Google Trends Scraper (Reliable, Bulk)

Get Google Trends data for many keywords × many countries in one run: interest over time, interest by region, and top + rising related queries. Clean JSON, CSV or Excel output that is ready for charts, spreadsheets, dashboards or AI agents.

Built to finish the job. Google limits Trends traffic hard. This Actor gives every query its own session and proxy exit. When Google answers "too many requests", it waits a little and retries on a fresh session. In our test runs (450 queries across 9 countries), 449 succeeded.

Answers "is it rising or dying?" at a glance. Every result has summary columns: trend (rising / falling / flat), percentOfPeak, peakDate and topRisingQueries. Paste in 200 product names and sort by trend; you don't need to open a single chart.

keyword	trend	percentOfPeak	peakDate	topRisingQueries
labubu	falling	3	2025-07-13	real labubu, macaron labubu, …
dubai chocolate	falling	10	2025-03-23	kataifi, knafeh, pistachio chocolate bar, …
ozempic	falling	48	2024-05-26	mounjaro, ozempic vs mounjaro, zepbound, …
matcha	flat	64	2026-08-02	ryze matcha, matcha suri, …

(Worldwide, today 5-y, run on 6 Nov 2026.)

You only pay for real data.

• Queries that fail are never charged.
• Keywords with too little search volume are not charged either. They are listed in the run's NO_DATA record, so you can see which ones were skipped.

What you can do with it

• SEO & content planning: see which topics are growing, and spot seasonality before you publish.
• E-commerce & product research: compare demand for products across countries and find rising queries.
• Market & investment research: track interest in brands, tickers and technologies over time.
• Ad campaign timing: put budget into the regions and weeks with peak interest.
• AI agents & automations: one call returns structured data. Use it with Make, Zapier, n8n, the Apify API or Apify MCP.

Input

Field	What it does	Example
keywords	Search terms. Each one is fetched separately, with its own 0–100 scale.	["air fryer", "slow cooker"]
geos	Country or region codes. Leave empty for worldwide. Every keyword is fetched for every geo.	["US", "GB", "DE", "US-CA"]
timeframe	now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m, today 5-y, all, or a custom range YYYY-MM-DD YYYY-MM-DD	today 12-m
category	Google Trends category ID (0 = all categories)	7 (Finance)
property	Which Google search to measure: web, image, news, Google Shopping or YouTube	web
sections	Which data to fetch. Leave empty to get everything.	["interestOverTime"]
maxConcurrency	Number of queries run in parallel (1–10)	10

Example:

{
  "keywords": ["bitcoin", "ethereum"],
  "geos": ["US", "GB"],
  "timeframe": "today 12-m"
}

That is 2 keywords × 2 geos = 4 results.

Output

You get one dataset item per keyword × geo (the values below are illustrative):

{
  "keyword": "bitcoin",
  "geo": "US-CA",
  "timeframe": "2024-01-01 2024-03-31",
  "category": 7,
  "property": "web",
  "hasData": true,
  "peakDate": "2024-03-04",
  "latestDate": "2024-03-31",
  "latestValue": 61,
  "percentOfPeak": 61,
  "trend": "falling",
  "trendChangePercent": -18,
  "topRisingQueries": ["ibit stock"],
  "interestOverTime": [
    { "date": "2024-01-01", "datetime": "2024-01-01T00:00:00.000Z", "timestamp": 1704067200, "value": 11, "isPartial": false },
    { "date": "2024-01-02", "datetime": "2024-01-02T00:00:00.000Z", "timestamp": 1704153600, "value": 25, "isPartial": false }
  ],
  "interestByRegion": [
    { "geoCode": "855", "geoName": "Santa Barbara-Santa Maria-San Luis Obispo CA", "value": 100 },
    { "geoCode": "807", "geoName": "San Francisco-Oakland-San Jose CA", "value": 97 }
  ],
  "relatedQueries": {
    "top": [{ "query": "bitcoin price", "value": 100, "formattedValue": "100" }],
    "rising": [{ "query": "ibit stock", "value": 42600, "formattedValue": "Breakout" }]
  },
  "scrapedAt": "2026-11-03T21:33:12.201Z"
}

What the fields mean:

• Summary columns (only when interestOverTime is fetched). They use complete periods only; the current, unfinished one is skipped.
  • peakDate, latestDate, latestValue: when interest peaked, and the latest complete value.
  • percentOfPeak: latest value as a % of the peak. 3 means interest is at 3% of its high.
  • trend / trendChangePercent: average of the last 12 points vs the 12 before them (weeks for today 5-y, hours for now 7-d). More than +10% = rising, less than −10% = falling, otherwise flat. Needs at least 24 points, otherwise null.
  • topRisingQueries: up to 5 fastest-growing related searches.
• interestOverTime:
  • Each value is relative search interest, from 0 to 100.
  • 100 is the peak for that keyword in that geo and timeframe.
  • The time step depends on the timeframe: minutes, hours, days, weeks or months.
  • Use datetime (UTC) for hourly and minute data.
  • isPartial: true marks the current, incomplete period.
• interestByRegion: subregions (states, metro areas) ranked 0–100 by share of searches.
• relatedQueries.top: the most popular related searches.
• relatedQueries.rising: the fastest-growing related searches. "Breakout" means growth of more than 5000%.

Pricing

Pay per event:

• Per result (one keyword × geo with data): $0.0015, which is $1.50 per 1,000 results.
• Per run start: a small fixed fee.
• No charge for failed queries or for keywords Google has no data for.

You can cap spending with the Max cost per run option in Apify. The Actor stops cleanly when the cap is reached.

FAQ

Why are values 0–100 and not search volumes? Google Trends publishes only relative interest, never absolute volumes. Each keyword is fetched on its own, so each has its own 0–100 scale. To compare keywords on one shared scale, compare them in Google Trends directly.

Which geo codes can I use?

• ISO country codes such as US, DE or JP.
• Subregion codes such as US-CA, GB-ENG or SK-KI.
• An empty list means worldwide.
• An invalid code fails right away, is not charged, and the log tells you which one it was.

Why is one of my keywords missing from the results? The term has too little search volume in that geo and timeframe, so Google returns only noise: fewer than 3 non-zero points, with no regional data and no related queries. It is left out of the dataset, is not charged, and appears in the run's No-data keywords output (NO_DATA record) together with whatever Google did return.

Are related topics supported? No. Google currently returns empty related-topics lists to automated clients. Related queries work normally.

Is the data identical to what I see on trends.google.com? It comes from the same backend endpoints the website uses. Google samples Trends data, though, so the website itself shows slightly different numbers when the same query is repeated.

Disclaimer

This Actor is not affiliated with or endorsed by Google. It collects aggregated, anonymous search-interest statistics and no personal data. You are responsible for using the data in line with the laws and terms that apply to you.