Consistency matters more than volume. One genuinely good, well-structured piece a week outperforms sporadic bursts, and a sustainable cadence you can hold for a year beats an ambitious one you abandon in March. Below roughly two pieces a month, momentum is hard to build at all.
The short version
- Sustained cadence beats total volume. Stopping undoes more than slowing down would have.
- Roughly weekly is a good target for most firms. Two a month is a workable floor.
- Restructuring existing pages counts, and usually returns more per hour than new writing.
- Output beyond your capacity to keep it specific and accurate is counterproductive.
- Named-author output is worth more than the same volume published anonymously.
Why consistency beats bursts
Citation authority accumulates through repeated corroboration over time. A model does not decide you are authoritative because thirty pages appeared in one week. It builds confidence as material accumulates, gets referenced, and holds up.
Bursts also fail for a mundane reason: they stop. The common pattern is a quarter of enthusiastic publishing followed by nothing, because the cadence was set at what was possible during a motivated month rather than what the business can actually sustain. The programmes that work are the ones whose pace survives a busy period, and that pace is almost always lower than the one people commit to at the start.
There is a real threshold effect at the bottom. Below roughly two substantial pieces a month, it is difficult to build topical density on anything, because coverage is spread too thin to make you the obvious answer on any one subject.
Where more output stops helping
Volume has a ceiling and it arrives sooner than most content plans assume.
Models reward depth on a topic over breadth across topics. A firm with twelve strong pieces on one specialism outperforms one with two hundred spread across everything. Once you are publishing enough to cover your specialism properly, additional volume in new directions dilutes rather than compounds.
The harder limit is quality. AI citation rewards specific, checkable claims. Specificity requires somebody who genuinely knows the subject to spend time on it. Push output past the rate at which your experts can genuinely contribute, and the content becomes generic, which makes it uncitable, which means the extra volume produces nothing. Doubling output while halving specificity is a net loss.
How to check it yourself
Count what you have actually published in the last six months and divide by six. That, not your plan, is your real cadence. Then ask how much of it contains a specific, checkable claim that a model could quote. The second number is usually much smaller than the first, and it is the one that matters.
What to do about it
- Set the cadence at what you can sustain for twelve months, then hold it.
- Count restructuring as output. Reworking an existing page for extraction often beats a new one, and it is faster. See does my existing content help.
- Go deep on one specialism before widening.
- Protect expert time rather than expert sign-off. Generic content published quickly is not a cheaper version of the same thing, it is a different and less useful thing.
- Publish under named authors with real credentials attached.