Google shows a panel when independent public evidence agrees so thoroughly that hesitating costs it more than displaying. One home, one story, many consistent witnesses. The part that takes months is the witnesses, because they are other people, and nothing you publish about yourself substitutes for them.
The one sentence the whole thing runs on
One home, one story, many consistent witnesses.
That is the method. It is not proprietary, nobody is hiding it, and the people who wrote the book on this say so themselves. What is hard is not knowing it. What is hard is that the third clause takes other people and therefore takes months.
Google shows a panel when independent public evidence agrees so thoroughly that hesitating costs it more than displaying. That is the whole decision. Everything below is a description of how to make evidence agree.
The four parts of an identity Google can read
The clearest way to hold this in your head is anatomical.
| Part | What it is | What breaks when it is missing |
|---|---|---|
| Skeleton | The entities themselves: you as a person, your company as an organisation, treated as separate connected things | Google merges you with your company, or splits you across two records that each stay weak |
| Organs | Every public surface where you can be observed: profiles, listings, directories, feeds | Nothing to cross-check, so nothing gets confirmed |
| Muscle | The actual work: case studies, talks, interviews, published pieces | The identity is asserted and never demonstrated |
| Blood flow | Signals that humans care: reviews, watch time, branded searches, repeat mentions | Everything looks correct and inert, which reads as manufactured |
Get all four agreeing and Google stops guessing. Most people arrive with strong muscle and no skeleton, which is why they feel invisible despite having done real work.
The five proof types, in the order they matter
1. The entity home
One canonical site that is unambiguously about you. Ideally on a domain matching your name. Separate from the company site and linked both ways. The first paragraph is your identity line in plain English with no adjectives — the format is literally Name is the Role at Organisation, based in City. Master headshot above the fold. Sections for media, speaking, press, contact.
If you take one thing from this guide: replace your link-in-bio service with a page on your own domain. A link aggregator on somebody else’s domain is a hub of your identity that you do not own and Google does not attribute to you.
2. Owned profiles, aligned
LinkedIn, YouTube, Instagram, X, podcast feeds. Their job here is not reach. It is agreement. Identical name spacing and capitalisation, the same headshot file, the same first eight to ten words of the headline, the same city, the same opening bio line, one link pointing home.
This is the cheapest item on the list and the most commonly skipped, and misalignment is the most frequent silent regression later — somebody updates one profile, and six months of consistency quietly develops a crack.
3. Third-party corroboration
Articles, podcasts, conference bios, association listings, vendor pages where credible other people name you and link to you. This is the part that takes months and the part that cannot be bought safely.
4. Rich media, and especially books
Talks and long-form podcasts published to your own site first, then mirrored outward with the first link pointing back. Say your exact canonical name aloud in the first fifteen seconds of every video and put that same string in the captions.
Books outrank all of it. An ISBN plus author and book pages on your entity home is the most durable accelerator available, which is why two of the eight names on our record page carry the descriptor Author.
This is proof type four working in public. Six ISBNs, six covers, six years, rendered by Google directly inside the panel. No other category of evidence gets its own shelf.
It is also the reason the descriptor reads author rather than any of the other things Chetan Bhagat does. Google picks the descriptor from the most consistently attested role, and published books are about as consistently attested as evidence gets.
5. Local signals
Google Business Profile, chamber and association listings, local press tying you to a place. Remote-first people routinely fail this one and cannot work out why their record stays soft. Pick the city you would name from a stage and use it everywhere.
The two tall bars are the two that take months. Everything short is an afternoon, and almost every stalled build we inherit has done the short bars and none of the tall ones.
The cross that spammers cannot fake
The most useful framing in this field: think of your authority as having two axes.
The vertical axis is your trade. Peers, suppliers, associations, conferences, industry podcasts. The people who would recognise your name at a trade event.
The horizontal axis is your place. Local press, chamber, allied trades, charities, civic organisations. The people who would recognise your name at a city function.
Independent witnesses at the intersection of those two axes are the pattern that cannot be manufactured, because building it requires actually being in both worlds. Anybody can buy a hundred generic links. Nobody can buy being the person the regional trade association and the local business journal both name in the same quarter.
Three to five nodes on each axis, every one linking back to the entity home and repeating the identity line as close to verbatim as you can get them to.
The horizontal axis assumes a dense layer of crawlable local institutions — chambers of commerce, city journals, civic directories — that in a lot of countries either does not exist online or is not indexed. In India in particular, the substitutes are regional-language press, industry bodies, university and alumni pages, event and speaker listings, and reputable trade publications. Same axis, different addresses. Anybody applying an American citation list to an Indian name is going to waste your first three months.
Why publishing more of your own pages does not move it
This is where most people waste their first year.
The instinct when you feel invisible is to publish more: more blog posts, more LinkedIn, more videos. All of that is useful for other reasons and almost none of it moves the entity needle, because everything you publish about yourself is a claim. Claims do not corroborate. Only independent sources do.
Think about what Google is doing. It is trying to decide whether to stake its own credibility on a statement about you. Your own website saying you are the leading advisor in your field is exactly as persuasive as a stranger telling you they are trustworthy. It is not that Google distrusts you. It is that self-description carries no information about consensus, and consensus is the only thing being measured.
Look at what is holding this one up: a Wikipedia article, a net worth figure attributed to Forbes, a Bloomberg video, a LinkedIn card. Four sources, none of them hers.
She has a company site and a substantial public profile of her own, and Google reached for none of it when deciding what to show. That is not a slight. It is the mechanism: self-description carries no information about consensus, so the system reaches for the sources that are independent of the subject.
Stop asking “what should I publish this month?” and start asking “who could credibly say my name in public this month, and what would make that easy for them?” The second question is harder, slower, and it is the one that works.
What Google is actually holding on you today
Enough theory. Check the name.
More than one result means more than one record answers to that name. Every extra record is a place Google’s certainty leaks, and whoever holds the highest match strength owns the name in practice.
The two stages, and where the gap actually is
Google does not flip a switch. Your record goes through two distinct states.
Stage one is an internal identifier and a thin file. Ask for the entity and Google will show you something. Type the bare name and it may show you nothing.
Stage two is the panel firing on the bare name.
Four of the eight records on our proof page are at stage one right now. Sagar Kathrotiya has eleven images, LinkedIn, Instagram, YouTube and Facebook all resolving to one person, and his panel still does not fire on his bare name. Katerina Khramova has an IMDb record, a Facebook profile and a match strength near a hundred, same situation.
Both of them have the structure right. What they are short of is recent independent mentions, which is what stage two is actually gated on. This is worth internalising, because it changes what you spend the next six months doing.
The order of operations, and why skipping fails
- Diagnose before promising. Google the name, read the Knowledge Graph, count the namesakes. Three verdicts, three routes, and one of them is much cheaper than the other two.
- Write the three sentences. Who you are. What you stand for. What result you drive, for whom. If these cannot be answered in plain language, nothing downstream can broadcast an identity nobody has decided on.
- Build the brand brain. One private file with every fact anyone will ever state about you. This is the highest-leverage hour in the project, because missing facts here are exactly where wrong facts come from later.
- Stand up the entity home. Then and only then.
- Ship the schema. Person on the entity home, Organization on the company site, wired together, complete profile list.
- Align every profile. One string everywhere.
- Instrument it. Search Console on an account you will still control in five years, because that same account is required for the claim.
- Manufacture real proof. Publish to the entity home first, mirror outward second.
- Densify the cross. Three to five witnesses on each axis. This is the six-month part.
- Add the accelerators. Authorship first, then Wikidata if and only if the notability bar is genuinely cleared, then cohort proximity through repeat podcast appearances.
- Claim it. Covered in the next guide.
- Hold it. Quarterly, forever. Drift is silent.
Skipping is the single most common cause of failure, because each step assumes the previous one is true. Corroboration pointing at a site that does not clearly say who you are just spreads the confusion wider.
Questions people ask next
How many independent mentions do I need?
There is no published number and anyone quoting you one is guessing. As a working target: three to five credible independent sources on your trade axis and three to five on your geographic axis, each naming you, linking to your entity home, and repeating your identity line closely. Relevance beats raw authority every time.
Does Wikidata guarantee a knowledge panel?
No. A clean, well-sourced Wikidata item is a strong corroborating signal and it is not a trigger. An unsourced item that gets deleted is worse than never having created one. Person schema on your own site quietly does most of what people hope Wikidata will do, and it is fully under your control.
Should I buy a domain with my name on it?
If it is available and reasonably priced, yes. A name-matched domain is the cleanest possible signal of an entity home. It is not required — a clearly-structured personal section on another domain works — but it removes an ambiguity you would otherwise spend effort compensating for.
Can I speed this up by buying links?
You can spend the money and it will not work, and it can do real damage. What Google is measuring is agreement among sources it independently trusts. Bought links are neither independent nor trusted, and a footprint of them is a liability that outlives the campaign.
My company has a panel but I do not. Why?
Because they are separate entities. A company earns a record through business signals: reviews, a Business Profile, press, registration data. A person earns one through personal signals. The fix is Person schema on your entity home wired to the Organization schema on the company site with the founder and worksFor relationships filled in, so Google can see they are two connected things rather than one confused thing.
Run your own name through the entity check, then score yourself with the readiness tool. Both are free and neither asks for an email address. If the reading comes back saying you are ready, the build page has the price and the terms.