How often should we publish new content?
Steadily and indefinitely beats fast and short. For most service businesses a sustainable cadence is a handful of answers per week, published on a schedule you can hold for a year, working one cluster at a time until it is complete. Publishing forty pages in a single week and then nothing for six months reliably produces worse results than the same forty spread out.
Why drip beats dump
A large batch of pages appearing at once is crawled slowly, evaluated as a block, and gives you no ability to learn. If the batch underperforms, you have forty pages and no signal about which decisions were wrong.
A steady stream gets crawled more predictably, lets you correct format and depth after seeing early results, and produces a site that looks maintained rather than campaigned. The last part is not a formal ranking factor but it correlates with everything that is.
Depth first, then breadth
The most common planning error is spreading a year of output thinly across every service line so nothing reaches critical mass anywhere. Twenty answers about one decision will outperform two answers each about ten decisions, because the first pattern builds a cluster and the second builds a scatter.
Pick the cluster closest to revenue, finish it, then move. Finishing means the obvious follow-up questions are answered and a reader can go from first curiosity to ready-to-call without leaving your site.
Your real constraint is review capacity, not writing capacity
Drafting stopped being the bottleneck. Verification did not. Before you set a cadence, measure how many pages your subject matter expert can actually check per week without resenting it — and then set the cadence at or below that number.
This is why cadence conversations that start with "how many pages can the system produce" go wrong. Production capacity is close to unlimited and irrelevant. Set the schedule from the human constraint and treat the automation as the thing that removes everything except the review.
When to deliberately slow down
Slow down when the question queue is running on invented questions rather than mined ones, when review is being rubber-stamped to hit a number, or when new pages are starting to overlap old ones. All three are signals that you have outrun your supply of things worth saying.
Pausing new pages to improve existing ones is almost always the higher-return move at that point. The archive is where compounding lives, and a hundred good answers maintained beats three hundred published and abandoned. That is the whole argument for treating the hub as an owned search asset rather than a campaign.
Topics: publishing cadence · editorial calendar · content operations · consistency
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.