Scrape Google News for Brand Monitoring at Scale (2026)

Q: Does the actor return full article text or just metadata?

Just metadata: title, source, published_date, description (Google's snippet), url, language, country. For full article body text, follow up with a content extraction actor against the url field. Most brand-monitoring use cases only need the metadata to detect mentions, attribute to source, and route to alerting; full text is only necessary for sentiment-analysis or quote-extraction pipelines.

Q: Can I filter to recent news only?

Yes. The timeRange input accepts shorthand strings: 1h (last hour), 1d (last day), 7d (last week), 30d (last month). For breaking-news monitoring use 1h or 1d; for trend research use 7d or 30d. Leave empty to include all-time results, which is rarely what you want for monitoring (clutter from old articles dominates fresh signal).

Q: Does this work in non-English markets?

Yes. Pass language and country to localize. Google News supports every standard ISO language and country code — en/US (default), es/MX (Spanish-Mexico), de/DE (German-Germany), hi/IN (Hindi-India), and so on. Same query in different language-country pairs returns substantially different result sets, so for global brand monitoring run separate queries per market.

Published April 27, 2026 · 1534 words · For ops

Thirdwatch's Google News Scraper returns Google News articles by keyword at $0.0015 per article — title, source, published_date, description, url, language, country. Built for PR teams monitoring brand mentions, competitive-intelligence functions watching peer coverage, content teams finding trending stories, investor-research desks following news flow on tickers, and academics studying news coverage longitudinally.

Why scrape Google News for brand monitoring

Google News is the largest news aggregator in the world. According to Google's News disclosures, the platform indexes 50,000+ news sources across 70+ languages, processed in real-time as publishers push articles. For PR teams, competitive-intelligence functions, and investor-research desks, Google News is the canonical aggregator for "what was published about X in the last week". The blocker for systematic access: Google News retired its public API in 2011, leaving structured-data scrapers as the only path.

The job-to-be-done is structured. A PR team monitors 20 brand keywords plus 5 executive names hourly to catch every mention. A competitive-intelligence function watches 30 competitor names daily to detect product launches, leadership changes, and crisis events. A content team finds trending stories per topic to inform editorial calendars. An investor desk tracks 50 portfolio company names plus their key competitors. All reduce to keyword + language + country + timeRange returning structured article rows.

How does this compare to the alternatives?

Three options for getting Google News data into a brand-monitoring pipeline:

Approach	Cost per 1,000 articles	Reliability	Setup time	Maintenance
Manual Google News browsing	Effectively unbounded analyst time	Low (human filtering errors)	Continuous	Doesn't scale
Paid media monitoring SaaS (Meltwater, Cision)	$5K–$50K/year per brand	High, includes sentiment + reporting	Weeks–months to onboard	Vendor lock-in
Thirdwatch Google News Scraper	$1.50 ($0.0015 × 1,000)	Production-tested, multi-language	5 minutes	Thirdwatch tracks Google News changes

Paid media monitoring SaaS is the gold standard for enterprise PR but priced for full marketing-comms departments. The Google News Scraper actor page gives you the data layer at pay-per-result pricing — most brand-monitoring teams build their own on top of it for 1-2% of the SaaS cost.

How to scrape Google News for brand monitoring in 4 steps

Step 1: How do I authenticate against Apify?

Sign in at apify.com (free tier, no credit card), open Settings → Integrations, and copy your personal API token. Every example below assumes the token is in APIFY_TOKEN:

export APIFY_TOKEN="apify_api_xxxxxxxxxxxxxxxx"

Step 2: How do I pull mentions for a brand watchlist?

Pass brand keywords as queries, set timeRange to your monitoring window, choose language and country.

import os, requests, pandas as pd

ACTOR = "thirdwatch~google-news-scraper"
TOKEN = os.environ["APIFY_TOKEN"]

BRAND_QUERIES = ["thirdwatch", '"thirdwatch.dev"',
                 "apify scraping", '"apify actors"',
                 "web scraping platform"]

resp = requests.post(
    f"https://api.apify.com/v2/acts/{ACTOR}/run-sync-get-dataset-items",
    params={"token": TOKEN},
    json={
        "queries": BRAND_QUERIES,
        "maxResults": 50,
        "language": "en",
        "country": "US",
        "timeRange": "7d",
    },
    timeout=600,
)
df = pd.DataFrame(resp.json())
print(f"{len(df)} mentions across {df.source.nunique()} sources")

5 queries × 50 results = up to 250 mentions, costing $0.38.

Step 3: How do I dedupe across syndicated coverage?

Major news stories run on the wire — Reuters, AP, Bloomberg, AFP — and appear under many publications with near-identical headlines. Dedupe with fuzzy title matching to surface unique stories.

from difflib import SequenceMatcher

def title_norm(t):
    return " ".join(str(t).lower().split())[:80]

df["title_norm"] = df.title.apply(title_norm)

# Greedy near-dup grouping
unique_stories = []
for title in df.title_norm.tolist():
    if any(SequenceMatcher(None, title, ut).ratio() > 0.8 for ut in unique_stories):
        continue
    unique_stories.append(title)

df["story_group"] = df.title_norm.apply(
    lambda t: next((s for s in unique_stories
                    if SequenceMatcher(None, t, s).ratio() > 0.8), t)
)
unique = df.drop_duplicates(subset=["story_group"]).copy()
print(f"{len(df)} mentions → {len(unique)} unique stories")

The unique rows are the canonical stories; the original df retains source-level granularity for reach analysis.

Step 4: How do I forward newly-found mentions to Slack and PR routing?

Persist seen URLs and forward only new mentions to a brand-monitoring Slack channel.

import json, pathlib, requests as r

snapshot = pathlib.Path("news-seen.json")
seen = set(json.loads(snapshot.read_text())) if snapshot.exists() else set()

new_mentions = unique[~unique.url.isin(seen)]
for _, row in new_mentions.iterrows():
    r.post("https://hooks.slack.com/services/.../...",
           json={"text": (f":newspaper: *{row.source}* — {row.published_date}\n"
                          f"*{row.title}*\n_{row.description[:200]}_\n{row.url}")},
           timeout=10)

snapshot.write_text(json.dumps(list(seen | set(unique.url))))
print(f"{len(new_mentions)} new mentions forwarded")

Schedule the actor on Apify's scheduler at hourly cadence (0 * * * *) and the loop runs unattended for weeks at the typical brand-watchlist scale.

Sample output

A single article record looks like this. Five rows of this shape weigh ~2 KB.

{
  "title": "Tech stocks rally as AI spending surges",
  "source": "Reuters",
  "published_date": "2026-04-09",
  "description": "Major tech companies reported increased AI infrastructure spending...",
  "url": "https://www.reuters.com/technology/...",
  "language": "en",
  "country": "US"
}

url is the canonical natural key for snapshot dedup. source is the publication name as Google News attributes it (Reuters, AP, Bloomberg, etc.) — useful for source-quality filtering (treating Tier 1 outlets differently from blogs and aggregators). published_date is the article's publication date as Google News indexes it; for time-zone-precise monitoring, use your scrape timestamp as a backup signal because Google News dates are sometimes day-precision only.

Common pitfalls

Three things go wrong in production brand-monitoring pipelines on Google News. Search-query precision — bare brand names get false-positive matches when the brand is a common word ("apple"-the-fruit articles in an Apple-the-company watchlist); always use exact-match quotes ("thirdwatch") or unique multi-word phrases. Localization tradeoffs — country: "US" returns US-focused coverage of your brand even when the brand is mentioned in non-US press; for global coverage run the same query across each target-market language-country pair separately. Time-range edge cases — timeRange: "1h" works at sparse-topic queries but returns very few results; for non-breaking-news brand monitoring, 1d or 6h is the practical floor.

Thirdwatch's actor returns language, country, and published_date on every record so multi-market filtering and time-series analysis are clean. The pure-HTTP architecture means a 250-article pull completes in under three minutes and costs $0.38 — small enough to run hourly even on a 20-keyword brand watchlist. Pair Google News with our Twitter Scraper for social mentions and Reddit Scraper for forum discussion to build a complete listening pipeline. A fourth subtle issue worth flagging: Google News occasionally surfaces older articles when the term has limited recent coverage, even with timeRange set to 1d — always cross-check the published_date in your downstream filter rather than trusting timeRange alone for time-sensitive alerting. A fifth pattern unique to brand monitoring: trademark-style keyword variants (Thirdwatch vs ThirdWatch vs Third Watch) sometimes return non-overlapping result sets on Google News; build a synonyms list and run all variants in the same batch.

Related use cases

Frequently asked questions

How much does it cost to monitor a brand on Google News?

Thirdwatch's Google News Scraper charges $0.0015 per article on the FREE tier and drops to $0.00085 at GOLD volume. A 10-keyword brand-monitoring sweep at 50 articles each costs $0.75 per refresh. Hourly cadence over 10 brand keywords costs ~$540/month at FREE pricing — well below typical media-monitoring SaaS subscriptions like Meltwater or Cision.

Does the actor return full article text or just metadata?

Just metadata: title, source, published_date, description (Google's snippet), url, language, country. For full article body text, follow up with a content extraction actor against the url field. Most brand-monitoring use cases only need the metadata to detect mentions, attribute to source, and route to alerting; full text is only necessary for sentiment-analysis or quote-extraction pipelines.

Can I filter to recent news only?

Yes. The timeRange input accepts shorthand strings: 1h (last hour), 1d (last day), 7d (last week), 30d (last month). For breaking-news monitoring use 1h or 1d; for trend research use 7d or 30d. Leave empty to include all-time results, which is rarely what you want for monitoring (clutter from old articles dominates fresh signal).

Does this work in non-English markets?

Yes. Pass language and country to localize. Google News supports every standard ISO language and country code — en/US (default), es/MX (Spanish-Mexico), de/DE (German-Germany), hi/IN (Hindi-India), and so on. Same query in different language-country pairs returns substantially different result sets, so for global brand monitoring run separate queries per market.

How do I deduplicate articles across publications?

Google News surfaces the same story from multiple publications when major outlets cover it (Reuters, AP, Bloomberg all running the same wire story). The actor returns each as a separate row keyed by url. For deduplicated coverage analysis, group by title (with fuzzy matching for minor headline variations) before counting. Keep all rows in your raw data; dedup is a downstream analytics step.

How does this compare to paid media monitoring SaaS?

Meltwater, Cision, Critical Mention, and similar services charge $5K-$50K/year per brand for media monitoring with sentiment analysis, influencer tracking, and reporting dashboards. Thirdwatch's Google News Scraper is the data layer only — no sentiment, no dashboard. Build your own monitoring on top at 1-2% of the SaaS cost; pair with our Twitter Scraper and Reddit Scraper for full social listening.

Run the Google News Scraper on Apify Store — pay-per-article, free to try, no credit card to test.