Google Trends Scraper: Compare Unlimited Keywords

Google Trends Scraper: compare unlimited keywords on one scale

Get Google Trends data for any number of keywords, all on one comparable 0-100 scale. Google's own tool compares only 5 terms at a time, and every request is rescaled so its own peak is 100, which means numbers from two different comparisons can't be compared. This Actor fixes that automatically: it links the batches through a shared anchor keyword and rescales everything onto one scale. Paste 50 keywords and see which ones really get more searches.

For every keyword you get:

• 📈 Interest over time (hourly to monthly points, 2004 to today)
• 🗺️ Interest by region: country, state, metro area (DMA) or city
• 🔎 Related queries: top and rising, with Breakout flags
• 🧩 Related topics: top and rising (optional; returned when Google provides them, which it often doesn't right now)
• 📊 A trend summary: average, latest, peak and peak date, change over the period, year-over-year change, momentum, breakout flag and seasonal peak month
• 📄 One CSV with every timeline (dates as rows, keywords as columns), ready for Excel or Google Sheets

There's also a Trending Now mode: today's rising searches for any country, with approximate search volume, growth %, start time, category and related searches. That's usually 300-700 trends per country per day, not the 10-20 the RSS feed gives you.

Why this Actor

	This Actor	Typical Google Trends scrapers
More than 5 keywords on one scale	✅ automatic	❌ separate scales, or "do it yourself"
Speed	7-9 keywords in ~10-20 s; 100 keywords in ~60-80 s	minutes per keyword with browser-based scrapers
Approach	Direct HTTP to the Trends data endpoints, no browser	Often a headless browser that waits for page widgets
Failed keywords	Not charged, and listed in the run summary	Often charged, or the whole run fails
Trend summary per keyword	✅	❌ raw data only
YouTube, Shopping, News and Image search	✅	sometimes
Proxy cost	included in the price	sometimes billed on top

How the stitching works

With more than 5 keywords, the Actor:

1. Asks Google for the first 5 keywords and picks the most popular one as the anchor (or uses the anchorKeyword you set).
2. Asks for the remaining keywords in groups of 4, plus the anchor each time.
3. Rescales each group so the anchor matches its value in the first group, then rescales everything so the overall peak is 100.

In our tests, stitched values matched Google's own direct comparisons to within 0-3.5%. The small differences come from Google rounding to whole numbers.

Google rounds inside each request, so a small keyword compared with a giant one ("excel formulas" next to "chatgpt") comes back as zero. The Actor spots these cases and asks again, this time next to an already-measured keyword of similar size, then rescales through it. In a 100-keyword test across 5 countries, that cut zero rows by up to 100% and removed every low-precision row, for about 5-8% more requests. Each row includes a stitch object with the batch number, scale factor and a quality grade (high / medium / low). If an anchor is tiny compared with a much bigger keyword, the row gets a warning. Set a mid-sized anchorKeyword to fix it.

Input example

{
    "searchTerms": ["air fryer", "instant pot", "slow cooker", "rice cooker", "pressure cooker", "toaster oven", "sous vide"],
    "geo": "US",
    "timeRange": "today 5-y",
    "searchType": "web",
    "compareMode": "together",
    "includeRegions": true,
    "regionResolution": "REGION",
    "includeRelatedQueries": true
}

Trending Now:

{ "mode": "trendingNow", "geo": "GB", "trendingHours": "24", "trendingCategories": ["Sports"], "maxItems": 100 }

Output example (one keyword; arrays trimmed to 1-2 items)

A real row from the input example above:

{
    "keyword": "rice cooker",
    "geo": "US",
    "timeRange": "today 5-y",
    "searchType": "web",
    "category": 0,
    "scale": "shared",
    "comparedWith": [
        "air fryer",
        "instant pot",
        "slow cooker",
        "pressure cooker",
        "toaster oven",
        "sous vide"
    ],
    "average": 5.98,
    "latest": 7,
    "peak": 13,
    "peakDate": "2026-06-21",
    "low": 4,
    "changePct": 76.4,
    "yoyChangePct": 0,
    "momentumPct": -13.8,
    "trendDirection": "rising",
    "isBreakout": false,
    "peakMonth": "Jan",
    "topRelatedQuery": "rice in rice cooker",
    "topRisingQuery": "best rice cooker 2023",
    "points": 262,
    "timeline": [
        {
            "date": "2026-10-18",
            "value": 7,
            "isPartial": false
        },
        {
            "date": "2026-10-25",
            "value": 6,
            "isPartial": true
        }
    ],
    "regions": [
        {
            "geoCode": "US-HI",
            "geoName": "Hawaii",
            "value": 100
        },
        {
            "geoCode": "US-WY",
            "geoName": "Wyoming",
            "value": 77
        }
    ],
    "relatedQueriesTop": [
        {
            "query": "rice in rice cooker",
            "value": 100,
            "formattedValue": "100",
            "isBreakout": false,
            "link": "https://trends.google.com/trends/explore?q=rice+in+rice+cooker&date=today+5-y&geo=US"
        }
    ],
    "relatedQueriesRising": [
        {
            "query": "best rice cooker 2023",
            "value": 12900,
            "formattedValue": "Breakout",
            "isBreakout": true,
            "link": "https://trends.google.com/trends/explore?q=best+rice+cooker+2023&date=today+5-y&geo=US"
        }
    ],
    "stitch": {
        "batch": 0,
        "anchor": "air fryer",
        "scaleFactor": 1,
        "anchorMeanInBatch": 57.1,
        "quality": "high",
        "isAnchor": false,
        "localMean": 5.98
    },
    "exploreUrl": "https://trends.google.com/trends/explore?q=rice+cooker&date=today+5-y&geo=US",
    "warnings": [],
    "scrapedAt": "2026-10-25T04:22:43Z"
}

Trending Now row (GB):

{
    "rank": 2,
    "title": "chelsea vs spurs",
    "geo": "GB",
    "searchVolume": 200000,
    "increasePct": 1000,
    "startedAt": "2026-10-24T13:30:00Z",
    "endedAt": null,
    "isActive": true,
    "categories": ["Sports"],
    "relatedQueries": ["chelsea vs spurs", "chelsea f.c. vs tottenham standings", "chelsea", "cole palmer"],
    "exploreUrl": "https://trends.google.com/trends/explore?q=chelsea+vs+spurs&geo=GB&date=now%207-d"
}

The Keyword overview table view shows the summary columns. TIMELINE.csv in the run's key-value store has every timeline side by side, and OUTPUT lists the status of each keyword.

What can you use Google Trends data for?

• SEO and content planning: rank a keyword list by real search interest, spot rising queries before the competition, and time content to the seasonal peak month.
• Product research and e-commerce: compare products on Google Shopping search, find which states or cities want them most, and catch breakout products early.
• YouTube strategy: compare video topics by YouTube search interest.
• Market and brand tracking: track your brand against competitors on one scale. Schedule the Actor weekly and watch momentum.
• News and social teams: pull Trending Now every few hours for content ideas.
• AI agents and automations: clean flat JSON, a summary row per keyword, and simple inputs. Works through the Apify API, MCP, Make, n8n and Zapier.

Pricing

You pay per result: one keyword report (everything above for one keyword) or one trending search. Proxy and compute are included. Keywords that fail aren't charged. If you set a spending limit on the run, the Actor stops before going over it and lists the keywords it skipped.

Tips

• Values are relative, not search counts. Google Trends shows relative interest (0-100), not absolute volume. Trending Now's searchVolume is Google's own approximate figure.
• Compare like with like. In together mode every value compares across keywords in the same run. To compare against a later run, keep the same anchor keyword.
• Ambiguous words: set a category (for example 71 = Food & Drink) to separate "apple" the fruit from Apple the company.
• Low-volume terms may return all zeros. That's Google saying "not enough data", not an error, and the row carries a warning.
• Few regions for small keywords? Like trends.google.com, regions with too few searches are left out by default, because their values are noisy. Turn on includeLowVolumeRegions to get every region.

FAQ

Is it legal to scrape Google Trends? The Actor reads public, aggregated, anonymized data that anyone can see on trends.google.com. No personal data is collected. Check that your use complies with Google's terms and your local laws.

Why can numbers differ slightly from the website? Google samples its data, so two identical requests made hours apart can differ by a point or two. Stitched values also carry Google's whole-number rounding.

How many keywords can I send? As many as you like. Large lists run in parallel; raise maxConcurrency (up to 10) for speed.

Something not working? Open an issue on the Issues tab with your run link. We read every one.