Your SEO Tool Reads Words. Gen Z Reads the Room.

Keyword tools sort queries into informational, commercial and transactional. That’s useful for filing a spreadsheet and less useful for deciding what to write. The label describes the syntax of a query. It says nothing about the person typing it, which is where the gap between search intent and keywords sits.

What your SEO tool means by intent, and what it can’t see

When a tool tags a query “commercial,” it has matched patterns: words like “best,” “review” or “vs” tend to appear when people are weighing options. That’s a reasonable shortcut, and it’s still a guess about wording.

What the tool can’t see is the situation. It doesn’t know whether the searcher is about to spend money, has just spent it, or is only curious. It also can’t see whether they’ve already been burned by a page like yours.

My view: treat the tool’s label as a first draft of intent. Intent is context plus a trust check, and both need a human to read them.

Same words, different situations: why one query can hold three different needs

Take “best budget laptop for students.” One person is a parent comparing models before term starts. Another is a student whose laptop just died and who needs something this week. A third is a content writer researching what students care about.

Three people at separate desks in a split-view scene, each with a different laptop situation: a parent with a notepad comparing printed spec sheets at a kitchen table, a student with a dead laptop with a black screen and a charger cable on a dorm bed, and a writer with sticky notes and a coffee at a

All three typed identical words, and a tool files them under one label. Each needs a different page: a comparison with clear tradeoffs, a short list with availability and a fast answer, or a deeper explainer. A page that tries to serve all three usually serves none well. Choose the situation you can answer best and make that obvious in the first screen.

The trust check: how younger searchers decide who deserves the click

Context tells you what someone needs. The trust check decides whether they believe you. Many younger searchers, by most accounts, don’t treat a top ranking as proof of anything. This is an observation about behaviour, not a statistic, and it will vary by person. They look at who is speaking and whether the claim holds up. They check whether the author is a person with visible experience or a faceless brand, and whether the advice matches what people say in forums, comments and videos.

For a marketer, that means a page has to pass two tests: it answers the query, and it gives the reader evidence to believe the answer. Named authors, stated limitations, disclosed affiliate links and specifics a reader can verify all help. Our guide to how Gen Z actually finds brands online goes further into where that checking happens.

Reading a results page like a cultural document

A results page is a record of what searchers have rewarded so far, so read it that way instead of scanning it for competitors. The mix of formats says a lot: if the top ten are forum threads and short videos, the searcher probably wants lived experience. If they’re glossy listicles from big publishers, the query may be one where people accept a general overview.

A top-down view of a desk where a printed search results page is laid out like an archaeologist's find, with highlighter marks, small colored tabs, and a magnifying glass resting on one entry, beside a phone showing a short video thread. Soft daylight, cream paper, yellow and teal accents.

Look at tone as well. Are the results cautious or confident? Do they sound like people or like brochures? A page that matches the intent but sounds like nothing in that results page can still stand out, provided it’s the more credible one.

A 20 minute intent audit you can run on one query without a tool

Pick one query and a timer. You need a notes document and the results page, nothing else.

A analog kitchen timer set to 20 minutes standing next to an open notebook with handwritten numbered notes and a pen, a mug beside it, on a plain wooden desk by a window. Bright morning light, calm neutral tones with a single red timer.

Minutes 1 to 3: list the situations. Write down three different people who could type this query and what each needs in the next hour.
Minutes 4 to 8: read the top ten. For each result, note the format, the author (named person, brand or unknown), the publish date and what the page assumes about the reader.
Minutes 9 to 13: run the trust check. For each result, ask who is saying this and whether you would cite it to a friend. Mark what earns trust: first-hand detail, sources, clear limits.
Minutes 14 to 17: check the edges. Look at autocomplete, related searches and the “People also ask” box. These show the follow-up questions real searchers have after the first answer.
Minutes 18 to 20: decide and rewrite. Choose the one situation your page serves best, then rewrite the opening paragraph so that person recognises themselves in the first two sentences.

The output is a decision about one page, which is more than a keyword export gives you.

Where AI search changes the job, and where it doesn’t

AI answers summarise before anyone clicks, so part of the trust check now happens inside the answer. What’s changing is how sources get chosen and shown, and that is still shifting, so I’d hold any claim about it loosely. If a page can’t be summarised cleanly, or says nothing a competitor doesn’t, it has less reason to be cited. We cover that in why your best-ranking page might be invisible to AI search.

The core job stays the same: work out which situation sits behind the query and give the reader a reason to believe you. AI search raises the cost of skipping that work, because a generic answer is the easiest one to summarise and the easiest one to replace.

Your next step

Pick one page of yours that ranks but doesn’t convert and run the intent audit from this piece on it. If you’d like a second pair of eyes on the result, get in touch about a content and search audit.