Search intent and the art of matching pages to queries
A page that perfectly answers the wrong question does not rank. That single sentence explains more stuck pages than any technical audit. Search engines are in the satisfaction business, and their first job for every query is deciding what the person actually wants. Pages that match that want get to compete. Pages that do not are simply not in the race, however good they are.
The classic four buckets, briefly
The traditional taxonomy sorts queries into informational (how does hreflang work), navigational (search console login), commercial investigation (best crm for small business), and transactional (buy standing desk). It is a useful vocabulary, but treating it as a filing system misses the point. Intent lives on a spectrum, mixes freely, and shifts over time. The buckets are training wheels; the real skill is reading individual SERPs.
The SERP is the answer key
Google spends billions learning what satisfies each query, and it publishes the conclusion openly: the current results page. Searching the target query and honestly reading what ranks answers most intent questions in thirty seconds.
If the results are product and category pages, the query is commercial and a blog post will struggle regardless of quality.
If the results are listicles and comparisons, searchers want options weighed, and a single-product pitch is mismatched.
If a featured snippet, People Also Ask boxes, or a video pack dominates, the appetite is informational and quick.
If the results mix formats, intent is mixed, and there is usually room for more than one type of page to win.
This is also where keyword difficulty scores mislead. A low-difficulty keyword is not an opportunity if every ranking page is a type the site cannot or will not build.
One page per intent, not per keyword
Keyword lists tempt people into creating one page per phrase, which manufactures near-duplicates that compete with each other. The healthier unit is the intent cluster: all phrasings that want the same thing share one page. A pricing question, a cost question, and a how-much question are one page. A comparison against competitor A and a comparison against competitor B are probably two, because the person deciding between different pairs wants different evidence.
The test is simple. Would one excellent page satisfy both queries completely? Same page. Would a searcher of one query feel the page was really about something else? Separate pages.
Diagnosing mismatch on existing pages
Intent mismatch has a recognizable signature. The page sits on the edge of page one or deep on page two for its target query and refuses to move. Impressions exist, clicks are weak, and engagement from the clicks that do come is short. Meanwhile some unrelated long-tail query quietly loves the page. That pattern usually means the content answers a different question from the one it targets.
The fix is rarely more words. It is alignment: reshape the page toward what the SERP demonstrates, or accept that the query belongs to a different page type and build that instead. Intent-first thinking feels slower than pumping out keyword pages, but it produces fewer, stronger pages that rank for entire families of queries. The search engine is telling everyone what it wants. The trick is to actually look.