Google Trends Intelligence

Built by Skootle for decision-ready trend data.

Fast answer: what this Actor is for

Google Trends Intelligence turns one to five Google Trends queries into structured timeline, related-query, and summary records. It helps SEO, content, product, and research teams distinguish a real low-data result from a blocked source before they use a trend signal in a decision.

• Run it in the Apify UI for a bounded demand or content-research pull.
• Schedule it or call it by API for recurring trend monitoring.
• Send the dataset to a spreadsheet, webhook, or an automation that consumes typed records and source links.

TL;DR

Provide keywords, a geography, and a time window. The Actor returns TREND_POINT, RELATED_QUERY, and SUMMARY records with normalized fields, a source URL, retrieval time, and an agentMarkdown explanation. Google Trends values are relative indexes, not search-volume counts. The latest verified Cloud run before this documentation update, run ywvdKNw0LMgXfnfik, succeeded on 2026-09-10 using build xab6roYMTXB8o8PKS.

Try a small dataset first. If it helps, leave an honest review. Use the Issues tab for reproducible bugs.

What does Google Trends Intelligence do?

A manual Google Trends workflow requires opening charts, changing the comparison, copying points, and separately reading related queries. This Actor writes one dataset instead. Filter recordType to use the shape you need:

• TREND_POINT: one row for a keyword and time bucket, including interest, rolling average, and partial-bucket status.
• RELATED_QUERY: a ranked top or rising related search, including a numeric growth value when Google publishes one.
• SUMMARY: one decision record per keyword, including direction, momentum, volatility, coverage, and source outcome.

Every record includes a stable recordId, schemaVersion, sourceName, sourceQueryUrl, trendsUrl, retrievedAt, fieldCompletenessScore, and agentMarkdown.

Why use Google Trends data?

Google Trends is useful for the shape and timing of interest in an exact query, geography, and time window. It is not an absolute-search-volume source. A buyer can use the output to compare terms in one request, inspect related demand, create a trend chart, or provide a traceable input to an automation. Keep the query context fixed when comparing runs because changing the timeframe, geography, property, or comparison set can change the relative 0 to 100 index.

Who needs this?

• SEO and content teams choosing topics from recent search-interest and related-query signals.
• Product and brand teams monitoring a fixed set of demand terms over time.
• Researchers building a traceable trend dataset with source links and timestamps.
• Analysts who need to separate a genuine no-match from an unavailable source.
• AI-agent and automation workflows that need typed numeric fields plus a concise agentMarkdown summary.

How to use Google Trends Intelligence

1. Open the Input tab.
2. Add one to five keywords and choose a geo and timeframe.
3. For a first run, use one or two keywords, today 3-m, and maxRelatedQueries: 10.
4. Start the run, then export the dataset or schedule the same bounded input.
5. Filter SUMMARY for one decision row per keyword. Keep TREND_POINT rows for charts and RELATED_QUERY rows for query research.
6. Before acting on a result, check sourceStatus, zeroMatch, dataCoveragePct, isPartialBucket, and the included source link.

How much does Google Trends Intelligence cost?

This Actor uses pay-per-result pricing with an Actor start fee and a charge for emitted dataset records. The current price shown in the Store Pricing tab is authoritative. The local pricing configuration for this build declares a $0.01 start fee and $0.003 per dataset-item event. Actual totals depend on records emitted and retries. The Actor's configured default run options have a 600-second timeout; check the live Console before relying on any run cap or price display.

ROI example

A conservative first-run budget is one keyword, today 3-m, maxRelatedQueries: 10, with both result types enabled. That produces a bounded dataset you can evaluate before scheduling a wider comparison. The value decision is whether the returned source-backed signal changes a research or monitoring workflow, not a claimed conversion or labor result.

Source, permissions, and legal use

The Actor reads public Google Trends web endpoints. It does not require a Google account and does not access a logged-in, paid, or private Google surface. It uses a residential proxy configuration for reliable retrieval and fails loudly when it cannot read trustworthy source data. It is not affiliated with, endorsed by, or sponsored by Google.

Google Trends values are sampled relative indexes, not absolute volumes. Review your intended use, applicable law, and source terms before redistributing data. Do not treat a source block as a demand signal, and independently verify material findings.

Examples

Example 1: Compare two demand terms in the United States

{
  "keywords": ["artificial intelligence", "machine learning"],
  "geo": "US",
  "timeframe": "today 3-m",
  "maxRelatedQueries": 10
}

Use SUMMARY to compare the two trend directions under the same scale.

Example 2: Weekly content research for one term

{
  "keywords": ["cold plunge"],
  "geo": "US",
  "timeframe": "today 3-m",
  "maxRelatedQueries": 25,
  "includeInterestOverTime": true,
  "includeRelatedQueries": true
}

Use RELATED_QUERY rows for top and rising query research, then preserve trendsUrl with any downstream recommendation.

Example 3: Trend chart without related queries

{
  "keywords": ["electric vehicle"],
  "geo": "US-CA",
  "timeframe": "today 12-m",
  "includeInterestOverTime": true,
  "includeRelatedQueries": false
}

Use TREND_POINT rows for the chart and SUMMARY for the partial-bucket and direction fields.

Example 4: YouTube-search interest comparison

{
  "keywords": ["home gym", "pilates"],
  "geo": "GB",
  "timeframe": "today 3-m",
  "property": "youtube",
  "maxRelatedQueries": 10,
  "proxyCountryCode": "GB"
}

Keep the property, geography, timeframe, and comparison set unchanged when comparing this result with a later run.

Input parameters

Field	Type	Default	Description
keywords	array of strings	required	One to five terms, each 1 to 100 characters. Multiple terms share one comparison scale.
geo	string	US	Google Trends geography, such as US, GB, or US-CA. Empty means worldwide.
timeframe	enum	today 3-m	One of now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m, today 5-y, or all.
category	integer	0	Google Trends category ID, from 0 through 1500.
property	enum	empty	Search surface: empty for web, or images, news, froogle, or youtube.
maxRelatedQueries	integer	25	Maximum rows from each top and rising list per keyword, from 0 through 50.
includeInterestOverTime	boolean	true	Return timeline points and compute time-series metrics.
includeRelatedQueries	boolean	true	Return top and rising related-query records.
language	string	en-US	Google Trends interface locale.
proxyCountryCode	string	US	Two-letter residential proxy exit country.
maxRetriesPerRequest	integer	4	Retry attempts per upstream request, from 1 through 8.

Google Trends Intelligence output format

Field	Type	Description and buyer use
recordType	string	TREND_POINT, RELATED_QUERY, or SUMMARY, for safe downstream routing.
recordId	string	Stable record identity for deduplication.
schemaVersion	string	Schema version for downstream pinning.
keyword, geo, timeframe, property	string	Query context that must be retained with the result.
interestValue, dateIso, isPartialBucket	number, string, boolean	Timeline value and whether the bucket is still filling.
relatedQuery, listType, growthPct, isBreakout	mixed	Related-query result and Google-published growth context.
direction, momentumPct, volatility, dataCoveragePct	mixed	Summary fields for a trend decision, not absolute demand.
sourceStatus, zeroMatch, notes	mixed	Distinguishes a usable no-match from an unavailable source.
sourceQueryUrl, trendsUrl, retrievedAt	string	Provenance and the human-readable Google Trends chart.
agentMarkdown	string	A concise source-backed explanation for an agent or automation.

Sample record

This redacted TREND_POINT is from successful Cloud run uXDXfIcTY1DVyr1Hi on 2026-09-09, dataset rBbfPkUAQd0RqyFHD. URL query parameters are omitted only for readability.

{
  "recordType": "TREND_POINT",
  "recordId": "gtrends:point:artificial-intelligence:US:2026-06-09:1780963200",
  "schemaVersion": "1.0.0",
  "queryLabel": "artificial intelligence vs machine learning | US | today 3-m",
  "keyword": "artificial intelligence",
  "geo": "US",
  "timeframe": "today 3-m",
  "dateIso": "2026-06-09",
  "interestValue": 51,
  "isPartialBucket": false,
  "rollingAverage7": 51,
  "sourceStatus": "ok",
  "zeroMatch": false,
  "fieldCompletenessScore": 90,
  "sourceName": "Google Trends public web endpoints",
  "trendsUrl": "https://trends.google.com/trends/explore?q=artificial+intelligence&date=today+3-m&geo=US",
  "retrievedAt": "2026-09-09T10:35:01.702Z"
}

During the Actor run

The Actor gets a session, requests Google Trends widget data, normalizes the returned records, and writes a machine-readable RUN_REPORT plus the default dataset. A blocked or malformed upstream response is retried within maxRetriesPerRequest and then fails the run rather than emitting an empty dataset. A genuine low-volume no-match emits a SUMMARY with zeroMatch: true and sourceStatus: "ok". The trailing partial bucket is emitted for visibility but excluded from summary statistics.

FAQ

How is this different from a chart-only Google Trends workflow?

The Actor returns typed records, source URLs, retrieval timestamps, and a run-level source outcome so the result can be exported or consumed by an automation. It does not convert the relative Google Trends index into search volume.

Why does this cost what it costs?

The Store charges for the Actor start and emitted dataset records. Result volume grows with timeline buckets, related-query limits, and the number of keywords. Start with a bounded input and check the live Pricing tab before scaling a scheduled workflow.

How reliable is it for recurring work?

The latest observed Cloud evidence before this update includes successful runs on 2026-09-10 using build xab6roYMTXB8o8PKS. Reliability for a future source request is not guaranteed. The Actor records a source status and fails loudly instead of silently turning an unavailable source into zero demand.

Will source changes break it?

Google can change its public response behavior. If the Actor cannot produce trustworthy data, it fails the run with a classified source error rather than returning a plausible empty result. Recheck the run report and source links after an upstream failure.

Can I use it with an API, spreadsheet, or AI workflow?

Yes. Export the default Apify dataset or consume it through the Apify API. Use recordType to route the three record shapes and retain source fields with any derived output.

What is the cheapest safe way to try it?

Run one keyword with today 3-m, maxRelatedQueries: 10, and a fixed geography. Verify that the dataset has sourceStatus: "ok" before broadening the input or adding a schedule.

Maintenance status

• Last verified: 2026-09-11
• Verified behavior: local tests, typecheck, build, and offline smoke passed before this information-page update. The latest observed successful Cloud run was ywvdKNw0LMgXfnfik, completed 2026-09-10 with build xab6roYMTXB8o8PKS.
• Maintenance promise: no response-time or future-maintenance promise is made here. Use the Issues tab with a run ID and reproducible input for a defect report.

Why choose Google Trends Intelligence?

• Returns three typed record shapes for timeline, related-query, and summary use cases.
• Separates a genuine no-match from an unavailable source with zeroMatch and sourceStatus.
• Keeps source URLs and retrieval timestamps on every record for downstream verification.
• Use it for relative interest and query discovery, not as an absolute-search-volume source.

Other Skootle Actors you might want to check

• Website SEO Audit Intelligence to evaluate a site after a demand hypothesis is selected.
• Website Tech Stack Intelligence to add technology and integration context to website research.
• Federal Contract Intelligence to pair public-sector demand research with award and agency evidence.

Support and contact

File bugs or feature requests through the Actor's Issues tab. Include the run ID, input, and the field or behavior that looks wrong. Do not include private data or credentials in an issue.