence. First, co-occurrence: when the brand is regularly mentioned in the context of relevant technical terms and related entities, the search engine learns which topics it belongs to. Second, internal linking: contextual links between topically related pages concentrate meaning and build a recognizable topic cluster. Third, topical authority: a domain that covers a topic in breadth and depth is perceived as a competent entity for that field. This is where structured internal linking pays off twice, because it increases both salience and topical cohesion.

Structured data as the foundation of entity SEO

Structured data translates brand facts into a format that machines can read unambiguously. The dominant format is JSON-LD, which according to the HTTP Archive Web Almanac is now used on 41 percent of pages (34 percent in 2022) and is considered the fastest-growing format. For entity SEO, three schema types are especially central.

• Organization schema: Describes the company as an entity with name, logo, founding data, contact details, and above all the sameAs property, which points to the official profiles.
• Person schema: Makes executives or expert authors tangible as entities in their own right, which strengthens the connection between the brand and the people who act for it.
• LocalBusiness schema: Relevant for companies with a physical location, because it makes address, opening hours, and service area machine-readable and thus supports local citations.

NAP consistency matters most: name, address, and phone number must be identical across the website, Google Business Profile, industry portals, and, in Austria, the WKO Firmen A-Z. If the company is registered in the Firmenbuch (company register, Austria) or the Handelsregister (commercial register, Germany and Switzerland), the company name and registered office on the website and in Wikidata should match the register entry. Contradictory information makes it harder to connect the entity and weakens the trust signal. JSON-LD is the format of choice because it can be kept separate from the visible content and Google names it as the preferred markup format.

Unlinked brand mentions and co-citations as a trust signal

Not every valuable mention is a link. Unlinked brand mentions, meaning mentions of the brand name without a clickable reference, count as a trust signal in their own right. They show search engines and language models that people are talking about a brand, regardless of whether an actual link was placed. Combined with co-citations, meaning a brand being named alongside established industry leaders or recognized sources, they create a picture of affiliation and authority.

For AI systems, this logic is especially pronounced. Language models work with patterns from large volumes of text, in which the brand name itself and its context often carry more weight than a single hyperlink. An international study on so-called ghost citations shows how wide the gap between linking and naming is: according to the study, around 62 percent of AI citations are ghost citations, in which the source is linked but the brand name is not mentioned in the answer. Overall, the dataset reaches a citation rate of 74.9 percent compared with a mention rate of only 38.3 percent. The takeaway for brands: being cited is not enough; the name has to appear in the text, and that is exactly what consistent brand mentions on the open web encourage.

AI citation optimization (GEO): entity authority in AI answers

Generative engine optimization (GEO) carries entity SEO over into the world of AI answers. A widely cited academic paper shows that targeted optimization can increase visibility in generative answers by up to 40 percent. Entity authority and brand signals are a major lever here, because AI systems prefer to draw on sources they classify as established and topically relevant.

The relationship to classic search is mixed. According to an Ahrefs analysis (March 2026), only around 38 percent of AI Overview citations now come from the organic top 10, down from about 76 percent in July 2025. Ahrefs suspects that the cause is Google increasingly citing sources that appear in the results of follow-up queries (query fan-out). BrightEdge, by contrast, measured that the overlap between organic rankings and AI Overview citations rose from 32.3 percent to 54.5 percent (plus 22.3 percentage points) between May 2024 and September 2025. According to BrightEdge, the increase comes mainly from pages in positions 21 to 100. Under BrightEdge's methodology, only 16.7 percent of citations come from the top 10, significantly less than in the Ahrefs analysis.

So a top-10 ranking does not guarantee a citation. Your chances are probably better if you are also organically visible for the sub-questions of a topic and are understood as an entity. On top of that, well-known brands show remarkable stability. In transactional categories such as retail, travel, and tech, brand agreement across five AI engines is 88 to 97 percent according to BrightEdge, which shows how strongly established entities dominate across engines.

DACH-specific sources for entity and brand signals

In the DACH region, some sources are especially effective for establishing an entity, because they are regarded as trustworthy in the region and deliver consistent facts.

• Firmenbuch and WKO Firmen A-Z in Austria: The Firmenbuch is the official company register with company name, legal form, and registered office. The Firmen A-Z is the member directory of the Austrian Economic Chamber (WKO). The responsible authority transmits the address registered for the trade license, while businesses maintain their contact details, logo, images, and videos themselves. If the details match the Firmenbuch, the Firmen A-Z is a strong NAP and entity anchor for Austrian businesses.
• Handelsregister in Germany and Switzerland: In Germany, the commercial register entries are available through the joint register portal of the federal states; in Switzerland, through the central business name index Zefix. As with the Firmenbuch, the company name, legal form, and registered office from the register are the reference that the website, Wikidata, and directories should align with.
• Industry portals and local citations: Regional and specialist directories reinforce address and industry consistency and deliver recurring, identically worded brand mentions.
• Trade media: Mentions in relevant industry and business media act as co-citations alongside established players and strengthen perceived authority.
• LinkedIn and XING: In B2B, LinkedIn is the central identity platform. According to the DACH long-term study “Social Media in der B2B-Kommunikation” (social media in B2B communication), around 95 percent of the surveyed B2B companies in the DACH region use LinkedIn (2024). XING repositioned itself as a job network in October 2023; a maintained profile still makes sense as a supplement for the German-speaking job market. Both profiles can be linked to the entity via sameAs.

Here, too, consistency is the common thread. Each of these sources should carry the same brand name, the same address, and the same short description so that Google and AI systems can attribute every trace to the same entity.

Measurement and KPIs: making entity visibility measurable

Entity SEO and brand signals need their own metrics, because classic keyword rankings reflect brand authority only partially. Useful KPIs fall into four areas:

• Share of voice in AI answers: The share of AI answers on relevant topics in which your brand is cited or named, compared with your competitors.
• Brand mention tracking: Monitoring linked and unlinked brand mentions on the web, including sentiment and source quality, to capture trust signals and PR impact.
• Knowledge panel monitoring: Tracking whether a knowledge panel exists, which facts it shows, and whether they are correct. Errors in the panel are direct correction targets.
• Entity visibility: Checking whether the brand is represented cleanly and without contradictions in Wikidata, in the Knowledge Graph, and in relevant directories.

Keeping citations and mentions apart is essential. With its gap between a 74.9 percent citation rate and a 38.3 percent mention rate, the international ghost citations analysis shows that both values have to be tracked separately. A high citation rate without the name being mentioned delivers little brand impact, which is why the explicit mention of the brand name in AI answers is one of the most meaningful target metrics.

GDPR and reputation: legally compliant mentions in a B2B context

Brand signals also come from reviews, mentions, and public reputation, which in the DACH region come with clear legal guardrails. The GDPR applies whenever personal data is processed, for example in reviews that name individual employees or when you publish Person schema. In B2B, pure company data is considered less sensitive than data on natural persons, but as soon as specific contact persons are named, you have to observe the legal basis and purpose limitation.

For legally compliant mentions and reviews, a few principles apply: reviews should be genuine and verifiable, references you collect need the consent of the people named, and Person schema should only be published with the consent of the individuals concerned. In a B2B DACH context, reputation management therefore does not mean embellishing facts. It means maintaining your own entity actively, accurately, and on the basis of consent across all relevant sources. This is exactly the kind of diligence that Google and AI systems treat as a long-term trust signal.

Going further: entity SEO as an ongoing task

Entity SEO is not a one-off project but a continuous alignment. If you define a unique entity home, connect it to consistent sources through structured data and sameAs, build topical authority, and maintain brand mentions and reviews in a legally compliant way, you lay the foundation for visibility in classic search and in AI answers. The data points show two things: a good organic ranking alone does not secure a citation in AI answers, and even a citation does not mean that the brand is named. Consistent brand signals are the lever that makes both more likely.