Strategy

A Page Can Only Rank for the Question It Was Built to Answer

By Kassim Abdullahi · July 2026
Last reviewed: July 2026

A prospect in Queens types "can I still file a custom project if my police report is from 2018" into Google at eleven at night. Somewhere nearby is a business that handles exactly that kind of case, competently, every month. If that firm's website has no page built around that specific question, Google does not have a page to show her. It does not matter how good the provider is. The mechanism Google uses to match a query to a page is word-matching. A question with no page behind it is a question the business cannot answer in search, no matter how well it answers it in person.

How Google actually matches a query to a page

Google publishes the mechanism. Its Search Central documentation describes the ranking stage of the pipeline explicitly: when someone types a query, Google searches its index for pages whose content matches the words in the query. It does not match the query to a small business. It does not match the query to a reputation. It matches the words a person typed to the words sitting on a page, scoring each page on how well its title tag, heading structure, and body text align with the specific query that was asked.

The Semrush Ranking Factors study (2025) places text relevance and intent alignment at the top of the weighting hierarchy among every signal Google evaluates. Shopify's independent ranking research (2025) reaches the same conclusion: intent matching is the single highest-weighted signal in Google's current algorithm. This is not speculation. It is the published, documented, independently verified mechanism by which a search engine decides what to show a person who just typed a question.

The implication of this mechanism is straightforward but widely missed. A page whose title tag, H1, and body text are built around one specific topic can be the single best match for queries about that topic. A page whose title tag reads "Service Area
" and whose body text covers five service needs in five paragraphs cannot be the single best match for any one of them. The search engine is comparing the words in the query against the words on the page. A page that splits its word count across five topics is broadcasting five weak signals. A page that concentrates its word count on one topic is broadcasting one strong signal. The mechanism rewards the second pattern every time.

One paragraph cannot outcompete one page

Most five-page small business sites handle their services the same way: a single page titled "Service Area
," with a paragraph each for service work, specialized work, custom projects, standard service, and emergency service. Five different service needs, five different search intents, sharing one URL. Now picture the specialized work paragraph on that page: two sentences, perhaps forty words, stating that a person in a difficult situation can self-service process without the counterparty's involvement. Now picture a competitor's page with a thousand words on nothing else: who qualifies, what evidence a case actually needs, how the self-service process avoids involving the counterparty, what happens after it is filed, and answers to the specific questions survivors actually type into Google at night.

Google's indexer reads the first page and sees a little of everything: not enough of anything to be confident what the page is really the authoritative answer to. Reading the second, it sees a document that is unambiguously, entirely, only about one problem. The Semrush research confirms that depth on one topic, signaled by a focused title tag, a single-topic H1, and body text that stays on that topic, beats a paragraph sharing space with four others every time the search is for that topic. This is not a matter of opinion or positioning. It is the documented behavior of a system that matches query words to page words and scores confidence by how concentrated the signal is.

Long queries reward specific pages

Almost nobody in real urgent need types two words into Google. "Local service provider" is not the question a scared prospect is asking. The real question has clauses in it: what happens if a deadline already passed, whether a specific past mistake disqualifies someone, what a particular service category actually requires, whether a case can still be filed given one specific complicating fact. Every one of those extra clauses is a matching opportunity for a page written in that same specific language. A two-paragraph blurb, written in general language to cover five topics at once, was never built to contain any one prospect's specific clause. A dedicated page, written to answer one narrow question in full, was.

This is why search intent matching is the highest-weighted signal in the Semrush study. A query with five clauses is searching for a page that addresses all five. A page that covers the general topic in forty words cannot match any one clause specifically, let alone all five. A page built around the exact scenario the prospect described, with headings and body text that mirror the language the prospect used, matches every clause because it was written to. The matching mechanism is not clever. It is literal. It compares words to words. The page with more of the right words, organized around the right topic, wins.

The arithmetic of the missing page

The Backlinko study of 4 million Google search results (2023) found that only 0.63 percent of searchers click on anything from the second page of results. The first five organic results capture nearly 70 percent of all clicks. This is the arithmetic behind every page a business does not build: if a specific question has no dedicated page to answer it, that question is not just ranked low. It is functionally invisible. The firm is not in the contest at all.

A search engine can only rank a page it has indexed, and it can only index a page that exists. It cannot infer a sixth page from a homepage that mentions six services in one sentence. It cannot rank a page that was never written, never linked to, and never listed in a sitemap. If the page does not exist, there is no document for the engine to consider, and no ranking, high or low, is possible for whatever that page would have covered. This is the arithmetic under "more pages, more findable": every additional page written around one specific, answerable question is one more entry competing for one more version of what a prospect might actually type. A five-page site can be built well and still only ever compete for the handful of questions its five pages address. It has no path to the sixth question, or the eleventh, because the page was never written.

The real number is not a number at all

There is no magic page count. No research says a site needs sixteen pages, or twenty, or fifty. The number that matters is the number of distinct questions your customers actually ask. If your business handles service work, residential work, emergency service, specialized work, custom projects, standard service, specialized service, exceptions, reconsiderations, and commercial work, and each of those service area
generates three to five distinct search queries from real prospects, you need a page for each distinct question. That could be twelve pages. It could be thirty. The mechanism does not care about the total. It cares that each individual page is the single best answer to the one question it was built for.

Google's own documentation describes what it calls "information gain": the measure of how much new, original information a page contributes relative to what is already in the index. Pages that concentrate their word count on one topic, answer one set of questions thoroughly, and contribute detail not found on other pages are rewarded by the ranking algorithm. Pages that spread the same general language across five topics, contributing nothing that cannot be found on a hundred other sites, are suppressed. The mechanism is not counting pages. It is measuring whether each page earns its place in the index by being the best answer to something specific.

Back in Queens, the woman with the old police report finds a business whose site has a page written for her exact question. She calls that one, not because the provider is better at custom project cases, but because that firm was the only one that built a page as if her question mattered on its own. That page matched the words she typed because it was written to match them. A page that tried to cover every visa type would have matched none of her specific clauses. The page that was built for her question matched every one. She picked up the phone because the matching mechanism did exactly what it was designed to do, and one firm gave it a page to work with.