B2B SaaS SEO when the whole market is 400 queries

A gym management platform, a game telemetry tool and a veterinary practice suite share a problem: the entire buying market types perhaps 400 distinct queries a month, and around 40 of them deserve a page. Everything past that number is invented demand. We build the map from those queries to pages to closed-won revenue, and the first recommendation is usually to delete most of the blog. How we work: We start in your CRM, not in a keyword tool. We pull closed-won and closed-lost from HubSpot or Salesforce for the last four quarters, read first-call transcripts in Gong or Fathom, and extract the language buyers used before they knew your category name. That list gets checked against the Search Console query export through the API — the full 16 months, not the 1,000 rows the interface shows — and then against the results themselves, because in these categories the page one you are entering is usually a G2 category page, a Reddit thread and two competitors. A page that cannot beat those three has no business being written. Then the page inventory. Every URL gets one of four verdicts: keep and improve, merge, redirect, or delete. Deletion is a real answer and we use it. A 200-post blog with eleven URLs that have ever touched a buying query is a liability. It dilutes internal linking, it spends crawl on a small site, and it makes the product pages harder to reach. The rest is structural. Docs subdomains competing with marketing pages over the same query, and the decision about which one should win. Integration and comparison pages built against real partner surfaces — the Unity Asset Store listing, the Shopify app page, the Zapier directory — rather than from a template. SoftwareApplication and FAQPage markup where it earns a result and nowhere else. Pricing pages that answer the price question instead of routing to a form, because in a market this small the buyer who bounces off a form goes to a competitor and does not come back. What we do not do: We do not write your content. We specify it: the argument, the evidence required, who inside your company has to be interviewed, and what the page must beat. Your team or your writer executes and we review against the spec. We do not buy links or run outreach. We do not produce a publishing calendar and will not agree to a post count. We do not report sessions, impressions or average position as headline numbers. The report is pipeline touched and where it came from. Who this is not for: Self-serve products under about 200 euro a year, where the economics need volume content and we are the wrong shop. Teams that cannot ship a template change inside a sprint, since most of what matters here is engineering work and recommendations that sit in a backlog for two quarters do nothing. Founders who need the number to move in six weeks. Category-creating products with no existing demand at all: search is not your first channel, and we will say that on the call rather than sell you a retainer. What lands: A demand map tied to revenue, a URL-by-URL verdict sheet, a technical specification your engineers can build from without asking us questions, and briefs for the pages worth building. All of it as working documents you keep, not a deck. We work in English and Swedish. Most of this practice is Nordic companies selling into the US and DACH, which means we spend real time on the question of whether the .se site should exist at all.