Skip to content

The fix, not the score

How to get cited by AI

The engines already told you which pages they read. The ones that get cited again and again, name a competitor every time, and have never named you, are a to-do list with names on it.

Start 7-day free trialHow we track citations

Card required, nothing charged for 7 days

Most advice about this is unfalsifiable

Write better content. Build authority. Earn mentions. All of it is probably true and none of it tells you what to do on Monday, because none of it names a page.

There is a shortcut that most of the advice skips: when an assistant answers a question about your category, it says where it looked. The sources are in the answer. Collect enough answers and the set of pages the engines keep returning to stops being a guess and becomes a list you can count.

71 sites for one brand, and why only eight are here

These are ours. Posteverywhere is our own company, and this is its real list, filtered in public rather than quietly.

The eight neutral editorial sites from our top twenty. Seven-day window, measured 2 October 2026. Our brand was named in none of them.
Site the engines readAnswers citing itNamed a competitorNamed us
zapier.com156across 5 engines0
socialpilot.co136across 3 engines0
sintra.ai116across 3 engines0
oneupapp.io95across 3 engines0
postplanner.com95across 1 engine0
beplan.io86across 3 engines0
technologyadvice.com85across 4 engines0
socialchamp.com76across 3 engines0
hubspot.com75across 1 engine0
techradar.com75across 2 engines0
dashsocial.com66across 3 engines0
napoleoncat.com66across 2 engines0
loomly.com65across 2 engines0
thecmo.com65across 2 engines0
thenextscoop.com65across 1 engine0
lyfemarketing.com64across 2 engines0

zapier.com was read fifteen times across five engines while assistants answered questions about our category. Sixteen of those twenty citations were one page, “The 10 best social media management tools in 2026”. It named six of our competitors and has never named us. That is not a ranking problem, and no amount of writing on our own site changes it. It is one listicle, and we are not in it.

The sources screen listing the domains AI engines read before answering, with a column counting the answers that cited a domain without naming your brand — the cited-but-not-named state.
The domains engines read, and how often they read one and named someone else. Our own workspace — Posteverywhere is our company, and these are its real numbers.

Why the raw number is bigger than the useful one

The product reports 71 for us over seven days and 92 over thirty. Eight are in the table above. We would rather show you the subtraction than a clean list you would catch us filtering, because anyone who opens the product sees the raw one.

Platforms with no editor to contact

3

User-generated content and jobs sites. The engines genuinely read them and they genuinely name competitors, but there is nobody to pitch, so they are not work a reader can do. This is the only reason anything is held back from the list below.

  • linkedin.com
  • reddit.com
  • indeed.com

The second one is a hole in our own rule and we are fixing it. The computation drops a tracked competitor's own domain, and we track six, so every other rival's blog passes straight through. For a social media tool that means a dozen rivals writing listicles about each other, and the largest single row over thirty days is a competitor. Pitching them is not advice we would give.

And the number moves, which is the point. Our demo video says 59, and it was right on 1 October 2026. Twelve more sites qualified within a day. A count without its window and its date is not a fact, which is why both appear beside every figure here and in the product.

What to do with the list

  • Start with frequency, not authority. A site cited in twenty answers about your category is worth more than a famous one cited in none. The list is already sorted by the only metric that applies.
  • Read the page before you pitch it. Most of these are roundups, comparisons and integration directories. The question is usually why you are not in a list, not whether you could write a guest post.
  • Check whether you are already on it and unnamed. Being present on a page the engine reads, and still not named in the answer, is a framing problem on that page rather than a coverage one. Fixing the paragraph beats publishing ten posts.
  • Ignore anything a competitor owns. Their own site naming them is not a gap, and the computation already drops those.
  • Re-check after the work lands. The engines have to read the page again before anything moves, and that takes weeks rather than days.

How this is computed

From stored answers, not from a crawl. Every answer we collect keeps its citation list, so a gap is a domain that appears in those citations at least twice in the window, appears alongside a competitor being named, and has never appeared alongside your own brand being named. Sites a competitor owns are excluded.

It is the Sources screen in the app, on every plan, and the same figures come back from the API as answers_citing_you_without_naming_you. If you would rather read the raw evidence, we wrote up what cited-but-not-named means and why it needs the opposite work to being absent.

Questions about getting cited

Answered from what the product computes.

How do you get cited by AI?
By being on the pages the engines already read. When an assistant answers a question about your category it cites a handful of sources, and those sources are observable: they appear in the answer. The work is to find the ones that are read often, name your competitors, and have never named you, then get yourself onto those specific pages. It is a list with names on it rather than a content strategy.
Is this the same as link building?
No. A backlink tool tells you who links to you, which is a fact about the web. This is computed from stored AI answers, so every row is a page an engine actually read while answering a question a buyer asked. A site can have enormous authority and never be cited in your category, and a modest site can be cited in half the answers. Only the second one matters here.
Why does being cited matter if I am not named?
Because they are separate states and they need opposite work. Being cited and not named means the engine read your page and recommended somebody else, which is a framing problem on that page: more content on the same topic will not fix it. Never being cited at all is a coverage problem. Telling them apart is the point of storing the answers.
How many of these sites are there usually?
It depends on the category and the window, and the raw count is bigger than the useful one. For our own brand the product reports 71 over seven days and 92 over thirty, but a chunk of those are platforms nobody can pitch and a chunk are competitors' own blogs running comparison posts. Eight of our top twenty are neutral editorial sites. The number also moves daily, which is the argument for watching it rather than auditing once.
Can I see this for my own brand?
Yes. It is the Sources screen in the app, computed from the answers collected for your brand, on every plan including the cheapest. It is also available over the read-only API as answers_citing_you_without_naming_you.

7-day trial

Get your own list

Add your brand and the prompts your buyers type. The sites the engines read appear as the answers come in, from $9 a month with all six engines on every plan.

Start 7-day free trialHow citation tracking works

Card required, nothing charged for 7 days.

On every plan

  • Every answer stored in full, with its citations.
  • Every rate shown with the number of checks behind it.
  • Cited-but-not-named reported as its own state.