Method
Motivate a New Topic from a Gap: The RankAI Method
A gap tells you something is missing. It does not tell you why your site should fill it. RankAI's five-field brief turns a hunch into a documented reason before anyone writes a word.
Before you open a new page, decide whether it should exist and why. Noticing that something is missing is not enough: you need to document the reason that page has to exist. At RankAI we propose a five-field brief, filled in before writing, that turns a gap into a checkable argument rather than a hunch. The brief is a RankAI working proposal, not product behavior and not a universal standard. If a field cannot be filled in against a checkable criterion, the page is not yet justified: the gap is real, but the argument is not.
A gap is not an argument
An information gap says something is missing. It does not say why RankAI should be the one to fill it, or at what angle. The risk is publishing pages that answer an internal curiosity rather than a real question. The brief separates the two moments: first you describe the gap, then you argue for the page.
The five-field method
- Observed gap. Checkable criterion: the gap is described in one sentence that names a concrete question and the point where it currently goes unanswered. An example that passes: “someone searching for how AI visibility is measured finds explanations of what visibility is, but nothing on how to build a measurement.” An example that fails: “the page is useful.”
- Audience and decision. Checkable criterion: state who reads the page and what decision they make after reading it. If the decision cannot be named, the field stays incomplete.
- Boundary with existing pages. Checkable criterion: list the closest RankAI pages and write in one sentence what this page adds that they do not cover. If the sentence is not distinguishable, the new page duplicates.
- Proof of non-redundancy. Checkable criterion: state the question the page answers and compare it with the titles of existing pages; if the question is already contained in a title, the field fails.
- Closing signal. Checkable criterion: describe the observation that, if it were missing, would make you close the page. An example that passes: “if we cannot point to an observable measurement, the page has no reason to exist.” An example that fails: “if it is not interesting, we close it.”
One completed brief, one incomplete
Take the search phrase “how AI visibility is measured” as an illustrative case. A complete brief describes the gap (visibility is discussed without saying how it is measured), names the reader and the decision (someone who has to set up a measurement), draws the boundary against existing pages on visibility, states the question, and sets an observable closing signal. An incomplete brief stops at the phrase “measuring visibility”: the gap is named, but audience, decision, boundary and closing signal are missing, so the page is not yet justified.
What this piece adds
The distinctive point of the method is not the fifth field but the link between the third and the fourth: the boundary with existing pages and the proof of non-redundancy have to be filled in together, because a page can look new against a single nearby page and turn out to duplicate the whole index. RankAI's working proposal is to write the boundary as a sentence that names at least two existing pages and the difference from both. If the sentence needs more than two lines to be clear, the gap is probably too broad and should be split into two topics. That makes the criterion falsifiable: anyone can read the sentence and check whether the new page adds something or repeats an angle that already exists. The brief does not replace writing, but it stops a page from being born without a documented reason.
A checkable limit
The brief does not prove the page will work: it only documents the reason it exists. If, after publication, the closing signal cannot be observed, reopen the brief and reconsider the page. For context on the limits of observation, see the in-depth page.