The Minimum Viable Entity: What Makes a Company Machine-Readable

In software development, there is a concept called the Minimum Viable Product. The smallest version of a product that can be shipped to real users and still deliver value. The idea is not to build everything. It is to build enough.

Entity infrastructure has the same dynamic. Most companies do not need a Wikidata entry, press coverage, ORCID profiles, and a 50-node verification network on day one. They need the minimum set of signals that makes them machine-readable. After that, they can iterate.

I call this the Minimum Viable Entity. It is the smallest configuration of digital signals that allows search engines and AI agents to recognize your company as a distinct, verifiable entity rather than just another website.

What "machine-readable" actually means

A company is machine-readable when a search engine or AI agent can answer three questions about it without relying on the company's own claims:

Question 1: Does this entity exist? Confirmed by at least two independent sources agreeing on the entity name, type, and basic attributes.

Question 2: What does this entity do? Confirmed by structured data (Organization schema) and corroborated by at least one external source.

Question 3: Can the claims be verified? Confirmed by cross-referencing the entity's own declarations (website, schema) with independent sources (Google Business Profile, directory listings).

If a machine can answer all three, you have an entity. If it can only answer one or two, you have a website. The difference matters because AI agents cite entities. They link to websites. Citation is trust. Links are traffic. You want both, but citation is what drives enterprise credibility.

The four components of a Minimum Viable Entity

After building entity infrastructure for three companies and studying how AI agents verify entities across dozens of industries, I have narrowed the MVE to four components. Remove any one of them and the system breaks. Add them all and you have a machine-readable entity.

Component 1: Organization schema on your domain

This is the foundation. A JSON-LD block on your homepage that tells machines who you are in a structured format they can parse.

At minimum, your Organization schema needs:

• @type: Organization (or a more specific subtype like Corporation)
• name: your exact registered company name
• url: your canonical domain
• description: what your company does (one to two sentences)
• address: your registered business address
• sameAs: array of URLs to your profiles on other platforms
• logo: URL to your company logo

This is not optional. Without Organization schema, your website is a collection of HTML pages. With it, your website is a declared entity. Every AI agent and search engine reads JSON-LD. It is the most direct way to communicate your identity to machines.

The sameAs property deserves special attention. It is the glue that connects your domain to your external verification nodes. Without sameAs, your schema declares an entity that exists in isolation. With it, your schema declares an entity that exists across multiple platforms.

Component 2: Google Business Profile (verified)

Google Business Profile is the single most impactful external verification signal for any company. Here is why:

Google's own AI systems (Gemini, AI Overview, Knowledge Graph) draw directly from GBP data. When Google has a verified GBP for your company, it has confirmed that a real business exists at a real address with a real phone number. This is verification that Google performed, not verification you claimed.

The verification process itself is important. Google sends a postcard or makes a phone call to confirm your business exists at the stated location. This is a level of verification that most other platforms do not perform. When an AI agent sees a verified GBP, it weighs that signal more heavily than an unverified directory listing.

Your GBP must match your Organization schema exactly. Same name. Same address. Same phone number. Same website URL. Any discrepancy introduces entity ambiguity, which is the opposite of what you want.

Component 3: One authoritative directory listing

Two sources (your website and GBP) are good. Three is where corroboration begins to feel definitive to an AI agent.

The third source should be an authoritative directory in your industry. For manufacturing companies, that might be Thomasnet or Kompass. For tech companies, Crunchbase. For Indonesian B2B, IndoTrading as a starting point with international directories as the next step.

The key word is "authoritative." A listing in a no-name directory that nobody, human or machine, has heard of adds no verification value. You need a directory that AI systems recognize as a legitimate source of business entity data.

One is enough for the MVE. You can add more later. But at least one external, authoritative, third-party confirmation of your entity is necessary to cross the verification threshold.

Component 4: Closed verification loop

This is the component that ties the other three together. A closed verification loop means every node in your entity network points back to every other node.

Your website's sameAs points to GBP and the directory listing. Your GBP links back to your website. Your directory listing links back to your website. Each node confirms the others. There are no dead ends. No one-way links. No orphaned nodes.

Without the closed loop, you have three independent data points that might or might not be about the same entity. With the closed loop, you have three mutually confirming data points that AI agents can traverse and verify as a single entity.

What the MVE is not

The Minimum Viable Entity is not the end state. It is the starting line. After establishing your MVE, there are many additional layers you can add:

• Wikidata entry (requires notability criteria)
• Press coverage and media mentions
• Academic citations or industry publications
• Additional directory listings
• ORCID profiles for key personnel
• Institutional partnerships and references
• Content publication velocity

Each of these adds strength to your entity infrastructure. The Trust Chain Methodology provides a framework for layering these additional signals systematically. But none of them matter if the MVE is not in place first.

Think of it like building a house. The MVE is the foundation, walls, and roof. Everything else is insulation, plumbing, electrical, and finishing. You need the structure before you can add the systems. And you need the systems before anyone wants to live there.

Common MVE failures

After auditing dozens of company websites across Indonesian B2B, here are the most common ways companies fail at the MVE level:

No Organization schema at all. This is the most common failure. The company has a website, maybe even a decent one, but zero structured data. To machines, it is just HTML. No entity signal whatsoever.

Schema with no sameAs. The company added Organization schema (or an SEO plugin generated it) but left the sameAs array empty. The schema declares an entity but provides no way for machines to verify it against external sources.

GBP exists but is unverified. An employee created a GBP listing years ago but never completed the verification process. Unverified GBP listings carry significantly less weight than verified ones.

Name inconsistency. The website says "Witanabe Engineering," the GBP says "PT Witanabe Integrasi Indonesia," and the directory says "Witanabe." Three names that might be three different companies, as far as an AI agent knows.

No external corroboration. Website and GBP exist, but no third-party directory listing. Two sources can establish plausibility. Three establish credibility. The third source is what pushes you from "probably real" to "confirmed real" in entity verification.

Implementation timeline

A company with an existing website can implement the full MVE in two to three weeks:

Days 1 to 3: Audit existing data. Establish canonical company name, address, description. Check for inconsistencies across any existing profiles.

Days 4 to 7: Add Organization schema to your homepage. Include name, URL, description, address, logo, and empty sameAs array (you will fill this in as you create the other nodes).

Days 8 to 10: Create or claim Google Business Profile. Begin verification process. Use exact same data as your Organization schema.

Days 11 to 14: Create one authoritative directory listing. Match all data exactly.

Days 15 to 17: Close the loop. Update your Organization schema's sameAs to include GBP and directory URLs. Ensure all platforms link back to your website. Verify every data point matches.

Days 18 to 21: Test. Ask ChatGPT or Perplexity about your company. Check Google's entity understanding using the Knowledge Graph Search API. Identify and fix any remaining inconsistencies.

Three weeks. No budget required beyond the time investment. The ROI is asymmetric: a few hours of structured work creates entity signals that compound over months and years.

The bottom line

You do not need to build a 50-node entity verification network to become machine-readable. You need four components: Organization schema, verified Google Business Profile, one authoritative directory listing, and a closed verification loop connecting all three.

This is the Minimum Viable Entity. It is the threshold between being a website and being an entity. Between being indexed and being verified. Between being found and being cited.

Build it first. Build it right. Then iterate.