Company News Scraper — Press, Funding & Announcements Feed

Company news for your accounts, as a clean feed: one deduplicated, typed row per story — funding, acquisitions, launches, partnerships, executive moves, layoffs — with every source that reported it.

If you sell to, invest in or keep an eye on a list of companies, you know the ritual. Before every call, every planning cycle, every Monday: search each company, wade through five outlets reprinting the same press release, skip the share-price noise, and try to remember whether you already saw that story last week. Your AI agent does the same thing, only faster and just as confused: it gets five near-identical articles, a stock ticker page and a K-drama star who happens to share the company's name.

We needed this for our own account agents, so we built it — and now we're sharing it. Give it company names or domains; it searches the news, reads each company's own newsroom and press releases, keeps only headlines that are really about that company, merges the same story from every outlet into one row, labels what kind of event it is, and remembers what it already told you so a scheduled run returns only what's new. No summaries, no AI rewriting: titles, dates, URLs and snippets come straight from the source, so your agent reasons over facts.

Who it's for

• Account-based sales & SDR teams — "what happened at my 50 target accounts this week?" before writing the next touch.
• AI agents and GTM automations — a raw, typed, MCP-callable news feed to poll on a schedule, with a stable id per story.
• Investors & analysts — "any funding, M&A or leadership changes at my portfolio companies?"
• Customer success & partnerships — spot expansions, layoffs, security incidents or new partners at your customers early.
• Market & competitor watchers — follow a topic ("SOC 2", "agentic commerce") or a set of competitors in one feed.

What you get

One row per news event (a story), newest first:

{
  "companyName": "HubSpot",
  "companyDomain": "hubspot.com",
  "eventType": "award",
  "eventTypes": ["award"],
  "title": "HubSpot named a Leader in the 2026 Gartner® Magic Quadrant™ for B2B Marketing Automation Platforms",
  "url": "https://www.hubspot.com/company-news/gartner-magic-quadrant-2026",
  "publisher": "HubSpot",
  "publishedAt": "2026-09-24T15:36:38.000Z",
  "snippet": "Marketing Hub recognized as a Leader for the sixth consecutive year",
  "snippetSource": "company-feed",
  "sources": [{ "url": "https://www.hubspot.com/company-news/gartner-magic-quadrant-2026", "publisher": "HubSpot", "publishedAt": "2026-09-24T15:36:38.000Z", "title": "HubSpot named a Leader in the 2026 Gartner® Magic Quadrant™ …" }],
  "sourceCount": 1,
  "isFirstParty": true,
  "attribution": "first-party",
  "isNew": false,
  "canonicalId": "cn_4f1c2a9b7d3e0a61",
  "firstSeenAt": "2026-09-26T08:12:40.000Z"
}

• One story = one row. The same announcement from TechCrunch, Yahoo Finance and three wire re-posts is one event with five sources — and one charge. Same-day share-price chatter about a company is one event, not twenty.
• Typed events — funding, acquisition, product, partnership, expansion, leadership, layoffs, earnings, legal, security, award, stock (share price, analyst ratings, insider trades) and other. Typed by transparent keyword rules, all matching types listed in eventTypes.
• The company's own voice — its newsroom, press feed or announcement blog is read too and merged with the coverage (isFirstParty: true).
• Only news that is really about the company. The headline has to name it. For names that are also everyday words (Stripe, Gong, Ramp, Notion) the headline needs extra evidence — "Stripe's", "Stripe launches", "at Stripe", a link to stripe.com — so football "stripes" and actors called Gong never reach you. Each row says how it was matched in attribution.
• Verbatim, never generated — snippet is the article's own description or lead paragraph (or the company's feed summary), copied as-is; snippetSource says which.
• New since last run — a stable canonicalId per story and a per-watch memory. Flip on Only stories that are new since my last run and a daily schedule becomes a clean feed of what changed; a story that gets picked up by more outlets later is not "new" again.

Input

Everything is optional — run it empty to get the latest company announcements worldwide.

• Companies — names (Datadog), domains (hubspot.com) or both (Gong (gong.io)). A domain is the most precise, and it unlocks the company's newsroom.
• Topics — with companies: narrow their news (stablecoin). Without companies: follow the topic itself (SOC 2, data center).
• Published within — 1–90 days (default 7). Event types — keep only the kinds you act on (leave out Stock & analyst chatter to keep business events only).
• Sources — newsroom (on), second news index (on), global news database (off, slower), open articles for the publisher URL and lead text (on).
• Delta & limits — only-new mode, Watch ID, caps per company and in total, and whether to include unconfirmed matches.

How to use it

1. In the Console: add your companies, press Start, open the News events view. Export to CSV/Excel/JSON or connect Google Sheets, Slack or your CRM through Apify integrations.
2. On a schedule: save the input as a Task with Only stories that are new since my last run on, and schedule it daily. Each run delivers only the stories you haven't seen.
3. From an AI agent (MCP): add the Actor through the Apify MCP server and ask "latest news on datadoghq.com and hubspot.com this week, funding and product launches only" — the agent gets typed JSON back.
4. From code: call it via the Apify API with {"companies": ["hubspot.com"], "daysBack": 7} and read the dataset.

And that's it.

Pricing

Pay per event: $0.005 per delivered news event (plus Apify's small start fee). Duplicate reports of the same story, filtered-out events, unconfirmed matches and empty runs are never charged.

A daily watch of 20 accounts typically delivers 20–60 new stories a day — about $0.10–$0.30 a day, and you pay once per story, not once per outlet that reprinted it.

Tip: turn on Only stories that are new since my last run for scheduled watches, and leave out Stock & analyst chatter if you only care about business events — you pay only for rows you keep.

Notes

• Sources are public news pages and the companies' own public newsrooms; no login, no API key, no personal data. Please use the data in line with the publishers' terms.
• The Actor delivers raw facts; it does not summarize or rewrite articles.
• Missing a source, a newsroom that isn't picked up, or a company that is matched wrongly? Open an issue in the Issues tab — we fix those quickly. Custom feeds and integrations: reach us through Apify.