AI & Infrastructure
Named Entities

Named Entities

Named entities are real-world things referred to by a proper name, such as a person, organization, place, product or event. “OpenAI” and “Paris” are named entities, while “company” and “city” are not.
June 27, 2026
Cihan Geyik
Table of Content

Ask an AI assistant for the best expense management tool for a 200-person company in Germany, then count the proper names in the answer. You’ll find vendors, maybe a review site, and Germany itself. Those are named entities. For a brand they are the part of the answer that counts, because being one of those names is what visibility means.

A named entity has a name of its own. “Ansvisor” and “Germany” qualify. “Software” and “country” are concepts, however much meaning they carry.

What counts as a named entity?

A named entity is one specific thing that a proper name picks out. “The company” could be any company. “Notion Labs” is one.

Practical systems stretch the idea. Most also tag dates, prices, percentages and quantities, because those values matter in text even though they aren’t names. Linguists sometimes call these numeric or value entities.

Named entity types with examples

TypeExamplesWhere it matters in AI answers
PersonFounders, executives, authorsExpert quotes and founder facts
OrganizationCompanies, agencies, universitiesVendor lists and comparisons
LocationCountries, cities, regionsLocal and market-specific answers
ProductSoftware, devices, AI modelsProduct recommendations
EventConferences, launches, recallsNews and timeline answers
WorkBooks, reports, studiesResearch cited as a source
ValuePrices, dates, percentagesPricing and stats repeated in answers

Standard entity type schemes compared

There is no single official list. Each tool draws its own lines, and the differences show up when you test your brand.

  • CoNLL-2003, the classic research benchmark, has four types. Anything that isn’t a person, organization or location goes into miscellaneous.
  • OntoNotes 5 uses eighteen, and spaCy’s English models follow it. Beyond the basics it has NORP for nationalities and groups, FAC for buildings, GPE for countries and cities, LAW and WORK_OF_ART.
  • Google’s Cloud Natural Language API lists thirteen type values, including CONSUMER_GOOD, PRICE, ADDRESS and PHONE_NUMBER.
  • schema.org, the structured data vocabulary, has a type hierarchy with Organization, Person, Place, Product, Event and CreativeWork near the top.

The same name can land in different buckets. Google’s list has no type just for software, so one tool may call a SaaS product an organization while another files it as a consumer good or other.

Named entities vs keywords vs concepts

Take the query “best CRM for agencies in Germany.” The whole string is a keyword. “CRM” and “agencies” are concepts. “Germany” is a named entity, and it changes the answer more than any other word, because it narrows the market. Matching the concepts is the job of semantic search.

A brand name such as HubSpot is also a named entity. Search engines resolve it to one record in a knowledge graph, where its facts and relationships live. Your company’s record is your brand entity.

How AI systems use named entities

Prompt names a category and a market → Retrieval pulls pages that name vendors → The model chooses which names to include → The answer lists them, with or without links

Two processing steps handle names on both sides of that flow. Entity recognition finds the names and their types. Entity linking works out which specific thing each name means. Named entities in the prompt shape retrieval, and named entities in the sources shape the answer.

What named entities mean for AI visibility

  • A named brand is a mention. A named brand with a link to its site is a citation. Track the two separately.
  • Your products are separate entities. Track product names as well as the company name.
  • People count. Founders and in-house experts get quoted by name in answers about your field.
  • Locations change the list. The same prompt for Germany and for the US can name different vendors.
  • Competitors are named entities in your answers too, and their presence is what you benchmark against.
Commercial answers end in a short list of names. Whether your brand is on that list, and in which position, is the core visibility question.

Our guide on tracking competitors in ChatGPT and AI search shows how teams watch those lists over time.

Common misconceptions

  • “Only companies and people count.” Products, events, laws and published reports are named entities too.
  • “Every tool uses the same types.” Schemes range from four types to eighteen or more.
  • “Any capitalized word is an entity.” Sentence starts, headings and title case fool simple rules.
  • “Being recognized means being trusted.” Recognition only shows the name was found and typed.

How Ansvisor helps

Ansvisor’s Competitor Tracking and Benchmarking shows which brands each AI platform names on every tracked prompt, next to yours. It compares mentions, citations and share of voice for each rival. The Gaps and Strengths view lists the prompts where rivals appear and your brand doesn’t. Add product names and market variants to your prompts so the picture holds across countries and languages.

Named Entity, Named Entity Types, Entity Categories, NE Types

Let AI Search Work for Your Growth

Each prompt you win brings more mentions, more citations and more buyers. Ansvisor keeps finding the next one.

AI Answers Mentions Citations Visibility Score AI Share of Voice AI Traffic Revenue
Win Back Search VisibilityTry Ansvisor 14-day Free →✓ Zero risk. Cancel anytime.

FAQ

Frequently asked questions.

What are Named Entities?

Named entities are specific things with a proper name, such as people, companies, places, products and events. “Microsoft”, “Berlin” and “ChatGPT” are named entities. Generic words like “software” or “city” are concepts, even though they carry meaning.

What are the main types of named entities?

Most schemes include person, organization, location and product, and many add events, works of art, laws and languages. Practical systems also tag dates, money and percentages. The CoNLL-2003 benchmark used four types, spaCy’s English models use eighteen and Google’s Natural Language API uses thirteen.

How do AI systems identify Named Entities?

Entity recognition models find names in text and assign each a type. Entity linking then connects each name to a knowledge base entry, so the system knows which specific person or company is meant. Both steps run on prompts and on the sources an AI answer draws from.

How do Named Entities affect AI visibility?

AI answers to commercial questions usually end in a short list of named brands. Being one of those names is a mention, and being linked as a source is a citation. Your products, founders and locations are named entities too, and each can appear or be missing.

What is the difference between a named entity and a keyword?

A keyword is any word or phrase people search, such as “best CRM for agencies”. A named entity is a specific thing with its own identity, such as HubSpot. A keyword can contain named entities, and the entities usually carry the most precise meaning in it.

Help us grow the AI Visibility Glossary

New terms are added regularly.

Help us improve the page or suggest a new term →
About the Author
Cihan Geyik

Cihan Geyik

Co-founder at Ansvisor

Cihan Geyik is the co-founder of Ansvisor, an open-source AI Visibility platform for AI Search. With more than 15 years of experience in digital marketing and growth, he writes about AI visibility, AI search, AEO, GEO, citations, and answer engines. He focuses on helping brands understand and improve their presence across ChatGPT, Gemini, Google AI Overviews, Bing AI Mode and other AI answer engines.

Google Preferred Sources
Get more Ansvisor insights in Google

Add Ansvisor as a preferred source and Google will show you more of our AI search, AEO and GEO guides.