Entities vs. Keywords
Session 1.2 · ~5 min read
In 2008, if you wanted to rank for "best pump supplier Jakarta," you optimized a page for that exact phrase. You placed it in the title, the H1, the first paragraph, and the alt text of an image. You built backlinks with that anchor text. If you did it well enough, you ranked. The words on the page matched the words in the query. That was keyword SEO.
In 2026, that same search triggers a different process inside Google. Google identifies the entities in the query: [pump supplier] as a business category, [Jakarta] as a Place entity. It then looks for Organization entities that are categorized as pump suppliers and have a geographic association with Jakarta. It does not just match strings. It matches things.
The Shift from Strings to Things
In May 2012, Google launched the Knowledge Graph with the tagline "things, not strings." This was not a marketing slogan. It described a structural change in how Google processes information. Before 2012, Google was primarily a text-matching engine. After 2012, it became an entity-understanding engine.
The transition has been gradual but relentless. Each major algorithm update since then has increased Google's reliance on entity understanding: Hummingbird (2013) improved semantic query interpretation, RankBrain (2015) added machine learning to query processing, BERT (2019) improved natural language understanding, and MUM (2021) enabled multimodal entity connections.
Query words = page words"] --> B["2012: Knowledge Graph
Things, not strings"] B --> C["2015: RankBrain
ML query understanding"] C --> D["2019: BERT
NLP entity extraction"] D --> E["2021+: MUM / AI
Entity-first ranking"] style A fill:#2a2a28,stroke:#c47a5a,color:#ede9e3 style B fill:#2a2a28,stroke:#c8a882,color:#ede9e3 style C fill:#2a2a28,stroke:#c8a882,color:#ede9e3 style D fill:#2a2a28,stroke:#6b8f71,color:#ede9e3 style E fill:#2a2a28,stroke:#6b8f71,color:#ede9e3
Keywords are what users type. Entities are what Google understands. You can rank temporarily with keyword tricks. You can only rank durably by being a recognized entity.
How This Changes SEO Strategy
Keyword-based SEO and entity-based SEO produce different strategies for the same goal. Here is how they diverge:
| Dimension | Keyword SEO | Entity SEO |
|---|---|---|
| Primary target | Specific search phrases | Entity recognition and topical authority |
| Optimization focus | On-page keyword placement | Structured data, entity signals, corroboration |
| Content strategy | One page per keyword | Topic clusters around entity expertise |
| Link building | Backlinks with keyword anchor text | Entity mentions, citations, sameAs connections |
| Local visibility | Location keywords on pages | GBP, local citations, LocalBusiness schema |
| Durability | Vulnerable to algorithm updates | Compounding, algorithm-resistant |
| AI search impact | Minimal | Direct feed into AI answers |
This does not mean keywords are irrelevant. Keywords remain the interface between users and search engines. Users still type words. But Google processes those words by identifying the entities behind them. If your entity is recognized and associated with the right topics, you rank. If not, you need to outcompete on pure content merit, which gets harder every year.
Entity Association: The Hidden Ranking Factor
When Google recognizes your company as an entity and associates it with a topic (e.g., "industrial pumps"), every piece of content you publish on that topic benefits. This is topical authority, and it is entity-dependent.
A recognized entity publishing content within its known domain of expertise receives a ranking advantage that no amount of keyword optimization can replicate for an unknown entity. This is why established companies can publish thin content and still outrank detailed articles from unknown sources.
The chart above is conceptual, not sourced from specific data. But it represents a real trend that SEO practitioners have documented over the past decade. Entity signals have become progressively more important in Google's ranking systems.
Practical Implication
If you are currently running a keyword-focused SEO strategy, you do not need to abandon it. You need to add entity infrastructure underneath it. Keywords tell Google what your page is about. Entity infrastructure tells Google who you are and why your content should be trusted on that topic.
Further Reading
- Google Search Gets Smarter with Knowledge Graph - Coverage of Google's 2012 Knowledge Graph launch
- Google Patent Analysis: Entities in Queries - Serpact's analysis of how Google uses entities in query processing
- Things Not Strings: Semantic and Entity SEO - OnCrawl's deep dive into the entity-based search paradigm
Assignment
List 10 keywords you currently target or want to target. For each, identify the underlying entities Google associates with that keyword. Example: "best pump supplier Jakarta" contains the entities [pump supplier] (category), [Jakarta] (place). Then ask: is your company connected to any of these entities in Google's understanding? If not, keyword optimization alone will not get you there.