Schema and the H1 disagree
JSON-LD prints one legal name. The visible heading prints another. Keyword targeting both in titles will not pick a thing. We write one official line.
Entity SEO
Entity SEO
Entity SEO from senior consultants who turn search demand into rankings, traffic and revenue for your business. We start with a free audit and a 90-day plan built around your goals.
- Visible
- IDs
- JSON-LD
Our process
How the work runs
- 01
Visible
We agree the official name your website will use for your organisation, everywhere.
- 02
IDs
We list identifiers you already hold, such as your legal entity, registry numbers and trademarks.
- 03
JSON-LD
We check your published structured data matches what the page says, so Organization or Person markup names the same entity.
- 04
Profiles
We list the profiles you control and keep active, and remove dead social links from your markup.
Position
The site already names a thing. We write which thing that is.
A homepage can say a trading name. The About H1 can say a parent. The JSON-LD can say a product. Search intent for the company people look up still has to land on one named thing. This page owns on-site identity: official name, identifiers, Organization or Person markup as identity, and sameAs URLs you will keep. It does not start from a Knowledge Panel. It does not sample assistant binds. It does not cover related concepts across a cluster.
Knowledge Graph optimization documents facts Google already shows in a panel. AI entity optimization tests whether models merged two companies. LLM brand visibility checks the spoken string in answers. Semantic SEO covers meanings on live documents. Digital PR owns new coverage. The ticket here is what you publish about the named thing on your own site.
We will not sell a Wikipedia page. We will not treat a schema plugin as the identity spec. We will not add every directory to sameAs. If legal cannot sign the official name, we stop. If the panel is wrong and the site is already consistent, that ticket belongs on Knowledge Graph optimization.
Compare packages if a published retainer already fits. Use Knowledge Graph optimization when a related service is the real job. If this page is still a guess, start with a free AI + SEO audit.
When this fits
When the site cannot agree which named thing it is
This page fits when official pages, identifiers, and markup disagree. If you needed a panel pass, a model bind test, a meaning pass, or a newsroom, start there.
Identifiers are missing or belong to a parent
No company number, the wrong DUNS, a product SKU used as the org. Topical relevance of a blog is not an identifier. We list what you will keep.
sameAs points at dead or sold profiles
A LinkedIn you no longer control, a Wikipedia draft, a directory you will not maintain. Semantic keywords will not clean the array. We keep only URLs you will own.
Someone promised a schema plugin would mint a panel
We will say so. An identity spec can still be useful. A panel is not a deliverable here. If the unknown is public facts already shown, that is Knowledge Graph optimization.
Process
Visible, IDs, JSON-LD, profiles
Four jobs, in that order. Each job produces a note legal and the CMS owner can keep. We do not start a panel campaign in the last job. The point is identity on your site, not a box Google owes you.
-
Visible
Outcome A dated official name line, including aliases that may still appear in copy.
-
IDs
Outcome An identifier list with gaps named, not a directory shopping list.
-
JSON-LD
Outcome A markup note: match, mismatch, or two speakers.
-
Profiles
Outcome A sameAs list with owners, not an infinite profile crawl.
Deliverables
What legal and the CMS can keep after we leave
You leave with a name line, an identifier list, a markup note, and a sameAs list. You do not leave with a Knowledge Panel, a Wikipedia article, or a ranking date.
Official name line
The string the site will stand behind. Dated. Last year's brand book is not this line if legal moved.
Identifier list
Registry and brand IDs you will keep. Parent IDs used by mistake are named.
Markup note
Whether Organization or Person JSON-LD matches the HTML. Two speakers are named and stopped.
sameAs list
Profiles you control and will maintain. Dead URLs are removed on purpose. Fake directories do not go in.
Owner map
Who can change the name, the markup, and each sameAs URL. An unread spec is not a failure of this page. A purchased panel would be.
Benefits
What a panel screenshot cannot replace
The site speaks one official name
H1, footer, and markup stop arguing. Extractors get one thing.
Identifiers have an owner
A parent company number stops pretending to be the child brand.
Dead profiles leave sameAs
The array uses what you control now. Old social URLs are retired on purpose.
The spec has a date
Someone can recheck after a redesign. Nobody pretends a plugin bought a Knowledge Panel.
Methodology
Who can sign the official name on the site
This works when legal can confirm the name, and someone can edit About and JSON-LD. If both are missing, we are collecting mismatches. We are not specifying identity.
We start with the legal name, the live About page, and the published markup. If any of those is missing, we name it first.
We fetch published HTML and JSON-LD. We do not trust a staging plugin preview. Does this H1 match the official line? Is another org in the same Organization block?
You may rename the roles below. You may not leave the named thing implied. An implied alias is how two brands share a site.
- Entity SEO lead SEOConsultants.ai writes the name line, the identifier list, the markup note, and the sameAs list. We do not become your PR agency because schema was messy.
- Legal or brand owner Confirms the official name and aliases. Every name needs a clear owner.
- Whoever edits the official site Can change About copy and Organization markup. They print the identity. They do not need to become panel editors.
- Whoever owns official profiles Can update URLs that remain in sameAs. They are not asked to build fake profiles.
- Leadership Accepts that identity on the site is not a purchased Knowledge Panel. An unread spec is not a failure of this page. A Wikipedia package would be.
First working session
Start with the H1 and the Organization block
The first hour is the visible name versus the name in markup. It is not a tour of Knowledge Panel tools.
We open the homepage and About. We list the visible name. We fetch JSON-LD. If they disagree, that is the job. If a public panel is the gap, that is Knowledge Graph optimization. If a model fusing two firms is the gap, that is AI entity optimization. If meanings on documents are the gap, that is Semantic SEO. If a Maps listing is the gap, that is Google Business Profile.
Then we list identifiers you already have. We list sameAs URLs you still control. The extra or missing names are usually the job. A Knowledge Graph API demo is not evidence of what your HTML states.
We leave with a first identity spec or a stop. A spec names the official line, the mismatches, and who can edit markup. A stop names what is missing: legal sign-off, a job that is actually a panel, or a job that is actually a model bind. Both are outcomes. A Wikipedia shopping list is not.
AI search
An identity spec is not an AI Overview
A consistent named thing on the site can help later extraction. This page cannot sell AI Overviews, a Knowledge Panel, or LLM visibility as a score.
People ask whether cleaner Organization markup will land the brand in Google AI Overviews (AIO). If the site still teaches two legal names, generative search has a messy thing to lift. That is ours to specify as identity. If the site already agrees and models still fuse you with a namesake, that is AI entity optimization. If the panel is wrong, that is Knowledge Graph optimization.
The official page has to state the name
AI search needs HTML it can fetch. A brand deck that never published does not help. Nobody can guarantee a spot in AI Overviews, but clear pages improve your chances.
On-site identity is not a public panel
We check H1, identifiers, and markup. Panel facts people already see live on Knowledge Graph optimization.
Model binding is another URL
AI entity optimization tests whether assistants merged two companies. This page specifies the thing on your site.
Schema
Markup is identity here. It is not a panel purchase.
Organization or Person JSON-LD must name the same thing the page states. This is the schema job on this page. It still does not buy a Knowledge Panel or a cite.
The name in markup must match the About page
If schema says one legal name and the H1 says another, extractors guess. We name the mismatch. We keep names clean instead of stuffing aliases.
sameAs only for profiles you will keep
Dead social URLs in markup train the wrong identity. We list which links may stay. Fake directories do not go in the array.
Extra types do not force a box or an Overview
Markup will not force generative search to cite you, and it will not force a Knowledge Panel. Good markup makes you eligible; it cannot buy a mention.
Who we work with
The person who can change About and JSON-LD
Someone has to sign the official name and ship the markup. We work with that person. We do not replace them with a second identity retainer.
Legal or brand owner
Confirms the name, aliases, and identifiers. Holds the line when marketing wants a slogan in the Organization name.
Whoever edits official pages
Can update About copy and markup after the spec. They print the identity. They do not need to become panel editors.
Whoever already runs schema plugins
Can keep going once one speaker stores Organization. If they try to run this as a panel campaign, two jobs collide. Panels stay on Knowledge Graph optimization.
FAQ
Entity SEO FAQs
What named thing does this page specify?
The organization, product, or person your HTML and schema will stand behind: official name, identifiers, Organization or Person markup, and sameAs you will maintain. Knowledge Graph optimization works public Knowledge Panel facts. A sameAs spreadsheet is not a panel correction. A panel screenshot is not an identity spec.
How is this different from AI entity optimization?
AI entity optimization tests whether an assistant fuses you with a namesake. This page is on-site identity. A merge sample is not an identity spec. An identity spec will not by itself unmerge two companies in a model. Hand a fused answer to AI entity optimization.
How is this different from Semantic SEO?
Semantic SEO covers related meanings on documents you publish. Entity SEO names the thing those documents are about when the thing is an organization, product, or person you will stand behind. A topical cluster is not an official name. A legal name is not a meaning pass.
Do we need a Wikipedia page to specify identity?
No. Wikipedia is not a product on this page. Many brands have no article and still need a consistent official name, identifiers, and markup.
Will Organization markup get us into AI Overviews?
No. Honest identity on the site can help later extraction. This work does not buy a place in Google AI Overviews (AIO), a Knowledge Panel, or an LLM visibility score. If the site already states one name and models still fuse you with a namesake, that measurement sits on AI entity optimization.
Identity the HTML will own
Does the homepage H1 disagree with the Organization name?
Paste the conflicting H1 versus schema names, plus the legal name you will keep. We will say if the identity spec is the gap, or whether a Knowledge Panel, a model bind, or a meaning pass should go first.