What Makes a Publication Citable? Authority, Crawlability, Freshness, Fit

A publication is citable when answer engines can reach it, already trust it, and consider it relevant to the question being asked. In practice that comes down to five checks: authority, real readership, crawler access and indexation, freshness, and topical fit, with editorial standards underneath all of them.

Domain metrics cover only the first of those. A site can carry a high Domain Rating and still be a poor bet for AI citations because it blocks AI search crawlers, its sponsored section isn't indexed, it never covers your niche, or its "authority" comes from links rather than readers. This guide explains each check, how to verify it yourself, and how InTheAnswer's AEO score turns the measurable parts into a single number.

If you're new to why publications matter at all, how backlinks help your brand get cited covers the chain from a placement to an AI citation. This guide is about choosing the publication.

The five checks at a glance

Before buying a placement for AI visibility, check that the publication passes all five.

1. Authority: the domain is established and well-linked, so its pages rank in the indexes AI engines search.
2. Real readership: it earns organic traffic in line with its authority, not just links.
3. Access and indexation: AI search crawlers are allowed in, and articles like yours get indexed and are eligible for snippets.
4. Freshness and editorial standards: it publishes regularly, dates its content, and has real authors and editing.
5. Topical fit: it already covers, and ranks for, the subjects your buyers ask about.

Authority, checked against real traffic

Answer engines that search the web inherit the ranking preferences of their index, and search indexes favor domains the rest of the web links to. That's why third-party authority scores such as Ahrefs Domain Rating (DR) and Moz Domain Authority (DA) make a sensible first filter: they approximate how well a site's pages are likely to rank.

But authority scores measure links, and links can be manufactured. The classic warning sign is a high DR with almost no organic traffic: plenty of sites link to it, yet search engines don't send it readers. That pattern usually points to a link network rather than a publication people read, and since AI engines retrieve from the same search rankings that pass over those sites, the high DR buys you little. Buying on DR alone is one of the errors covered in backlink mistakes that hurt AI visibility.

Real traffic is the second opinion. A publication that ranks for many queries and draws readers from search is, almost by definition, one that search-grounded engines encounter often. The traffic doesn't need to be huge; it needs to be consistent with the authority the site claims.

Crawler access and indexation: the gate most buyers skip

A placement can only be cited if the engine can retrieve it. Two things decide that: whether the publication's robots.txt admits the right crawlers, and whether the specific article gets indexed.

Search crawlers versus training crawlers

AI companies run separate crawlers for search and for model training, and publishers often treat them differently. For citations, the search side is what counts: OAI-SearchBot for ChatGPT search, PerplexityBot for Perplexity, Claude-SearchBot for Claude, Googlebot for AI Overviews and Bingbot for Copilot. Training crawlers such as GPTBot and ClaudeBot, and Google's Google-Extended token, govern what goes into future models.

So a publisher that blocks GPTBot but allows OAI-SearchBot can still appear in ChatGPT search answers, while one that blocks OAI-SearchBot won't, according to OpenAI. Perplexity likewise documents PerplexityBot as the crawler that surfaces sites in its answers (Perplexity's crawler documentation). Google says Google-Extended doesn't affect inclusion in Google Search (Google's crawler documentation); for AI Overviews, Googlebot access and snippet controls are what matter. The engine-by-engine detail is in how AI answer engines choose sources.

How to check a publication in five minutes

1. Open /robots.txt on the publication's domain and look for user-agent groups that name the search crawlers above. A Disallow: / under one of them blocks that crawler from the whole site.
2. Read the User-agent: * group as well, since it applies to any crawler without a group of its own.
3. Find an existing sponsored or contributed article on the site and search for its exact title in Google and Bing to confirm it's indexed.
4. View that article's source for a robots meta tag. In Google, noindex keeps a page out of the index entirely, and nosnippet or max-snippet limits what can be shown, including in AI Overviews.
5. If sponsored content lives on a separate subdomain or section, check that location specifically, because its rules and indexing can differ from the main site.

The free AEO audit automates the first two steps: give it a URL and it reports which AI search and training crawlers the site's robots.txt allows.

Freshness and editorial standards

Engines answering time-sensitive questions (this year's best tools, current pricing, recent funding rounds) likely give weight to recency, and a publication that dates its articles and publishes steadily gives them recency to work with. A site that hasn't published in months is a weaker bet even with a strong archive, partly because new pages on rarely updated sites tend to be crawled less often. Both points are reasoned inferences rather than documented rules, but they match how search engines generally allocate crawling.

Editorial standards matter for a related reason. Real bylines, author pages, corrections and an identifiable editorial team are signs of a publication that readers trust and search engines tend to rank. They also matter to the person who clicks through from a citation and lands on your mention. A site made up entirely of unsigned paid posts is unlikely to rank, and so unlikely to be retrieved.

What to look for:

• Recent articles in your niche, with visible publish or update dates
• Named authors with a track record on the site
• A real mix of editorial and sponsored content, not sponsored posts alone
• Sponsored content that is labeled rather than disguised

Topical fit: be where the question gets answered

Retrieval is topical. When an engine searches for "best expense management software for startups", it gets back pages about expense management and startups from sites that rank for that subject. A placement on a respected travel site won't be in that list, however strong the domain.

Fit works at three levels:

• Vertical: the publication covers your industry at all.
• Section: your article would sit in a section that ranks for your topic, such as a finance site's small-business coverage rather than its celebrity net-worth pages.
• Article: the piece itself addresses a question your buyers ask, rather than being a generic brand profile.

Some verticals are also drawn on more heavily than others. InTheAnswer's method gives its highest topical-citability weight to news, tech and SaaS, business and finance, and science publications, reflecting the kinds of sources answer engines tend to cite most for factual and commercial questions. That's a tendency, not a rule: a well-ranked specialist site will often beat a general news site for a narrow question in its own field. You can browse publications by vertical, for example business and finance or tech and SaaS.

How InTheAnswer's AEO score turns this into a number

InTheAnswer scores every publication in its catalog from 0 to 100 on how likely answer engines are to retrieve and cite it. The score combines four weighted signals:

• Authority (55%): Ahrefs Domain Rating, with Moz Domain Authority as the fallback when DR isn't available.
• Real traffic (20%): monthly organic traffic on a log scale, so the jump from a few thousand visits to a few hundred thousand counts for far more than the jump from ten million to twenty million.
• Topical citability (10%): a boost for the verticals answer engines cite most, led by news, tech and SaaS, business and finance, and science.
• Link hygiene and placement (15%): whether organic traffic is consistent with the site's authority (so high-DR, no-traffic link farms score low), whether the link is dofollow, whether the placement is available from more than one vendor, and whether it carries a sponsored label.

Scores map to four tiers:

• Elite (75–100): top-tier news and widely cited domains
• Strong (60–74): high-authority niche publications that engines retrieve for their topics
• Standard (45–59): mid-market sites with real traffic, useful for diversifying
• Emerging (0–44): lower-authority sites, best used as supporting links

The score is a way to prioritize, not a prediction for any single prompt. It can't see whether a specific article will be indexed, whether a publisher changes its robots.txt next month, or whether a given section fits your topic, which is why the manual checks above still matter. Traffic data is verified for 6,806 placements, and the full methodology is on the method page.

To see the highest-scoring options for your own market, the niche link finder ranks publications by AEO score for your niche, budget and audience.