A search label leaves most of the design undecided. Knowing that a question is informational does not tell me which answer belongs first, how much context to provide, or where the person should go after reading it.
For this site, I make those calls in a page brief. It lives in the page map that connects the site. The entry for my search psychology page is a useful example because the finished page also displays the brief in its “Who this page is for” block. The planning decision and the work can be checked against each other.
The question I gave the page
My psychology page has to let a fellow SEO professional judge whether I understand the people using a page. That is more specific than making a page about “user intent.” It tells me what the examples need to demonstrate.
The recorded question is: “Do you actually know who is searching, and can you build the page they need?”
I need to show how a person's situation changes my choices. A definition of search intent would leave that question unresolved. I already have a guide to the types of search intent; this page needs to demonstrate my judgment in practice.
What goes in an entry
Every page on this site gets the same short entry before it's built: who it's for, where they arrive from, the question they bring, what the first screen has to answer, the one next step and what the page leaves out. The psychology page shows its own entry in its “Who this page is for” block, so I won't repeat it here. What's worth showing is what the entry does once it exists.
Two of those fields are decisions, not findings. Where a reader arrives from is a route I'm designing for, not a measured source of visits. The question is the job I've chosen for the page, not an interview result. Keeping that straight lets me make a clear call without claiming to know more about the audience than I do.
What the entry catches
The entry earns its keep when it disagrees with the page. The build reads every entry and fails a page that ships with no next step, with two, or with one that goes somewhere the entry doesn't say. It also fails a page nothing points to.
On October 30, 2026, that last rule failed two posts I had just finished. Both had entries and both read well, but no other page named either one as a parent, sibling or next step, so nobody following the site could reach them except through a list. The fix wasn't more writing. Each became the sibling of the older post that hands its reader off to it, and the check passed.
The same night I caught a worse mistake by reading an entry back. I had started one new post's entry by copying three fields from another post's entry, and for a few minutes the brief described a different page than the one I was writing. Nothing in the page looked wrong. The entry did, the moment I read it next to the draft.
What the brief becomes on the page
On the psychology page, the entry's first-screen line is why the page opens on “Who is searching, and what do they need at that moment?” and puts the premise before the examples. Its next-step label is why the page ends on a whole site built for one kind of fan: after seeing single choices, the reader gets to inspect a site built around one audience. The link continues the thought the page just resolved, instead of sending someone to a directory to work out the connection.
What an entry leaves out matters as much. A new panel or a second button needs a reason to belong in that sequence; being useful somewhere on the site doesn't give it a place on this page. I make the same call when deciding whether a subject needs a new URL at all, which how I find topic-cluster opportunities covers.
What I can claim from the result
The result here is a page whose opening, examples and ending can be compared with its recorded purpose. I can inspect that correspondence directly. I have not established that this brief caused more visits, better understanding or more case-study clicks.
A reader task would answer a different question: after using the page, can someone explain a choice I made for a particular audience and find the case that develops it? That is a check to run, rather than a result I can write into the story now.
How I use the entry before asking for a draft
A brief gives me a concrete way to direct work and judge what comes back. I can ask an AI to compare an outline with the recorded question, identify a section that does not help answer it, or point out an ending that leads somewhere unrelated. I still decide whether the interpretation and the page make sense.
For another page, I start by finishing the entry before asking for more content. Then I compare the opening and ending with what I wrote down. If they describe different jobs, I resolve that difference before adding another section.
The useful output is a page I can account for: why it starts there, why each example belongs, and why the next link is the next thought.
Written for this site and first published October 30, 2026. Every number in it comes from my own exports and is dated where it appears.