Query-to-Page-Type Intent Map for Large Inventories

A query list is not a page plan. On the last six inventories we mapped, between 58 and 71 percent of the queries a client handed us resolved to a URL they already owned, and the right answer was an edit to that page rather than a new one. What we return is a decision per query: which URL owns it, which page type the results page is currently rewarding, and which queries we are telling you not to build for. What you get back: One sheet, one row per query, carrying the owning URL, the page type that wins the SERP today (category listing, editorial guide, comparison table, tool, forum thread, video), the action — own, edit, merge, ignore — and a sentence of reasoning we sign. A second tab holds the queries we removed from your list and why. The removals are the part clients argue with, so each one carries the date the SERP was read and the count of user-generated results in the top ten. How we decide what a query wants: We pull sixteen months of query-URL pairs through the Search Console API rather than the interface, so the export is not truncated at a thousand rows, and we join it against your internal site search log. For marketplaces the site search log is the better source. It carries the words your buyers use for a category before they learn your taxonomy, and it separates the two audiences sharing one domain: people who want to buy and people who want to list. Those two intents rarely belong on the same URL, and on the home and furniture marketplaces we have worked on they were merged onto one page about half the time. Then the results page itself. We read the top ten by page type, not by domain. If four or more results are forum threads or user reviews, an editorial page from a commercial domain does not enter that SERP, and we mark the query closed rather than let a writer spend a week on it. If the top ten splits between a category listing and a guide, the query holds two intents, and the map says which one your domain has a claim on and which one it does not. Where we ignore volume: Third-party volume estimates for Singapore and the wider region rest on samples too thin to act on, and they collapse queries across en-SG, en-MY and Chinese-language variants that behave as separate results pages with separate winners. We use them to sort rows, never to decide one. Priority comes from your own impression data, your site search counts, and revenue per session on the pages already ranking. Not included: We do not write pages, produce briefs, or specify a hub structure inside this engagement. Those are separate pieces of work and bundling them does not make the map better. We do not do keyword research in the sense of handing back a longer list than you gave us. And we do not deliver a spreadsheet of volumes and difficulty scores; if that is the artefact being asked for internally, buy it from a tool vendor for a few hundred dollars a year. Wrong fit: If your site has fewer than roughly two hundred indexable URLs, there is not enough behaviour in your own data to map from, and the hub planning engagement is the better purchase. If nobody on your side can change a page template within the quarter, the map will sit unread. And if the answer you need is that all six hundred queries are worth pursuing, we are the wrong practice. We will come back with a shorter list than the one you sent, and we will not soften it.