Pull Top Websites and Google Trends in one Trends MCP account, then join domain-plus-category rows against live search queries and google search indexes.
Live data as of 2026-08-27
A Top Websites vs Google Trends comparison on Trends MCP is two get_top_trends calls: type Top Websites and type Google Trends. From the August 27, 2026 pull, Google Trends stamped 2026-08-27T08:01:41 UTC and filled ranks 1 through 200, then count 0 at offset 200. Top Websites still stamped 2026-08-02T04:01:35 UTC, 25 days earlier, and kept returning 50 rows at offset 250, so 300 mixed-board rows were scanned without a close. Mixed-board rank stayed 1 through offset 100. Rank 2 began at offset 114 with netflix.com under arts-and-entertainment. Google Trends rank 1 was dolly parton. paypal.com led the mixed finance path. Zero exact titles overlapped. Free tier access is 100 requests per month.
From the Trends MCP pull of August 27, 2026, Top Websites mixed board was still on as_of 2026-08-02T04:01:35 UTC, Google Trends closed at 200 rows on as_of 2026-08-27T08:01:41 UTC, and a default limit 25 join misses every Google query after 25 plus every Top Websites domain after the first page.
| Rank | Top Websites mixed (domain / path) | Google Trends |
|---|---|---|
| 1 | paypal.com / finance | dolly parton |
| 1 | bitlabs.link / gambling/bingo | tim curry (Google rank 2) |
| 1 | impress.co.jp / marketing-and-advertising | nepal (Google rank 3) |
| 1 | famousbirthdays.com / holidays-and-seasonal-events | jed york (Google rank 4) |
| 1 | ancestry.com / ancestry-and-genealogy | full moon august 2026 (Google rank 5) |
| 1 | mercadolibre.com / auctions | bitcoin price (Google rank 6) |
| 1 | linktr.ee / online-marketing | target halloween costume controversy (Google rank 7) |
| 1 | canva.com / graphics-multimedia-and-web-design | nationwide indus foods chicken recall (Google rank 8) |
| 1 | youtube.com / tv-movies-and-streaming | social security cola 2027 announcement date (Google rank 9) |
| 1 | homedepot.com / home-and-garden | shelley fabares (Google rank 10) |
Top Websites mixed rank is not a global 1-to-N list on the first 100 rows. Every scanned row through offset 100 carried rank 1 because each line is the leader inside a category path. Offset 50 still opened with coupang.com, facebook.com, amazon.com, google.com, and spotify.com, all rank 1. Offset 100 switched to rank 2 at netflix.com under arts-and-entertainment. Offset 200 stayed on rank 2 through apple.com and temu.com, then rank 3 appeared later in that page at walgreens.com under health. Offset 250 was still count 50 on rank 3, including x.com under social-networks-and-online-communities. The Top Websites API page still describes a limit 25 scan from August 2, 2026 and says the rank field stays 1. This pull kept filling past 300 rows, and rank 2 started at offset 114.
Same-stamp Google Trends 1d against 2026-08-26T02:01:36 UTC returned count 11, led by jelly roll at rank 31, prev_rank 135, rank_change 104. Wikipedia Trending 1d on that stamp returned count 12, led by Buddy (2026 film) at rank 69, prev_rank 188, rank_change 119. Instagram sat on that Wikipedia climber list at rank 185, prev_rank 186, rank_change 1. instagram.com sat at Top Websites mixed rank 2 under social-networks-and-online-communities. Instagram was absent from Google Trends 200. App Store Top Free on the Google stamp listed ChatGPT at rank 1 and Google at rank 10. ChatGPT missed Top Websites 300 and Google Trends 200.
Top Websites sort rank_change window 1d returned count 0, with a note that no baseline sat within 1 day of 2026-08-01, that the feed updates roughly monthly, and that the client should retry window 30d. Window 3d pointed at 2026-07-30. Window 7d pointed at 2026-07-26. Window 14d pointed at 2026-07-19. All four short windows were empty. Window 30d resolved against 2026-07-02T04:01:31 UTC and filled 50 climbers at offsets 0, 50, and 100, so at least 150 mover rows. The top 30d climber was ltk.com at rank 9, prev_rank 47, rank_change 38, under heavy-industry-and-engineering/construction-and-maintenance. That within-category rank 9 is not the mixed-board leader. The mixed board listed procore.com as rank 1 on that construction path.
Two get_top_trends calls against the same Trends MCP key return Top Websites and Google Trends as separate ranked lists, and REST mode get_top_trends bills each type and each page as its own request, so paginating Top Websites through offset 250 plus Google Trends through offset 200 is ten requests before any get_growth follow-up.
{"mode": "get_top_trends", "type": "Top Websites", "limit": 50, "offset": 0}
{"mode": "get_top_trends", "type": "Google Trends", "limit": 50, "offset": 0}
MCP clients send the same fields on get_top_trends. Top Websites on this stamp filled 50 six times through offset 250. Google Trends filled 50 four times, then 0 at offset 200. Offset and limit behavior is on get_top_trends pagination. For the Google Trends sibling board, see Google Trends Trending Now. For climber fields on the search board, see Google Trends rank_change.
Some mixed-board rows omit the category path. google.com appeared once as [1, "google.com"] with no third field, and youtube.com appeared later as [2, "youtube.com"] the same way. A parser that expects rank, domain, path on every row will drop those lines.
Passing category on Top Websites is not a filter on the mixed August 2 board. It can select a different snapshot. On August 27, 2026, category finance stamped 2026-08-02T04:01:35 UTC, matching the mixed board, and returned paypal.com at rank 1, tradingview.com at 2, chase.com at 3. Category games on the same call pattern also stamped August 2, with twitch.tv at rank 1 and roblox.com at 2. Category news-and-media stamped 2026-07-02T04:01:31 UTC, a full month earlier, and returned yahoo.co.jp at rank 1, yahoo.com at 2, nytimes.com at 5, bbc.com at 7, cnn.com at 9. That July 2 stamp is the same clock as the mixed-board 30d rank_change baseline.
{"mode": "get_top_trends", "type": "Top Websites", "category": "finance", "limit": 25}
{"mode": "get_top_trends", "type": "Top Websites", "category": "news-and-media", "limit": 25}
Official category strings are lowercase kebab-case: news-and-media, finance, games, e-commerce-and-shopping. They are listed on data sources. Omitting category returns the mixed leader list. Do not assume every category string shares the mixed board's as_of_ts.
Exact string equality between a domain and a Google Trends query produced zero hits across 300 Top Websites rows and 200 Google queries. Brand follow-ups still resolve on get_growth with source google search.
| Brand or query | Board fact | google search index | 7D | Recent date |
|---|---|---|---|---|
| paypal.com | mixed finance rank 1 | PayPal 53.0, volume 23,700,000 | -7.02% from 57.0 | 2026-08-22 |
| youtube.com | mixed tv-movies rank 1 | YouTube 72.0, volume 1,510,000,000 | -7.69% from 78.0 | 2026-08-22 |
| canva.com | mixed graphics rank 1 | Canva 62.0 | +8.77% from 57.0 | 2026-08-22 |
| homedepot.com | mixed home-and-garden rank 1 | Home Depot 55.0, volume 31,200,000 | -9.84% from 61.0 | 2026-08-22 |
| temu.com | mixed marketplace rank 2 | Temu 59.0, volume 33,100,000 | -4.84% from 62.0 | 2026-08-22 |
| instagram.com | mixed social rank 2 | Instagram 64.0, volume 468,000,000 | -1.54% from 65.0 | 2026-08-15 |
| google docs | Google Trends rank 30 | 52.0 | +4.0% from 50.0 | 2026-08-22 |
| dolly parton | Google Trends rank 1 | 1.0 flat | 0.0% from 1.0 | 2026-08-22 |
| ChatGPT | App Store Top Free rank 1 | 80.0 | +8.11% from 74.0 | 2026-08-22 |
dolly parton 30D compared 1.0 against baseline 0.0, so that percent is omitted. Instagram's recent_date sat on 2026-08-15, a week behind the other google search follow-ups from this run.
False token joins that a lowercase contains() would accept: google.com against google docs at Google rank 30, usps.com against trump usps voting rule lawsuit at Google rank 42, savefrom.net against save america at Google rank 48. Temu: Shop Like a Billionaire sat at App Store Top Free rank 19 on the Google stamp, matching temu.com on the mixed board, and still missed Google Trends 200. Category leadership and a live search spike are different questions. Join them with a brand table, then keep the two as_of timestamps in the same row so a 25-day lag is visible.
FAQ