How to get a Knowledge Panel for a person (without hacks): the system model

Most people think a Knowledge Panel is a reward. It's closer to a confidence threshold.

Google shows a person panel when it believes it can do two things consistently:

1. Resolve identity (this person is who the query means)
2. Attach facts (and not be embarrassed by contradictions)

This is why "optimizing" random pages rarely works. You're not fighting ranking factors. You're feeding an entity resolution system.

What a Knowledge Panel actually is

A Knowledge Panel is a UI layer built on Google's entity graph.

It can appear (or not appear) depending on:

• the query (exact name vs name + context)
• the country and language
• device (panels are often more visible on mobile)
• ambiguity (your name collides with other people/entities)

Your job is not "trigger the panel". Your job is to make the person entity stable enough that surfacing it is low-risk.

The decision model: why panels appear for some people

Think in gates. Panels appear when all gates are "good enough".

Gate 1: identity is stable (one canonical version)

The system wants one consistent identity:

• a single name spelling
• one canonical site
• one canonical handle (or a small set of handles)
• one consistent short bio

If you drift (different names, different sites, different bios), you are creating ambiguity. Ambiguity is risk.

Gate 2: disambiguation is easy (you are not "confusable")

If there are multiple people with the same name, Google needs disambiguation anchors:

• stable niche/occupation ("SEO & AI strategist" is better than "entrepreneur")
• stable geography (if relevant and consistent)
• stable associations (company, project, publication, channel)

Disambiguation is why one clear occupation line helps more than a long story.

Gate 3: corroboration exists (your site is not the only source)

Your site is the entity home. It is allowed to be "self-asserted". But the system still looks for corroboration:

• profiles on major platforms (LinkedIn, X, YouTube, etc.)
• consistent bios across those platforms
• pages that mention you outside your own domain (press, guest posts, podcast pages, event pages)

Not because you need "authority". Because the model needs agreement.

Gate 4: persistence is proven (signals don't vanish)

Panels don't love churn. If the system sees that your identity signals are stable over time (same site, same handles, same topical footprint), it gets easier to treat the entity as persistent.

This is also why constant rebranding slows everything down. Each change forces a new round of uncertainty.

What actually increases certainty (high impact)

1) Make one page the entity home (and commit to it)

Pick a single canonical URL that represents "you", and make it the most consistent place on the web.

For most personal sites, that can be the homepage or /about. For this site, the canonical entity home is more explicit:

• /person/mikhail-drozdov

The homepage and /about should support that page, not compete with it.

2) Use sameAs and rel="me" for identity, not for SEO decoration

Two mechanisms, two roles:

• sameAs in JSON-LD: declares identity equivalence across URLs
• rel="me" on profile links: asserts "this profile is me"

They don't "force a panel". They reduce identity ambiguity.

3) Build proof pages and close the loop

If you want the model to stop treating you as a random blog author, give it citation-like pages:

• /press: mentions, appearances, talks, guest posts
• /research: durable research or publication objects
• official profiles listed on the canonical person page

Link structure matters:

• homepage -> person / press / research
• person -> press / research
• press/research -> person

Closed loops are easy to interpret.

The important distinction: proof pages can be useful without all being search-facing pages. A social profile archive or video archive may stay noindex while still supporting discovery and attribution. The canonical person page should remain the source of truth.

4) Keep your bios boring (boring is machine-readable)

Your bio should be consistent and factual:

• who you are
• what you do (one line)
• what topics you cover (a small set)
• one alias if you use it

Save philosophy for essays. The identity block is infrastructure.

What doesn't work (or creates negative side-effects)

• "Add more schema types" (spammy schema doesn't create confidence)
• "Request changes everywhere weekly" (churn increases ambiguity)
• "Buy citations" (low-quality mentions are agreement noise)
• "Force the panel" thinking (wrong mental model)

A minimal plan you can execute in a weekend

1. Pick your canonical identity: name + site + handle.
2. Put a one-line identity block on the homepage (and keep it stable).
3. Make the canonical person page the stable identity source.
4. Add /press and /research pages and link them from the person page.
5. Ensure your author block on every post points to the person page and uses the same name.
6. Ensure your Person schema includes sameAs for the profiles you actually use.
7. Add rel="me" to your outbound profile links.
8. Make sure your social profiles link back to the site (and use the same bio line).
9. Publish 3–5 strong cluster posts in one topic (coherence > variety).
10. Stop changing identity signals for at least 6–8 weeks. No hacks - just reducing ambiguity and increasing agreement.

The clean version for Casinokrisa

For a full Knowledge Panel, the site should not look like a broad magazine. It should look like a stable expert/entity source.

The clean stack is:

• Person: /person/mikhail-drozdov
• Evidence: /press
• Research: /research
• Topic expertise: Indexing and visibility
• Entity model: Entity-based SEO

Everything else should either support those URLs or stay out of the index until it is strong enough to be a destination page.

Related reading (same model, different layer)

• Entity-first SEO for personal brands (checklist)
• Google indexing explained
• E-E-A-T meaning: what Google is actually trying to prevent