Research creates possibilities, not obligations
Keyword exports, user questions, AI suggestions, and competitor pages can all produce a long list of plausible topics. None of those lists knows what your current pages already cover, what evidence you own, or which work matters to the business. Treating discovery as a publishing queue skips the decision that gives the research value.
The useful output may be a stronger paragraph on an existing page. It may be a comparison table built from current primary sources. It may be a crawl that proves two pages overlap. Those are real outputs even though they do not increase the URL count.
The cost of saying yes too early
Every new page needs a distinct promise, internal links, evidence, maintenance, and later review. Weak pages also create ambiguity: people and search systems must decide which URL is the real owner of the topic. The operating burden accumulates quietly because publishing is visible while maintenance is not.
A conservative decision rule makes that cost explicit before the draft exists. If the team cannot name the owner, evidence, and lifecycle of the page, the idea is not ready.
Count decisions, not drafts
A healthier review tracks how many pages were extended, given a new section, merged, held for evidence, created, or left unchanged. That record shows whether research is improving the site rather than merely expanding it.
The goal is not less content as a doctrine. It is fewer pages that exist for no durable reason, and more attention on the pages and evidence that already carry a real job.