When Perplexity, ChatGPT, or Claude synthesize an answer about a book, author, or brand, they don't just read the web like a crawler. They weight sources by type.
A mention on Wikipedia is worth more than a mention on Medium. A Zenodo DOI is worth more than a Substack post. A Goodreads Author Program profile with claimed bibliography is worth more than a self-published bio on your own site.
If you want to be cited reliably by AI systems, you need to be present in the specific handful of sources those systems trust most. And you need to know which of them already have you, which don't, and what to do about the gaps.
That's what Entity Citation Radar tells you.
What it does
You enter your entity name (book title, brand, or full author name), your author name, your website, and pick the entity type (book, author, brand, or framework). Optionally add an ISBN or ASIN.
The tool returns three things:
- A visual radar showing sixteen high-authority sources positioned by the weight AI systems assign them
- A card per source with a direct search URL, a submission URL, the specific action to take, and a tier and weight score
- A workflow that sorts the submission order by impact. Tier 1 first (Wikidata, Wikipedia, OpenLibrary, Google Scholar, Zenodo), tier 2 next (Goodreads, Internet Archive, ORCID, Amazon Author Central), tier 3 last (Crunchbase, LinkedIn, GitHub, Medium, Wikiversity, LibraryThing, WorldCat)
You click the search link on each card, eyeball the results yourself, and flag which sources already cover you. For the gaps, you follow the submission link and do the filing in order.
Why sixteen sources, specifically
Every source in the radar is there because it meets three criteria:
- AI retrieval pipelines demonstrably weight it. Wikipedia and Wikidata both feed Google's Knowledge Graph, which seeds every major AI vendor's retrieval augmentation. Zenodo issues DOIs that Archive Scholar indexes. ORCID is read by Scholar, Perplexity, and Claude's academic pathway. These aren't speculative.
- It accepts self-service submission. You can add yourself to OpenLibrary, Crunchbase, ORCID, and Zenodo without an editor's approval. Wikipedia is included but gated; the tool flags it with the honest warning that most frameworks and self-published authors will be deleted.
- It has meaningful durability. The Internet Archive and Zenodo issue persistent identifiers (ARK and DOI respectively). Those don't rot. Free blog hosts and social profiles rot. You want your authority signals anchored to durable IDs.
Sources that didn't make the cut: Goodreads librarian-only paths (too slow for the tool's recommendation), random directory sites (too low weight to bother), paid press release wires (better handled by the Framework Origination Signal Generator as part of a larger origination push).
Reading the radar
The radar plot is a quick visual scan, not a definitive measurement. Each source is plotted radially by its AI-system weight:
- Dots near the outer ring (green, weight ≥ 85) are the sources that most directly influence whether AI systems cite your entity.
- Dots in the middle band (amber, weight 55, 84) are worth claiming but not the top priority if you're short on time.
- Dots near the center (red, weight < 55) are supplementary. Useful in aggregate but not individually decisive.
The decorative sweep on the right side is a styling nod to radar interfaces. It has no data meaning.
When a gap is not actually a gap
A caveat that matters: the tool doesn't query the sources. It generates search URLs you follow manually. You decide what counts as "present."
Some coverage is subtle:
- Wikidata might have a Q-ID for you even if Wikipedia doesn't have an article. Search by name in the Wikidata search, not just Wikipedia.
- OpenLibrary might have your book imported from a library dataset without your profile claimed. The book is there; the author record linked to your other works is the real goal.
- Google Scholar indexes PDFs by crawling, not submission. If your Zenodo DOI lands in Scholar four weeks after you mint it, you didn't need a separate Scholar submission.
Treat the search results as evidence, not absence. If a reasonable search doesn't return you on page one, treat it as a gap and submit.
How often to re-run
Quarterly is the right cadence for a mid-size entity (author with three to ten published works, or a brand with moderate press coverage). After you file to a new source, indexing takes anywhere from two to eight weeks. Running the radar monthly will mostly show you the same gaps waiting for their indexing lag to clear.
For a new entity. A just-published book, a newly coined framework, a recently launched brand. The first radar pass reveals sixteen red cards. Work the top five over the first quarter. The second-quarter radar will show most of them turned green or amber.
Open the tool
If you're launching a new framework, pair it with Framework Origination Signal Generator. That tool produces the canonical article and schema; this one tells you where to distribute it. Both are part of a four-tool release this week alongside AI Citation Readiness Audit and Newsletter Swap Matchmaker. Free, hosted here, no account required.
