Answers
Does publishing more content get you cited by AI assistants?
Short answer
Does publishing more content get you cited by AI assistants?
Rarely, past a small floor. Across our own measurement of 901 citations spanning 82 distinct hosts, 91.5 percent of cited pages were off-site, meaning the surface deciding most recommendation answers is not one you can add pages to. Publishing more on your own domain helps you occupy more of your own 8.5 percent, and it does nothing to the 91.5 percent that is somebody else's roundup, comparison post or community thread. A vendor shipping 20 to 28 content units a month is mostly filling a number a buyer can count, and the work that actually moves a citation, getting named accurately on the pages already being retrieved, does not scale with page count at all.
The short version
901 citations across 82 hosts, our own measurement on our own money queries.
01
Where the 901 citations we measured actually pointed
We built our own money-query set and measured, rather than assumed, where the answers assistants gave were actually sourced from. Across 901 citations and 82 distinct hosts, 91.5 percent pointed at a page nobody in the answer owned: comparison roundups, community threads, and documentation belonging to a different project entirely. Only 8.5 percent were the vendor's own site being cited directly.
That split sets a hard ceiling on what publishing volume can do. Even a team that doubled its own page count would still be optimizing the smaller slice of a split that runs nine to one the other way, and the pages doing the deciding would be untouched by any of that work.
Our own measurement, 901 citations
Doubling your own page count still competes for the smaller slice of a split that runs nine to one the other way.
02
Citations decay, and volume does not fix decay
A separate finding worth holding next to the volume question: cited content decays. A competitor AI-visibility platform tracking citation age publicly reports that half of all cited content is under 13 weeks old, meaning what gets pulled into an answer skews recent rather than accumulating with age the way search rankings can.
That looks like an argument for constant publishing, and it is not quite that. Decay argues for timed refresh of the pages already proven to get cited, not for growing the pile, because a fresh update to a page an assistant already retrieves is a different intervention than a new page competing for the same retrieval slot from zero authority.
A competitor platform's citation-age reading
Half of cited content is under 13 weeks old. A timed refresh on a proven page is a different move than a new page from zero authority.
03
What the monthly worklist actually counts, and why it is not a page count
Our own delivered worklist is deliberately not sized by publishing volume. It ships 6 to 12 owned answer pages a month depending on tier, and 6 to 15 off-site placements a month on top of that, roughly matching or exceeding the owned-page count rather than treating it as the main lever. That ratio is a direct reflection of the 91.5 percent split: most of the monthly work targets pages we do not control.
A vendor whose entire monthly deliverable is a page count is optimizing a number a buyer can verify at a glance, not the number that determines whether an assistant cites you. The two are not the same axis, and conflating them is how a slower, off-site-heavy program gets outbid by a faster-looking one that moves nothing.
Shipped monthly, Own it tier
Off-site and community placement roughly matches or exceeds the owned-page count, a direct reflection of the 91.5 percent split.
Asked next
The questions that follow this one
Should I stop publishing new pages entirely?
No. On-site pages are still where you get to state your own answer capsule cleanly and where roughly 8.5 percent of citations land directly. The claim is narrower: past a working floor, more volume does not move the 91.5 percent that decides most answers.
How much content is actually enough?
Enough to cover your own money queries with a clean, extractable answer each, which our own tiers size at 6 to 12 pages a month. Past that, the marginal page competes with off-site placement work for the same budget and usually loses on impact per hour.
Does refreshing an old page work better than a new one?
Often, yes, when the old page is already being cited. A citation-age reading of under 13 weeks median suggests a timed refresh on a proven page can re-enter the retrieval set faster than a brand-new page building authority from zero.
Why do some vendors sell by content volume then?
Because a page count is the easiest thing to show a buyer and bill against. It does not track the 91.5 percent off-site split we measured, so it is a proxy for effort rather than a proxy for the thing that actually decides a citation.
Before adding more pages
Free gap report
When the model answers,
be the one it names
Tell us your category and the questions your buyers ask. We run them against live AI answers and walk you through what came back. If you are already winning, we will tell you that too.
12 queries · 5 repeats each · 5 working days
Free · Walked through live · Five working days
Queries we run
Cited instead of you
reddit.com38%
g2.com26%
news.ycombinator.com21%91.5% of citations point off-site