Skip to main content
Autobound logo

Market Intelligence

News Event Signals

Published Updated

Know when a company raises money, launches a product, or changes leaders.

39 news event subtypes covering funding rounds, acquisitions, product launches, executive hires, partnerships, and market expansions, for public and private companies.

Read the API docs
Companies With a Record
305K+
Event Subtypes
39
Signals / Month
~86K
Languages
9+
Delivery
REST API, GCS Push, Generate Insights API, Flat File
Format
JSONL or Parquet
Bucket
gs://autobound-news-v4/

What Are News Event Signals?

Daily refresh

News signals turn company news into structured events: funding, acquisitions, launches, partnerships, expansions and executive moves. They cover public and private companies, so they reach far beyond SEC filings.

We read company news all day, every day, in 9+ languages. A model pulls out each event, ties it to the right company domain and quotes the sentence it came from. When many outlets report one event, you get one signal, from the most authoritative source.

Before a record ships, it passes noise gates (about 40% of raw events are removed), company resolution, three dedup layers with a 60-day lookback and a strict contract check. The section How we build this signal, below, shows each step.

Example subtypes

Product launchlaunches

From a real record

Mastercardmastercard.comSep 27, 2026

Mastercard launched its Mastercard Move money movement solution in Jordan through a partnership with Network International.

Source sentence, verbatim
“Mastercard's General Manager for Saudi Arabia, Bahrain, Jordan, and the Levant, Saud Swar, said the launch of Mastercard Move in Jordan in partnership with Network International would support financial innovation and economic growth, while enhancing financial inclusion and the efficiency...”
Partner
Network International
Product
Mastercard Move
Location
Jordan
Confidence
0.90

Source: petra.gov.jo

From a news story to your next move

What we detect, what you get, and what you do with it. Pick the path that fits your team.

Sales, marketing and RevOps: Slack, CRM fields and plays.

  1. 0103

    Companies make news every day

    We read news coverage, find the company in each story, label the event and pull out its details: the product, amount, people and place.

    1. techcrunch.com
    techcrunch.comJun 18, 2026

    Wiz launches AI-powered cloud detection and response product

    Wiz today announced the general availability of its AI-driven detection and response product for multi-cloud environments.

    Cloud security company Wiz announced the general availability of Wiz Defend, an AI-powered cloud detection and response offering that correlates runtime signals across AWS, Azure, and GCP...

    Extracted
    Wizwiz.iocompanyProduct launchsignal_subtypeWiz DefendproductJun 18, 2026published_atAlready happenedplanning0.91confidence

  2. 0203

    You get one clean record

    Each record names the company and its domain, labels what happened, and links back to the source.

    2. Signal record

    Wiz

    wiz.ioComputer & Network Security

    Product launchlaunches

    Wiz launches AI-powered cloud detection and response product

    “Wiz today announced the general availability of its AI-driven detection and response product for multi-cloud environments.”
    Product
    Wiz Defend
    Published
    Jun 18, 2026
    Status
    Already happened
    Confidence
    0.91
    Detected
    Jun 19, 2026
    Source
    techcrunch.com

    signal_id 1cd1f9c7…

  3. 0303

    Act on it where you work

    Send it to Slack, write it to your CRM, or give it to an AI agent.

    When
    signal_subtype = launches on a target account
    Do
    Outreach that ties your product to the launch
    3. Your team acts
    or your own code
    #account-signals

    Signal alert from your workflow

    Wiz· Product launch

    Wiz launches AI-powered cloud detection and response product

    Play: Outreach that ties your product to the launch

    Start the playOpen the source

    Illustration. The record fields are real.

Try it on your own accounts

1,000 free credits. Query news in the Signal API or from Claude.

Get a free API keyUse it in Claude via MCP

How we build this signal

How we build each news signal

Every article passes six stages before it becomes a signal. Each task goes to the model best suited to it, and code checks wrap every model so the output cannot drift.

Feeds polled
150+
Raw events removed as noise
~40%
Dedup lookback
60 days
Refresh
Daily
  1. 01

    Collection and translation

    We poll 150+ press wires, publisher feeds and Google News queries around the clock. Broken text is repaired, and a model translates non-English articles, so an event reads the same in any language. A new source waits for a clean spot-check.

  2. 02

    Extraction

    A model reads each article into a fixed schema: event type, companies, date, amount and one verbatim sentence as evidence. It cannot add fields. Events under 0.8 extraction confidence are dropped.

  3. 03

    Relevance and noise gates

    A second model checks that each event is real, central to the article and useful to B2B buyers. About 40% stop here: recalls, sports, politics, obituaries, law-firm solicitations. A separate check blocks roundups that name many companies in passing. Every drop is logged with a reason.

  4. 04

    Company resolution

    Each company is resolved to its domain: the article's links and boilerplate first, then an 82-million-company reference graph, then web search. Direction is checked too, such as who raised and who invested. One article can yield several signals, each on the right company.

  5. 05

    Source selection and deduplication

    Articles on one event are clustered, and one source is kept: the issuing newsroom or wire, else the most authoritative outlet, never an aggregator. Three dedup layers follow: a shipped ID never ships again, a fact match catches re-reports within 21 days, and a model checks 60 days of the company's same-type events.

  6. 06

    Validation and delivery

    Every record must pass a strict contract check, or the file does not upload. A day under 500 records ships nothing and pages an engineer, so a broken run never looks like a quiet day. Rule-based dedup that would drop over 15% of a day rolls back. Each delivery logs what was dropped and why.

News Event Signal Schema

News signals include event classification, source attribution, and sales context enrichment. Each signal is linked to resolved company and contact entities.

Request, response and price →Free list built from this feed →
{
  "signal_id": "1cd1f9c7-468e-434d-bfac-c95fe9a32b65",
  "signal_type": "news",
  "signal_subtype": "launches",
  "association": "company",
  "detected_at": "2026-06-19T09:15:24Z",
  "company": {
    "name": "Wiz",
    "domain": "wiz.io",
    "industries": ["Computer & Network Security"]
  },
  "data": {
    "signal_name": "Launches",
    "summary": "Wiz launches AI-powered cloud detection and response product for multi-cloud environments.",
    "category": "launches",
    "found_at": "2026-06-18T14:10:24Z",
    "confidence": 0.91,
    "article_sentence": "Wiz today announced the general availability of its AI-driven detection and response product for multi-cloud environments.",
    "planning": false,
    "product": "Wiz Defend",
    "body": "Cloud security company Wiz announced the general availability of Wiz Defend, an AI-powered cloud detection and response offering that correlates runtime signals across AWS, Azure, and GCP...",
    "url": "https://techcrunch.com/2026/06/18/wiz-launches-cloud-detection-response",
    "published_at": "2026-06-18T14:10:24Z",
    "title": "Wiz launches AI-powered cloud detection and response product"
  }
}

Bulk data

Need this signal for 10,000+ companies?

Get a free sample file or a bulk feed (API, GCS push, flat file). The sample is real News records, as a CSV.

Talk to sales

Customer story

RocketReach logo

How RocketReach builds on our event signals

RocketReach used to track three event types (funding, M&A and executive changes) from a hand-maintained list of RSS feeds. Today Autobound is the primary source behind its event signals, delivered daily and loaded straight into Person and Company Search. Users pick a signal type and a time range to find the people and companies behind each event.

Read the RocketReach case study →
buying triggers in search, from funding to hiring
48
from first GCS delivery to live in production
About 1 month
“RocketReach data gives our users the who. Integrating signals data gives our users the when.”
Jamie GullbrandJamie GullbrandGroup Product Manager, RocketReach

Measured coverage

What's inside News Event Signals

A dated snapshot counted in our production database, not a live count. Minimums, not ceilings.

Coverage for all 35 signal types →
50,048
Companies with fresh data per month
3-month average, records through Jul 24, 2026
86,320
Records created per month
3-month average, records through Jul 24, 2026

Top subtypes

39 total
Launches227K · 22%
Attends event178K · 18%
Partners with122K · 12%
Hires93K · 9.2%
Recognized as45K · 4.4%
Is developing35K · 3.5%
Receives award35K · 3.5%
Invests into assets29K · 2.8%

Geography

all 199 countries →
🇺🇸United States65K · 21%
🇬🇧United Kingdom24K · 7.9%
🇮🇳India10K · 3.3%
🇨🇦Canada8.7K · 2.8%
🇦🇺Australia7.0K · 2.3%
🇩🇪Germany4.2K · 1.4%
🇫🇷France3.2K · 1.0%
CA 24KNY 14KTX 10KFL 8.1KMA 5.8KIL 5.8K

Top industries

all industries →
Software Development7.4K · 2.4%
Financial Services6.8K · 2.2%
Non-Profit Organizations6.3K · 2.1%
Hospitals and Health Care5.9K · 1.9%
IT Services and IT Consulting5.1K · 1.7%
Leasing Real Estate4.7K · 1.5%
Retail3.9K · 1.3%

Join keys

fill rate
Domain99%
LinkedIn URL26%
Company name99%
Records measured 1,013,260+Companies reached 305,915+

Company pages

See this data on real companies

Each company page shows the latest rows from this data for free, with the date and a link to the source.

See for yourself

Check out sample data: real News Event records with the same fields the API returns.

Read the API docs
“Autobound's triggers keep us current with news and job openings, making personalization effortless.”
Oleksandr Demianyk, Demand Generation Manager, TechMagic

Frequently asked questions

What types of news events do you track?
We track 39 event types including funding rounds, acquisitions, product launches, executive hires, partnerships, market expansions, office openings, IPO filings, layoffs, rebranding, awards, and regulatory actions. Each event is extracted from news sources and classified with confidence scoring.
How quickly are news events detected?
News event signals refresh daily. Most significant events, including funding rounds, executive appointments, and acquisitions, reach you within days of the public announcement. Coverage includes public and private companies, from news in 9+ languages.
How do news signals compare to 8-K filing signals?
8-K signals are sourced from SEC regulatory filings, so they're limited to public US companies but carry legal certainty. News event signals cover a much broader universe (public and private companies worldwide) but are sourced from news articles rather than regulatory filings.
Can I filter news events by event type or company size?
Yes. The API supports filtering by event subtype, company domain, industry, employee count, and date range. This lets you build targeted workflows, for example every company in your territory that raises a Series B or later.
How do you remove duplicate stories?
Three layers. An exact-identity check makes sure no signal we have delivered is delivered again. A deterministic layer catches re-reports of the same event by matching its facts. A final model compares each event with the same company's events of the same type from the prior 60 days and decides: the same event, or a new one?
When many outlets report one event, which source do you use?
One. We keep the single most authoritative source: the earliest confirmed report, the most complete one, and the one closest to the primary source. The other reports of the same event do not become extra signals.
Will I get the same event twice?
No. Every signal we deliver is tracked, and the exact-identity check stops it from shipping again, even when the story is reported again days later.
How do you know which company a story is about?
We resolve every company an article names to its real domain: from the article's own text first, then an 82-million-company reference graph, then web search. One story can produce several signals, each pointed at the right company. A partnership plus an investment in one article gives a partners_with signal for one company and a receives_financing signal for the other.
How do you filter out noise?
A second model reviews every event: is it real, is it central to the article, does it matter to B2B buyers? Roughly 40% of raw events stop there, for example recalls, sports, politics, obituaries and law-firm solicitations.