Pass source google search for a weekly Google Search interest series, or source amazon for a monthly shopper-phrase series with independent volume. The live boards use different type strings.
Live data as of 2026-09-27
A Google Search vs Amazon search comparison on Trends MCP is two keyword sources plus two live boards. Source google search on get_growth or get_time_series returns a weekly 0-100 Google Search interest series, with estimated volume on some keywords. Source amazon returns a monthly 0-100 shopper-phrase series with independent volume. On September 27, 2026, air fryer closed 51.0 on Google Search (recent_date 2026-09-19, 261 weekly points from 2021-09-25, estimated recent_volume 3610000.0) and 54.4 on Amazon (recent_date 2026-08-31, 50 monthly points from 2022-07-31, recent_volume 5882739.0). 7D through 2W collapsed on Amazon. Type Google Trends and type Amazon Best Sellers Top Rated are separate get_top_trends feeds. Mixing those type strings into source is a 400.
The September 27, 2026 pull is the split that still matters next month: Google Search is a five-year weekly strip, while Amazon shopper demand is about four years of month-end rows. get_time_series for air fryer on source google search returned 261 weekly points from 2021-09-25 (value 31.0, volume 2193599) through 2026-09-19 (value 51.0, volume 3608824). The series peak was 100.0 on 2026-05-09 at volume 7076125. Every row carried a volume key and a source field. Dates stamped 00:00:00+00:00.
get_time_series for air fryer on source amazon returned 50 monthly points from 2022-07-31 through 2026-08-31. The field name on those rows was datatype amazon, not source. Opening row: value 67.1, volume 7261799. Close: value 54.4, volume 5882739. Peak index 100.0 fell on 2025-11-30 at volume 10823067. Dates printed YYYY-MM-DD with no timezone.
Google Search air fryer 12M was +2.0% from 50.0 on 2025-09-20. Amazon 12M was -8.57% from 59.5 on 2025-08-31, with volume_growth -8.67% from 6441264.0 to 5882739.0. Same phrase, opposite year. Google 7D fell 13.56% from 59.0 on 2026-09-12 while Amazon 7D, 1W, 14D, and 2W each returned error collapsed_range. Amazon 30D matched 1M, +5.63% from 51.5, volume_growth +5.46%. Google 1W matched 7D, 2W matched 14D (-15.0% from 60.0), and 1M matched 30D (-13.56% from 59.0 on 2026-08-22).
| Keyword | Google close | Google 12M | Google points | Amazon close | Amazon 12M | Amazon points |
|---|---|---|---|---|---|---|
| air fryer | 51.0 on 2026-09-19 | +2.0% from 50.0 | 261 | 54.4 on 2026-08-31 | -8.57% from 59.5 | 50 |
| crocs | 66.0 on 2026-09-26 | +6.45% from 62.0 | 261 | 38.3 on 2026-08-31 | -27.46% from 52.8 | 50 |
| running shoes | 31.0 on 2026-09-26 | +24.0% from 25.0 | 261 | 57.1 on 2026-08-31 | -25.75% from 76.9 | 50 |
| nike | 56.0 on 2026-09-26 | -1.75% from 57.0 | 261 | 47.0 on 2026-08-31 | -27.47% from 64.8 | 50 |
air fryer on Google Search closed a week earlier than nike, crocs, and running shoes on the same pull. nike recent_date printed 2026-09-26 with no timezone. air fryer printed 2026-09-19 00:00:00+00:00. Quote recent_date per keyword. Do not assume one Saturday for the whole Google strip.
chatgpt shows a different scale, not just a different sign. Google Search closed 63.0 on 2026-09-26 across 261 points, with 12M -33.68% from 95.0 and no volume field. Amazon closed 15.2 on 2026-08-31 at volume 10529.0, across 45 monthly points from 2022-12-31, with 12M -60.42% from 38.4. The Amazon peak was 100.0 on 2023-10-31 at volume 69119. Web curiosity and shopper volume are not interchangeable units.
Two source strings, one shopper phrase. REST uses mode get_growth. MCP uses the get_growth tool. The Search keyword page is Google search data. The Amazon keyword page is Amazon search trends. Window defaults live on the get_growth contract page.
{"mode": "get_growth", "source": "google search", "keyword": "air fryer", "percent_growth": ["7D", "1W", "14D", "30D", "1M", "3M", "12M", "YTD"]}
{"mode": "get_growth", "source": "amazon", "keyword": "air fryer", "percent_growth": ["7D", "1W", "14D", "30D", "1M", "3M", "12M", "YTD"]}
{"mode": "get_growth", "source": "google search, amazon", "keyword": "air fryer", "percent_growth": ["7D", "30D", "12M"]}
{"mode": "get_growth", "source": "google search", "keyword": "air fryer"}
{"mode": "get_growth", "source": "amazon", "keyword": "air fryer"}
The last two calls omitted percent_growth. Google Search still used period 12M, recent_value 51.0, baseline_value 50.0, growth 2.0. Amazon still used period 12M, recent_value 54.4, baseline_value 59.5, growth -8.57. That matches the 12M row inside each multi-window block.
The comma-separated air fryer call returned sources_requested ["google search", "amazon"], sources_processed 2, sources_successful 2, plus trend_analysis.uptrend_ratio 1/2. That envelope is not two stacked single-source payloads. Read source_results for the per-source windows. metadata.total_data_points was 261 on Google Search and 50 on Amazon.
get_time_series rejected the same comma list. HTTP 400 invalid_source listed 16 keyword sources and did not list all. Source Google Search and source Amazon case-folded to 200 and matched the lowercase rows.
A third commerce calendar, weekly shopping intent with no volume key, lives on Google Shopping vs Amazon search. Mixing Search and Amazon first keeps one weekly index next to one monthly volume series.
Google Search volume is derived from the 0-100 index when the pipeline prints it. Amazon volume is a separate measurement. On air fryer, Google get_growth set volume_available true, volume_estimated true, recent_volume 3610000.0 (the series close was 3608824), and volume_growth null with reason volume is derived from the trend value for this source, not an independent measurement. Amazon set volume_available true, recent_volume 5882739.0 matching the series close, and volume_growth -8.67% on 12M next to index growth -8.57%.
Volume is not guaranteed on Google Search. chatgpt and running shoes Google windows on this pull carried no volume_available field. nike Google 12M printed estimated recent_volume 21800000.0. crocs printed 6900000.0. air fryer printed 3610000.0. Check the field. Do not fill it in.
The top-level volume_weighted_growth block mixes those two definitions, and it can mix windows. The 7D, 30D, 12M comma call printed volume_weighted_growth -1.77 with weight_coverage 2/2 sources and total_volume 9492739.0. Google Search took 38.03% of the weight at estimated volume 3610000.0 and growth -13.56 (the 7D index). Amazon took 61.97% at 5882739.0 and growth 5.46 (the 30D volume_growth, because 7D had collapsed). That is estimated weekly Google interest blended with independent monthly Amazon volume, on two different lookbacks. Quote the per-source rows. Do not treat -1.77 as a shared week.
Amazon 3M on air fryer was +18.26% from 46.0 on 2026-05-31 (volume_growth +18.22%). Google Search 3M was -7.27% from 55.0 on 2026-06-20. YTD on Google was -22.73% from 66.0 on 2026-01-03. YTD on Amazon was -33.82% from 82.2 on 2025-12-31 (volume_growth -33.88%). Amazon YTD baselines on the December month-end, not the first Google Saturday of the year.
Source strings and live-board type strings are not interchangeable. Type google search on get_top_trends returned HTTP 200, echoed type Google Trends, and put latin pop at rank 1 with as_of_ts 2026-09-27T06:01:56.589117+00:00. Omitting limit on type Google Trends returned limit 25, count 25, offset 0, same stamp. Source Google Trends on get_growth returned HTTP 400 invalid_source, echoed google trends in lowercase, and listed 16 keyword sources plus all.
Type amazon on get_top_trends returned HTTP 400 invalid_request and listed 24 valid feed names. Source Amazon Best Sellers Top Rated on get_growth returned HTTP 400 invalid_source, echoed amazon best sellers top rated in lowercase, and listed the same 16 sources plus all. Pass google search or amazon to growth and series tools. Pass the exact feed label to get_top_trends.
Omitting limit on type Amazon Best Sellers Top Rated returned limit 25, count 25, offset 0, and as_of_ts 2026-09-21T04:01:39.189390+00:00. Rank 1 was an Amazon Basics 20-pack AA listing. Rank 3 was Crocs Unisex-Adult Classic Clog, Clogs for Women and Men. That board stamp is six days behind the Google Trends stamp from the same morning. Rank is not a search-volume sort. The pairing of shopper phrases against that board lives on Amazon search vs Amazon Best Sellers.
An ASIN is not a shopper phrase. B08GC6PL3D on source amazon returned HTTP 404 no_data. The full Crocs listing title from rank 3 returned the same 404. The short keyword crocs closed 38.3 on 2026-08-31 at volume 6697760.0, with Google Search flat at 66.0 on 2026-09-26 for 7D. Truncate the listing, then attach the series in code.
get_growth bills one request per source plus keyword, including every window inside percent_growth. A two-source comma string is two units. The Free plan covers 100 requests per month. Starter is $19 for 1,000 requests, Pro is $49 for 5,000, and Business is $199 for 25,000.
FAQ