{"slug": "what-42270-measured-monthly-searches-for-mcp-actually-contain", "title": "What 42,270 measured monthly searches for MCP actually contain", "summary": "A team building an MCP gateway measured 73 MCP-related search phrases using paid keyword data, finding they total 42,270 monthly searches, with the top three phrases (\"model context protocol,\" \"mcp protocol,\" \"a2a protocol\") accounting for 47.6% of that volume. Clustering the live result pages for 48 phrases produced 43 distinct clusters with only 4 merges, indicating that nearly every MCP query owns its own results page rather than collapsing into shared topical clusters. The team concludes that winnable demand sits in specific, low-difficulty phrases like \"mcp vs api\" and \"streamable http\" rather than the highest-volume terms.", "body_md": "We build an MCP gateway. That means we have to answer \"where is the demand for MCP tooling\" with\n\nnumbers instead of instinct — and when we finally measured it, the shape of the answer was not what\n\nthe topic's noise suggests. This is that measurement, with the method attached so anyone can repeat\n\nit and disagree with our reading.\n\nThe MCP ecosystem produces a lot of *content* and very little *demand data*. Every week brings another\n\n\"awesome MCP servers\" list, another protocol explainer, another launch post. What none of them carry is\n\nthe number that decides what to write next: how many people search for a given phrase in a month, and\n\nhow hard the results page for it is to enter.\n\nWe had a second reason to care. Our own pages were indexed, our sitemap was clean, and our Search\n\nConsole still showed almost no non-branded impressions. When that happens, the useful question is not\n\n\"is our prose good\" but \"does anyone search for this at all, and with what words\". A demand\n\nmeasurement answers both at once.\n\nThe pool comes from paid keyword data, not estimate tables scraped from a free tool:\n\nThe 73 phrases sum to **42,270 searches per month**. One phrasing carries more than a quarter of it:\n\n| Phrase | Monthly searches | Share of pool | \n|---|---|---|\n| model context protocol | 12,100 | 28.6% | \n| mcp protocol | 4,400 | 10.4% | \n| a2a protocol | 3,600 | 8.5% | \n| mcp tools | 2,400 | 5.7% | \n| agentic rag | 1,900 | 4.5% | \n| mcp vs api | 1,900 | 4.5% | \n| everything else (67 phrases) | 15,970 | 37.8% | \n\nTwo readings follow, and they point in opposite directions. The optimistic one: there is a real\n\naudience — 42,000 searches a month is not a hobbyist niche. The pessimistic one: **the top three phrases are 47.6% of it**, and those three are the \n\nDifficulty here is a 0–100 estimate of how hard the first page is to enter. Sort the pool by volume\n\nand a pattern appears — the big phrases are defended, and the *usable* ones sit in the middle:\n\n| Phrase | Monthly searches | Difficulty | \n|---|---|---|\n| anthropic mcp | 1,600 | 69 | \n| mcp protocol | 4,400 | 64 | \n| model context protocol | 12,100 | 58 | \n| mcp tools | 2,400 | 56 | \n| mcp spec | 1,300 | 32 | \n| mcp server list | 880 | 22 | \n| mcp vs api | 1,900 | 12 | \n| streamable http | 880 | 12 | \n\nThe two phrases with the lowest difficulty in the whole table are also two of the most specific:\n\n\"mcp vs api\" compares two architectures, and \"streamable http\" names one transport that MCP\n\nstandardised. Both are the kind of query a reader has *after* learning what MCP is, and both are\n\nanswerable by anyone who has actually shipped the thing rather than read the specification.\n\nThat asymmetry is the practical finding: **the demand that a new entrant can realistically win is not the demand that looks biggest.** Chasing the 12,100 phrase means competing with the protocol's\n\nThis was the surprise. We expected the 48 phrases to collapse into a handful of topical clusters, the\n\nway \"cheap flights\" and \"low cost flights\" belong together. Instead, clustering the live result pages\n\nturned 48 phrases into **43 clusters** — with only 4 merges and 11 adjacent pairs across the whole set.\n\nTranslation: for MCP, *almost every phrase owns its own result page*. \"mcp spec\" and \"mcp\n\nspecification\" are not the same query if the results pages disagree; \"mcp tools\" and \"mcp servers\" sit\n\nnext to each other without merging. The topic looks like one huge subject from the outside and\n\nbehaves like a mosaic of small ones from the inside.\n\nThis has a direct consequence for anyone planning content here, and it is the opposite of the usual\n\nadvice to build one comprehensive pillar page. A single page cannot be the canonical answer to 43\n\ndifferent result pages. What works is a **small centre per phrase family** — the protocol overview, the\n\ntransport comparison, the client-configuration walkthrough — each one narrow enough to actually match\n\nits result page, and each one linking to its neighbours so that a reader who lands on any of them can\n\nreach the rest. Our own overview of the protocol's actors, lifecycle and transports is written to that\n\nshape, rather than to a keyword count.\n\nThe last piece of the measurement is not about search demand at all; it is about what our own site\n\nsaw while we were publishing into this pool. In the four weeks ending 22 September 2026, our Search\n\nConsole recorded **54 impressions and 7 clicks** — and every one of them was a branded query. Not one\n\nof the non-branded phrases in the table above produced an impression during that window.\n\nThat is worth stating plainly, because it is the number that disciplines the rest: a pool of 42,270\n\nmonthly searches, fifty-odd published articles, and a non-branded impression count of zero. Demand\n\nbeing real does not mean demand is reachable; the pages existed before the audience did, and Google has\n\nto decide to serve them before any of the arithmetic in this article matters.\n\nThe same period's edge logs add a useful counterweight: of the page reads recorded at our CDN in one\n\nweek, 282 carried AI-crawler user agents and 144 carried search-crawler user agents. Treat that as a\n\nsignal rather than a measurement — a user agent is a claim, not an identity, and machine files like\n\n`robots.txt` inflate both numbers unless they are filtered out. The direction is still informative:\n\nagents are reading this material earlier and more often than search engines are.\n\nThree things, all of them reversible if the next snapshot disagrees:\n\nThe honest list, because a demand table invites more confidence than it earns:\n\nIf the useful summary is one paragraph long, it is this: for MCP the measurable demand is real but\n\nconcentrated in a handful of definitional queries, the entry price is lowest exactly where the\n\naudience is most specific, and the topic fragments into far more result pages than its vocabulary\n\nsuggests. Plan accordingly — a few narrow pages that answer a phrase completely, linked into a\n\nneighbourhood, will outperform one broad page that answers nothing precisely.\n\nThe pool, the clustering and the diffable snapshots stay where the content lives. If you want the\n\nstarting point rather than the summary, the protocol's own mechanics are documented in our\n\n[MCP overview](https://smartgate.network/industry/model-context-protocol-explained), and the\n\npublished-server side is collected in the\n\n[server list](https://smartgate.network/industry/mcp-server-list).\n\n*Disclosure: this measurement was assembled and written with AI assistance from our own paid keyword requests and Search Console exports; the numbers are quoted from those files rather than re-typed.*", "url": "https://wpnews.pro/news/what-42270-measured-monthly-searches-for-mcp-actually-contain", "canonical_source": "https://dev.to/smartgate/what-42270-measured-monthly-searches-for-mcp-actually-contain-3hmk", "published_at": "2026-09-24 11:17:51+00:00", "updated_at": "2026-09-24 11:31:48.800435+00:00", "lang": "en", "topics": ["agent-protocols", "generative-engine-optimization", "ai-search"], "entities": ["MCP", "Anthropic", "A2A"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/what-42270-measured-monthly-searches-for-mcp-actually-contain", "markdown": "https://wpnews.pro/news/what-42270-measured-monthly-searches-for-mcp-actually-contain.md", "text": "https://wpnews.pro/news/what-42270-measured-monthly-searches-for-mcp-actually-contain.txt", "jsonld": "https://wpnews.pro/news/what-42270-measured-monthly-searches-for-mcp-actually-contain.jsonld"}}