News Trending Detector 📈

Find out what's accelerating in the news — not just what's being published.

Most "news scrapers" hand you a firehose of articles and leave you to figure out what matters. This actor is different: it tracks the velocity of every topic across multiple sources and surfaces the ones that are spiking right now versus their own recent baseline.

---

Why velocity, not volume?

A topic with 50 mentions isn't news if it always has 50 mentions. A topic that jumped from 4 → 20 in the last hour is. This actor keeps a rolling baseline for every topic (persisted between runs) and flags:

Status	Meaning
emerging	Brand-new topic, never seen in previous runs
accelerating	Mention volume is significantly above its own baseline
baseline	First run only — establishing the reference (no velocity yet)
steady / cooling	Flat or fading — only returned as a fallback (see below)

Accelerating and emerging topics come first, so you get signal, not noise. When nothing is accelerating in a given run, the actor returns the current top topics (labelled steady/cooling) instead of an empty result — so a scheduled run always gives you a readable snapshot of the landscape.

⏱️ Important: velocity is measured across runs. The first run just builds the baseline. Schedule the actor to run on a recurring basis (e.g. every hour) — from the second run onward you get real accelerating / emerging signals. Apify Schedules make this one click.

---

Two modes

1. Watchlist mode — pass keywords_filter (e.g. openai, nvidia, bitcoin). Each topic is queried directly against Google News for a strong, focused volume signal, then tracked for acceleration. Perfect for "tell me when MY topics surge."

2. Discovery mode — leave keywords_filter empty. The actor aggregates all sources and surfaces whatever is trending globally.

---

Sources (all public, no API keys)

• Hacker News — tech / startup signal
• Google News — broad + per-keyword search
• Bloomberg — markets & business
• CNBC — business & tech
• BBC — general / world
• The Verge — tech / consumer
• ➕ your own custom RSS feeds

---

Use cases

• Traders / hedge funds — catch market-moving stories early
• Brand & PR monitoring — know the moment your brand or a competitor spikes
• Growth / VC scouts — spot emerging themes before they're obvious
• Newsletter & content creators — auto-sourced "what's hot" sections
• Researchers — study how stories propagate and accelerate

---

Input

Field	Type	Default	Description
keywords_filter	string	—	Comma-separated watchlist. Empty = global discovery.
include_hackernews	boolean	true	Include Hacker News.
include_rss_feeds	boolean	true	Include Google News, Bloomberg, CNBC, BBC, The Verge.
custom_rss_feeds	array	[]	Your own RSS feed URLs.
min_frequency	integer	3	Minimum mentions to qualify as a topic.
max_results	integer	100	Cap on trends returned per run.

Example

{
  "keywords_filter": "openai, nvidia, bitcoin",
  "include_hackernews": true,
  "include_rss_feeds": true,
  "min_frequency": 3
}

---

Output

Each dataset item is one trending topic:

{
  "topic": "OpenAI",
  "status": "accelerating",
  "current_mentions": 18,
  "baseline_mentions": 6.2,
  "velocity_score": 12.4,
  "acceleration": 11.8,
  "sources": ["gnews", "hackernews", "theverge"],
  "sample_articles": [
    { "title": "OpenAI raises mega round", "source": "gnews", "url": "https://…" }
  ]
}

• velocity_score — single ranking number (higher = accelerating harder / fresher)
• acceleration — current mentions minus baseline
• sample_articles — up to 5 example headlines behind the trend

---

Pricing

Pay per result: $0.002 per topic returned. A typical run returns a handful to a few dozen items, so most runs cost a few cents. No subscription, no rental — you pay only per item in your dataset.

---

Tips

• Run on a schedule (hourly is a good start) — that's what powers the velocity signal.
• Tighten min_frequency if you want only strong trends; loosen it to catch weak/early ones.
• In watchlist mode, fewer, well-chosen keywords give cleaner signal than a huge list.

---

Built for the international market. All data comes from public, freely available sources.