Credible Roots / Field Notes / Wikidata
Reference
Wikidata: what it is, whether you qualify, and how to do it properly
A far lower bar than Wikipedia, read directly by Google's knowledge graph, and misunderstood in both directions. What it actually is, the three inclusion criteria, and the steps that keep an item alive.
More than five years in personal branding. Runs atharvsankpal.com and works with founders and businesses in the US on their digital presence. Previously produced AI video for brands.
Published 2026-08-14
Researched against primary sources, drafted with AI assistance, then reviewed and approved by the named author. How we write · About the author
The short answer
Wikidata is a free, open database of structured facts about entities, run by the Wikimedia Foundation alongside Wikipedia. Where Wikipedia stores prose for humans, Wikidata stores machine-readable statements: this person holds this position, at this institution, from this date, sourced to this reference. It accepts an item if any one of three criteria is met, and the second — a clearly identifiable entity describable from serious, publicly available references — is reachable for most professionals with a documented career. It is not a profile you own. Anyone can edit it, every statement needs a reference, and that trade is the entire point.
- Structured facts, not prose. Statements a machine can parse.
- Read directly by Google's knowledge graph, which is why it affects knowledge panels.
- Serious is not the same as significant. The bar is verifiability, not fame.
- Most deletions are not about notability — they are about references and tone.
On this page
What it actually is
Every entity in Wikidata — a person, a company, a book, a chemical, a city — gets an item with an identifier, and everything known about that entity is expressed as statements attached to that item.
The difference from an encyclopedia is worth sitting with. An encyclopedia article says a person was appointed to a chair in a particular year, in a sentence, in context. Wikidata says the same thing as a property and a value with a date qualifier and a reference, in a form a machine can read without interpreting language. That is less satisfying to read and enormously more useful to a system trying to answer a question.
It is also public infrastructure rather than a directory listing. There is no account that owns your item, no dashboard, and no support line. What there is instead is a permanent, referenced record that other systems can consume.
Why search engines and AI read it
Google's knowledge graph reads Wikidata directly. That is the practical reason people in this field care about it: the knowledge graph is what produces knowledge panels, and it is fed by structured sources it can parse rather than prose it has to interpret.
The same logic applies to answer engines. A system grounding an answer in verifiable facts prefers a source where the fact is stated as a fact with a reference attached, rather than inferred from a paragraph. A well-referenced item makes you legible to those systems in a way that a beautifully written bio page does not.
Keep the claim honest
An item is an input, not a lever. It does not produce a knowledge panel by itself, and anyone telling you it does is selling something. See how panels actually work.
The three inclusion criteria
Wikidata accepts an item if at least one of three things is true. You do not need all three.
| Criterion | What it means | Who it usually covers |
|---|---|---|
| A Wikimedia sitelink | The subject already has a page on Wikipedia or a sister project, in any language | Anyone with an existing article anywhere |
| Identifiable and describable | A clearly identifiable entity that serious, publicly available references can describe | Most professionals with a documented record |
| Structural need | The item is needed to make statements on other items useful | Co-authors, employers, connecting entities |
Note what is absent. There is no requirement for significant coverage, no requirement for independence in the Wikipedia sense, and no assessment of whether you are important enough to write an encyclopedia article about.
What counts as a serious reference
This is where criterion two does its work, and where people get it wrong in both directions. Serious and publicly available is a lower bar than Wikipedia's significant, reliable and independent, but it is still a bar. It means a source a reasonable person would treat as dependable and that anyone can go and check.
Institutional pages, professional registries, authority files, indexed publications, official databases and established publications all qualify comfortably. What does not qualify is material that is self-authored or promotional. Your own website describing how experienced you are is not a serious reference for that claim. Your own site confirming that you hold a stated role at a stated organization is a much weaker case than the organization's own record of the same fact.
How this differs from Wikipedia
| Wikipedia | Wikidata | |
|---|---|---|
| Asks | Has the world written about you in depth? | Can you be identified and described from public sources? |
| Needs independence | Yes, strictly | Less strictly, but promotional material still fails |
| Needs significance | Yes | No |
| Format | Prose for people | Statements for machines |
| Typical blocker | Coverage is all self-generated | Statements are unreferenced, or the tone is promotional |
| Reachable for | A minority of professionals | Most professionals with a documented career |
The relationship runs one way. A Wikipedia article in any language automatically satisfies Wikidata's first criterion. A Wikidata item does nothing to help you qualify for a Wikipedia article — it is not a stepping stone, a trial run, or evidence of anything a Wikipedia reviewer weighs. The two get conflated constantly, including by people selling both. What Wikipedia actually requires.
Doing it yourself, step by step
Mechanically simple, and easy to do badly. Two things to settle before you start: whether you meet one of the three criteria, and what position you are in. Editing about yourself is a conflict of interest and is treated as one. If you or anyone you are paying is doing the editing, the Wikimedia Terms of Use require disclosure of employer, client and affiliation.
- Search first. Search your exact name, then variants. Duplicates get merged and statements get lost in the merge, so this step saves more trouble than any other.
- Create the item with a neutral label — your name as it is normally written, nothing else.
- Write a short disambiguating description. This is the trap. The description exists to tell two people with the same name apart, not to position you. “Indian software entrepreneur” is right. “Visionary founder transforming logistics” gets stripped, and marks the item as promotional while it lasts.
- Add core statements using properties, not free text: instance of, occupation, employer, position held, educated at, country of citizenship, with date qualifiers where they apply.
- Reference every statement. An unreferenced statement is the thing that gets removed. A reference to the organization's own record of a role is stronger than a reference to your site.
- Add external identifiers — ORCID, VIAF, professional registries, official profiles. These are what let other databases resolve you as the same entity.
- Stop before you overbuild. A short, fully referenced item is stronger than a long one with gaps.
If you would rather not
This is the layer we usually build first for clients, because it is the cheapest structured signal available and it is reachable long before Wikipedia is. It is also entirely possible to do yourself in an afternoon, and this page is the whole method. Ask us if you want it done, not because you cannot.
Why items actually get removed
- Unreferenced statements. The single most common cause, and the easiest to avoid.
- A promotional description or label. Marketing language gets stripped, and it colors how the whole item is read.
- Duplication. An item created without searching, then merged, usually losing detail.
- Genuine failure of the criteria. Real, and rarer than the other three. If no serious public source describes you, nothing you do to the item will save it.
Questions people actually ask
Do I need a Wikipedia article to be on Wikidata?
No. An article in any language satisfies the first criterion automatically, but it is only one of three. Most professionals qualify under the second.
Does a Wikidata item get me a Google knowledge panel?
Not on its own. The knowledge graph reads Wikidata directly, so a well-referenced item is a real input, but a panel appears when Google is confident about you across many corroborating sources.
Can I edit my own item?
You can, and it is a conflict of interest. If money is involved, disclosure of employer, client and affiliation is required under the Terms of Use.
Why did my item get deleted?
Usually unreferenced statements, a promotional description, or a duplicate — not notability.
Sources
- Wikidata. Notability policy, the three inclusion criteria.
- Wikidata. Help pages on items, statements and sources.
- Wikimedia Foundation. Terms of Use, section 4, paid contributions disclosure.
- Google. About knowledge panels — on how entity information is sourced.
How this was made: researched against the primary guidelines rather than summaries of them, drafted with AI assistance, then reviewed and approved by the named author before publication. This page replaces five shorter pages that covered the same ground separately; the merge is logged on our corrections page. Our editorial standards.
Related reading
- Wikipedia notability — the much higher bar, and the routes by profession.
- Google knowledge panels — what the structured data actually feeds.
- How to get cited by AI search
- How this actually ends up at Wikipedia
The whole method is on this page
If you want it built rather than explained, that is what we do. If you would rather do it yourself in an afternoon, nothing here is missing.
Book a 30 minute call