What is an ontology, and how do you use it for SEO?

What an ontology is

Think of a family tree crossed with a dictionary. A family tree shows who belongs to whom. A dictionary explains what each thing is. An ontology does both for a whole topic, and it is built from just three ingredients:

Written as text in the format the visualizer reads, the same picture looks like this:

[[bald eagle]] is a type of [[bird]] [isA]
[[bird]] is a class of [[vertebrate]] [isA]
[[bird]] has [[feathers]] [hasAttribute]
[[bird]] comes from [[theropod dinosaur]] [derivedFrom]
[[bald eagle]] lives in [[North America]] [locatedIn]

Double brackets mark the things. The word in single brackets at the end names the kind of fact.

Ontology vs keyword list vs topical map

People mix these up all the time. They are three different things, used one after another:

A keyword list on its own misses the connections, and an ontology on its own ignores what people search for. Combine them and you get a topical map that is both complete and in demand.

Each kind of fact creates a kind of page

This is the practical heart of it. Every relation type in your ontology turns into something concrete on your website:

Step by step: from ontology to a ranking zoology site

Follow these in order. The label on each step tells you whether you can skip it.

1. 1

   Decide who the site is for and how it makes money

   Necessary

   This is your source context. The same topic gives a completely different plan for a student study site, a pet shop or a safari tour company. Decide this first, because it filters everything after it.

   Example: a zoology study site for high-school and university students, earning from ads and paid revision notes.

2. 2

   Pick the central entity

   Necessary

   This is the one thing every page on the site is about in some way. Make it broad enough to cover what competitors cover, and narrow enough that your site can genuinely help with it.

   Example: “animals” (the kingdom Animalia). Every page, from “What is a mammal?” to “African elephant diet”, connects back to animals.

3. 3

   Generate the ontology and visualise it

   Necessary

   Use the AI prompt in the Ontology Visualizer (or paste your own data, such as a species list as CSV). Read the graph. Do the groups make sense? Is anything obviously missing?

   Example: 80 facts like “bald eagle is a type of bird”, “bird has feathers” and “habitat loss causes population decline”.

   Use: Ontology Visualizer
4. 4

   Fix the quality check

   Necessary

   A good ontology is a web, not a star. Connect any disconnected islands, make sure the important entities have at least two facts, and use different kinds of relation (types, features, causes, places).

   Example: if “conservation” floats alone, connect it: “poaching causes population decline in the african elephant”.

   Use: Quality check tab
5. 5

   Turn the family tree into your site structure

   Necessary

   The “is a type of” and “is part of” facts form a hierarchy. That hierarchy becomes your categories, URLs, breadcrumbs and navigation.

   Example: /animals/vertebrates/mammals/african-elephant/ with breadcrumbs Animals › Vertebrates › Mammals › African elephant.

   Use: Family tree tab
6. 6

   Build one page template per entity type

   Necessary

   Entities of the same type share the same features (root attributes). Every species has a habitat, diet, lifespan and conservation status, so every species page answers those questions in the same order. Features only some species have (rare attributes) become extra sections where they apply.

   Root: classification, appearance, habitat, diet, lifespan, behaviour, reproduction, conservation status. Rare: echolocation, migration, hibernation, venom, metamorphosis.

7. 7

   Check every planned page against real searches

   Necessary

   An ontology describes what is true, and search data tells you what people ask. Export your keywords from Google Search Console, Ahrefs or Semrush and check each planned page. Keep pages that are searched or needed to define the topic. Merge or drop the rest.

   Example: “african elephant lifespan” and “how much does an elephant eat” become sections, and “elephant habitat map” may become its own page.

   Use: Content Gap Analyzer
8. 8

   Write each page as questions and answers

   Necessary

   Start with a direct definition, then answer the root-attribute questions in order: numbers with units, lists for types, tables for comparisons. Keep facts identical across pages, so the same lifespan is quoted everywhere.

   Example: “The African elephant is the largest land mammal… It lives 60–70 years in the wild.” Then diet, habitat, behaviour…

   Use: Content brief (Gap Analyzer)
9. 9

   Link pages the way the ontology links entities

   Beneficial

   If two entities have a fact between them, their pages should link, using the entity name in the anchor text. Hub pages link down to their species. Species link sideways to their habitat, diet type and adaptations.

   Example: the bottlenose dolphin page links to “echolocation”, which links to “bat” and back up to “adaptation”.

10. 10

    Write the bridge content

    Beneficial

    The visualizer's “Missing links” are groups that your ontology barely connects. Articles that explain those connections are often the most original content on the site.

    Example: “How climate change affects animal migration” bridges the conservation and behaviour groups.

    Use: Missing links tab
11. 11

    Add structured data with entity IDs

    Beneficial

    Mark up each page with Article/WebPage schema, an “about” entity linked to its Wikidata ID (sameAs), and BreadcrumbList from the family tree. It removes ambiguity about which “jaguar” or “python” you mean.

    Example: the African elephant page uses “about” → African elephant (Wikidata), plus breadcrumbs.

    Use: Entity Graph Mapper (schema)
12. 12

    Publish one group completely before starting the next

    Beneficial

    Finish one hub (e.g. all mammal pages) before spreading thin across every group. It is easier to interlink, easier to measure, and practitioners widely recommend it. It is not a proven ranking rule.

    Example: launch the Mammals hub with its 30 species pages, then Birds, then Reptiles.

13. 13

    Keep a formal ontology file (OWL / RDF)

    Optional

    Search engines don't read your ontology file. A simple list of facts is enough for planning. Formal OWL or RDF only matters if you are building software, a database or a knowledge graph product.

    Example: a zoology site doesn't need OWL. A spreadsheet of species and their facts is enough.

Checklist: necessary, beneficial or optional?

What the zoology website plan looks like

Run the zoology example through the visualizer and its groups translate into a site like this:

Core = what your audience mainly comes for (here: students studying animals). Outer = supporting topics that add depth and links. Build the core first.

The species page template (from root attributes)

H1  African Elephant: Facts, Habitat, Diet & Lifespan
    Direct definition (1–2 sentences: what it is, its class, one identifying fact)
H2  Classification            ← is a type of / is part of
H2  Appearance and size       ← has the feature (numbers with units)
H2  Habitat and range         ← lives / is found in
H2  Diet                      ← herbivore / carnivore / omnivore
H2  Behaviour                 ← rare attributes go here only if they apply
H2  Reproduction and lifespan ← numbers with ranges
H2  Conservation status       ← comes from the IUCN Red List; causes of decline
H2  FAQs                      ← real questions from your keyword data

Common beginner mistakes

• A star instead of a web

  Every fact starts with “animal”. Connect the other things to each other, too: eagle → bird → feathers.

• Treating the ontology as the page list

  Not every entity deserves a page. Check search demand. Many entities become sections, not pages.

• Stuffing entities into text

  Mentioning 50 animal names doesn't cover a topic. Defining things and explaining how they connect does.

• Facts that disagree across pages

  If one page says elephants live 60 years and another says 70, fix it. Consistency is part of being a trustworthy source.

• Spreading across every group at once

  Ten half-finished hubs are weaker than two complete ones.

Quick answers