News Intelligence v4: 40 Subtypes, Daily Cadence, and 16 New Schema Fields

News signals upgraded to v4: 40 subtypes (up from 30), daily delivery cadence, and 16 new schema fields including sentiment, confidence scores, and second company enrichment.

Why This Matters

News signals are the #1 most consumed signal type across our OEM partners. By volume, by API calls, by downstream activation. When ZoomInfo, 6sense, or RocketReach surface a "Company X acquires Company Y" event to their users, that data originates from this pipeline.

The problem with v3: weekly cadence meant events could be up to 7 days stale by the time they reached end users. In a world where sales teams compete on timing, a week-old acquisition signal is barely actionable. And the schema was missing fields that partners needed to build richer experiences—things like sentiment, confidence scoring, and structured data on the second company involved in M&A/partnership events.

v4 fixes all of it.

Daily Cadence: 24-Hour Event Surfacing

News signals now deliver daily. Company events surface within 24 hours of publication, down from up to 7 days.

What this means in practice:

• A company announces an acquisition on Tuesday morning. Your users see it Tuesday night or Wednesday morning.
• A competitor loses a major client. Your platform surfaces it same-day.
• An earnings report drops. The structured signal is available within hours, not next week.

For GCS delivery partners: daily files land by 06:00 UTC each morning. File naming shifts from news_weekly_YYYY-MM-DD.jsonl to news_daily_YYYY-MM-DD.jsonl. Your ingestion pipeline should handle both patterns during the transition window (through July 15, 2026), after which weekly files stop.

For API partners: no change needed. The /v1/companies/search endpoint returns signals as they become available. You'll simply see fresher data.

Typical daily volume: ~12,000-18,000 news records per delivery, compared to ~60,000-90,000 in the old weekly batches. Same total volume, just distributed across 7 days instead of one dump.

All 40 Subtypes

We went from 30 to 40 news subtypes. Here's the complete taxonomy:

Corporate Structure (8 subtypes)

• acquires – Company acquires another company
• merges_with – Merger announcements
• spins_off_company – Corporate spin-offs and divestitures (New in v4)
• spins_off_division – Division-level spin-offs (New in v4)
• splits_into – Company splits into multiple entities (New in v4)
• rebrands_to – Company rebrands or name changes (New in v4)
• declares_bankruptcy – Bankruptcy filings and proceedings (New in v4)
• goes_public – IPO announcements

Partnerships & Relationships (6 subtypes)

• partners_with – New partnerships announced
• ends_partnership_with – Partnership dissolutions (New in v4)
• wins_client – New customer wins
• loses_client – Customer churn (named client lost) (New in v4)
• invests_in – Investment activity
• is_invested_in_by – Receives investment/funding

Financials (6 subtypes)

• raises_funding – Fundraising rounds
• has_valuation – Valuation announcements/updates (New in v4)
• has_earnings – Quarterly/annual earnings reports (New in v4)
• has_revenue – Revenue disclosures and milestones (New in v4)
• gets_acquired – Company is acquired (target side)
• receives_grant – Grant funding received

Product & Operations (8 subtypes)

• launches_product – New product or feature launches
• expands_to – Geographic or market expansion
• opens_office – New office or facility openings
• closes_office – Office closures or consolidation
• wins_award – Industry awards and recognition
• faces_lawsuit – Litigation (defendant side)
• files_lawsuit – Litigation (plaintiff side)
• lays_off – Layoffs and workforce reductions

Leadership & People (6 subtypes)

• hires_executive – Executive hires
• promotes_executive – Executive promotions
• loses_executive – Executive departures
• appoints_board_member – Board appointments
• hires_mass – Mass hiring announcements
• restructures – Organizational restructuring

Market & Industry (6 subtypes)

• enters_market – Entry into new market segments
• exits_market – Market exits or sunset announcements
• faces_regulation – Regulatory actions or compliance events
• wins_contract – Major contract wins
• loses_contract – Contract losses
• pivots_strategy – Strategic pivots or direction changes

16 New Schema Fields

v4 adds 16 fields to the news signal schema. All are additive—existing fields remain unchanged and these appear alongside them. If your parser ignores unknown fields (as it should), zero code changes required.

Code Example: Querying News Signals via API

curl -X POST "https://signals.autobound.ai/v1/companies/search" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "signal_types": ["news"],
    "subtypes": ["acquires", "raises_funding", "has_earnings"],
    "filters": {
      "confidence_score_min": 0.85,
      "source_credibility": ["tier_1", "tier_2"],
      "article_sentiment": ["positive", "neutral"],
      "date_range": {
        "start": "2026-06-14",
        "end": "2026-06-15"
      }
    },
    "limit": 50
  }'

Response (truncated):

{
  "results": [
    {
      "signal_id": "news_20260615_acq_7f3a9b",
      "signal_type": "news",
      "subtype": "acquires",
      "company_name": "Datadog",
      "company_domain": "datadoghq.com",
      "batch_id": "news_daily_2026-06-15_001",
      "data": {
        "summary": "Datadog acquires CloudSecOps for $340M...",
        "source_url": "https://techcrunch.com/2026/06/14/...",
        "published_date": "2026-06-14T14:30:00Z",
        "event_date": "2026-06-14T00:00:00Z",
        "second_company_name": "CloudSecOps",
        "second_company_domain": "cloudsecops.io",
        "article_sentiment": "positive",
        "confidence_score": 0.97,
        "news_category": "corporate_structure",
        "geographic_region": "US",
        "source_credibility": "tier_1",
        "is_breaking": true,
        "extraction_model_version": "news-v4.1.0"
      }
    }
  ],
  "total": 847,
  "has_more": true,
  "next_cursor": "eyJvZmZzZXQiOjUwfQ=="
}

Filtering by New Fields

The new fields are particularly useful for building quality-aware integrations. Some patterns we recommend:

High-confidence only (precision-focused)

"filters": {
  "confidence_score_min": 0.90,
  "source_credibility": ["tier_1"]
}

This gives you signals from major publications where the model is 90%+ confident in classification. Expect ~40% fewer records but significantly higher precision.

Second company enrichment (relationship signals)

"filters": {
  "has_second_company": true,
  "subtypes": ["acquires", "partners_with", "invests_in", "ends_partnership_with"]
}

Roughly 5,000 records per daily delivery include structured second-company data. Previously this was buried in the summary text and required your own NLP to extract.

Breaking news prioritization

"filters": {
  "is_breaking": true
}

Surfaces only signals detected within 6 hours of article publication. These are the freshest, most time-sensitive events. Good for real-time alerting use cases.

Migration Notes

The short version: this upgrade is non-breaking. Here's the detailed breakdown:

What changed

• Cadence: Weekly delivery files are being phased out. Daily files started June 1, 2026. Weekly files continue through July 15, 2026, then stop.
• File naming: news_daily_YYYY-MM-DD.jsonl replaces news_weekly_YYYY-MM-DD.jsonl
• New subtypes: 10 new subtype values will appear in the subtype field. If your integration uses an allowlist enum, add the new values.
• New fields: 16 fields added to the data object. Additive only, no existing fields removed or renamed.

What did NOT change

• All existing fields retain their names, types, and semantics
• The 30 original subtypes are unchanged
• API endpoint paths and authentication are unchanged
• GCS bucket paths are unchanged (only file names within them)
• Signal ID format is unchanged

Action items by integration type

GCS Push partners:

1. Update file pattern matching to accept news_daily_* files (do this before July 15)
2. Optionally update your subtype enum to include the 10 new values
3. Optionally parse the 16 new schema fields

Signal API partners:

1. Optionally update your subtype enum
2. Optionally parse new fields from the response
3. Consider adjusting poll frequency (signals are now fresher, more frequent polling gives better results)

No action required if your integration handles unknown subtypes gracefully and ignores unknown fields in the JSON response.

Volume and Performance

• Daily delivery volume: 12,000-18,000 news records per day
• Subtype coverage: 40 distinct event types across 6 categories
• Second company enrichment: ~5,000 records/day include resolved second_company_domain
• Source coverage: 15,000+ news sources monitored globally
• Language coverage: English primary, with emerging coverage in DE, FR, ES, PT, JA
• Extraction latency: median 4.2 hours from article publication to signal availability
• Confidence distribution: 72% of signals score above 0.85, 91% above 0.70

What's Next

Real-time streaming delivery for news signals (sub-hour latency) is in development for Q3. We're also expanding second-company enrichment to cover 90%+ of relational subtypes, up from the current ~60%.

Resources

• Delivery Changelog (full technical details on every schema change)
• Signal Catalog (browse all 40 news subtypes with example payloads)
• API Reference: Search Signals
• May 2026 Product Update (broader context on everything else we shipped)

Questions on migration or want to discuss how to use the new fields in your integration? Reply directly or reach out via your Slack Connect channel.