Course → Module 1: Identity Consistency: NAP and Beyond
Session 2 of 7

Your brand name seems obvious to you. You know that "Hibrkraft," "HIBRKRAFT," "Hibr Kraft," "HibrKraft," and "Hibrkraft Indonesia" all refer to the same company. A human reading any of those would understand immediately.

An algorithm does not have that luxury. To a matching algorithm, each of those strings is a distinct candidate entity. It has to decide, based on context and corroborating signals, whether they all point to the same thing. Every variation makes that decision harder and less confident.

To a human, these are obviously the same brand. To an algorithm, these are five different potential entities.

How Name Variations Fragment Your Entity

Google's entity recognition works by collecting mentions across the web and clustering them. When it finds "Acme Corp" on your website, "Acme Corp" in a directory listing, and "Acme Corp" on your LinkedIn page, it clusters those as belonging to the same entity. Confidence goes up with each matching mention.

But when it finds "Acme Corp" on your website, "ACME Corporation" in a directory, "Acme Corp." on LinkedIn, and "Acme" on Yelp, the clustering algorithm has to make judgment calls. Some of these might cluster together. Others might not. The result is a fragmented entity with lower confidence scores across the board.

The left side shows four platforms using the exact same name. Google confidently clusters these into one entity. The right side shows four platforms with variations. Google might split these into two candidate entities, each with low confidence. Neither gets a Knowledge Panel.

Common Name Variation Types

Variation Type Example: Canonical Example: Variation Risk Level How Often It Happens
Capitalization Hibrkraft HIBRKRAFT Medium Very common (social media defaults)
Spacing Hibrkraft Hibr Kraft High Common (people "correct" compound names)
Legal suffix Arsindo Teknik PT Arsindo Teknik Prakarsa High Common (directories pull from registrations)
Abbreviation International Business Machines IBM Depends on recognition Common for established brands
Punctuation Acme Corp Acme Corp. Low-Medium Very common (trailing periods)
Location qualifier Acme Corp Acme Corp Jakarta Medium Common in directory auto-generation
Language Arsindo Teknik Arsindo Engineering High Common in bilingual markets
Character encoding Muller Müller High Common for names with diacritics

The Decision: Which Name Is Canonical?

You need to pick one version and commit to it everywhere. The decision criteria are:

  1. What does your website use? Your website is your entity home. Its version should be canonical.
  2. What does your Google Business Profile use? GBP is the second most important source for Google.
  3. What is most recognizable to your audience? If you are known by an abbreviation, consider whether the full name or abbreviation should be primary.
  4. What is searchable? If people search for you using a specific version, that version should be your canonical name.

Once you decide, write the canonical name in your master NAP record. This is the only version that should appear anywhere online.

Handling Legitimate Variations

Some businesses genuinely use different names in different contexts. A company might have a legal name ("PT Arsindo Teknik Prakarsa"), a trading name ("Arsindo Teknik"), and a brand name ("Arsindo"). All three are legitimate. The question is which one you want Google to recognize as your entity name.

The answer is usually the trading name or brand name, because that is what customers search for. The legal name appears in registrations and contracts. The brand name appears in search queries. Optimize for the brand name, and use structured data to connect the legal name as an alternate.

graph LR A["Legal name:
PT Arsindo Teknik Prakarsa"] -.->|"alternateName in schema"| B["Canonical name:
Arsindo Teknik"] C["Short name:
Arsindo"] -.->|"alternateName in schema"| B B -->|"Used on website"| D[Entity Home] B -->|"Used on GBP"| E[Google Business Profile] B -->|"Used on all profiles"| F[Social + Directories] style B fill:#2a2a28,stroke:#c8a882,color:#ede9e3 style A fill:#2a2a28,stroke:#8a8478,color:#ede9e3 style C fill:#2a2a28,stroke:#8a8478,color:#ede9e3 style D fill:#2a2a28,stroke:#6b8f71,color:#ede9e3 style E fill:#2a2a28,stroke:#6b8f71,color:#ede9e3 style F fill:#2a2a28,stroke:#6b8f71,color:#ede9e3

The canonical name (solid border) is used everywhere. Legal and abbreviated names (dashed lines) are connected via the alternateName property in structured data. This tells Google: "These are all the same entity, but this specific version is the primary name."

Personal Names

The same principles apply to personal entities. If you are building a personal brand, decide on one version of your name. "Ibrahim Anwar," "Ibrahim A.," "Hibranwar," and "Ibrahim Anwar Hibran" are all valid. But only one should be canonical across platforms.

For personal names, consistency in given name, family name, and any professional titles or suffixes is critical. If your LinkedIn says "Ibrahim Anwar, Director" and your Twitter says "Hibranwar" and your personal website says "Ibrahim A. Hibran," Google sees three potential person entities.

Further Reading

Assignment

  1. Search Google for your brand name (exact match). Then search for common variations (with and without legal suffix, abbreviated, different capitalization). Note how many different versions appear in results.
  2. Check five of your most important online profiles (website, GBP, LinkedIn, Facebook, one directory). Record the exact name string used on each. Are they identical?
  3. If you have legitimate alternate names (legal name, abbreviation, translated name), list them. Decide which is canonical and which are alternates.
  4. Update your master NAP record with the canonical name and a list of alternate names to connect via structured data later (Module 2).