B2B Sales Intelligence Scraper — Company Signal Scoring & Lead Prioritization

Score and prioritize your prospect list by real buying signals — funding rounds, hiring surges, executive changes, and news activity — so your sales team reaches out at exactly the right moment.

No subscriptions. No bloated data platforms. Paste a list of companies and get back an A/B/C/D priority grade with a one-sentence rationale for each. At $0.10 per company, it's the fastest way to add trigger-based intelligence to any outbound workflow.

---

What It Does

Given a list of company names (and optionally domains / LinkedIn slugs), this Actor runs a parallel intelligence sweep and returns one structured JSON record per company:

Signal	What's Collected
Hiring Velocity	Job count (30-day), active departments, senior role mix, headcount delta
Funding Activity	Latest round type, amount, date, lead investors, recency bucket
News & Press	Article count (30 days), themes (funding / expansion / leadership / product)
Composite Score	A/B/C/D priority grade + one-sentence rationale explaining the grade

Results are sorted grade A → D so the highest-priority accounts are always at the top.

---

Who It's For

SDRs and AEs running outbound campaigns who need to know which accounts are in an active buying window right now — not six months ago.

RevOps and sales ops teams enriching CRM records or scoring an account list before a campaign launch. Drop the output directly into HubSpot, Salesforce, or any spreadsheet.

Founders and solo sellers who do their own prospecting and want the equivalent of a weekly analyst report at a fraction of the cost.

VC analysts doing lightweight deal sourcing or monitoring a watchlist of portfolio-adjacent companies for new rounds and leadership changes.

AI agent builders connecting this Actor as an MCP tool — it exposes a clean compact output mode purpose-built for LLM pipelines.

---

Pricing

$0.10 per company profile — charged against your Apify compute units at run time. No subscription, no seat fees.

This is an introductory rate while the filtering system is actively being tuned. The biggest ongoing challenge in a scraper like this is cross-contamination — company names like "Arc," "Forge," "Wave," and "Mesh" appear constantly in unrelated contexts (Intel GPUs, SpaceX rockets, Wi-Fi routers). Filtering these out accurately is a continuous engineering effort, which is reflected in the pricing. If you run it and a specific company seems mis-scored due to irrelevant articles, reach out and a patch is typically deployed within 24 hours.

---

Quick Start

No-Code Input (paste names, one per line)

Open the Actor's Input tab, find the Company List field, and paste names directly:

Salesforce
Notion Labs
Linear
Rippling
Ashby

That's all you need. Optional metadata (domain, LinkedIn slug) can be added in the Advanced JSON input for better accuracy.

API / JSON Input

{
  "companies": [
    { "name": "Notion Labs", "domain": "notion.so", "linkedinSlug": "notionhq" },
    { "name": "Linear", "domain": "linear.app" },
    { "name": "Figma" }
  ],
  "outputFormat": "compact",
  "maxConcurrency": 3
}

domain, linkedinSlug, and crunchbaseSlug are optional but improve signal accuracy. When crunchbaseSlug is provided and domain is omitted, the Actor resolves the domain automatically from Crunchbase.

---

Output

Compact Format (default — recommended for most use cases)

{
  "companyName": "Notion Labs",
  "domain": "notion.so",
  "grade": "A",
  "score": 72,
  "priorityFlag": true,
  "topSignal": "Series C $275M, 45d ago",
  "rationale": "Grade A — Reach out now. Actively hiring at high velocity (24 roles in 30d); Series C $275M (45d ago) — post-round spend window open; 8 news articles (expansion, product).",
  "runTimestamp": "2026-04-13T14:23:00.000Z",
  "errors": []
}

Full Format (outputFormat: "full")

Returns the complete signal payload including all job listings, funding round history, article headlines, and score breakdown. Useful for CRM enrichment, deeper analysis, or building custom scoring on top of the raw signals.

{
  "companyName": "Notion Labs",
  "domain": "notion.so",
  "scoreBreakdown": {
    "hiringScore": 28,
    "fundingScore": 30,
    "newsScore": 14,
    "totalScore": 72,
    "grade": "A",
    "rationale": "..."
  },
  "hiringSignal": {
    "jobCount": 41,
    "jobCount30Days": 24,
    "topDepartments": ["Engineering", "Sales", "Product"],
    "seniorRoles": 5,
    "hiringVelocity": "high"
  },
  "fundingSignal": {
    "latestRoundType": "Series C",
    "latestRoundAmountUsd": 275000000,
    "latestRoundDate": "2024-10-15",
    "leadInvestors": ["Sequoia Capital"],
    "daysSinceLastRound": 45,
    "fundingRecency": "recent"
  },
  "newsSignal": {
    "articleCount30Days": 8,
    "keyThemes": ["expansion", "product"],
    "hasExpansionNews": true,
    "hasLeadershipNews": false,
    "hasProductNews": true,
    "overallSentiment": "positive",
    "topHeadlines": ["Notion launches AI meeting notes...", "..."]
  },
  "runTimestamp": "2026-04-13T14:23:00.000Z",
  "errors": []
}

---

Signal Scoring

How the A–D Grade Works

Component	Max Points	Key Drivers
Hiring	40	Job count (30d), senior role mix, department diversity, velocity
Funding	40	Round recency, round size, lead investor quality
News	20	Article count, themes (funding/expansion boost), sentiment penalty
Total	100

Grade	Score	Recommended Action
A	70–100	Reach out now — multiple strong signals
B	45–69	Follow up this week
C	20–44	Monitor — revisit next cycle
D	0–19	Deprioritise — acquired, dormant, or no signal

priorityFlag: true is set for any Grade A company, or any company with a recent funding event (< 6 months) regardless of grade.

---

Input Options

Field	Type	Default	Description
companiesText	string	—	No-code input. One company name per line. Easiest way to get started — no JSON required.
companies	array	—	Advanced input. JSON array with optional domain, linkedinSlug, crunchbaseSlug per company for better accuracy.
outputFormat	compact | full | mcp	compact	compact returns grade + rationale (best for most uses). full returns all raw signals. mcp is for AI agent pipelines.
maxConcurrency	integer 1–10	3	Companies processed in parallel. Lower values reduce proxy costs.
includeRawSignals	boolean	false	Attach full job listings, articles, and funding rounds to output. Only applies in full mode.
cacheTtlHours	integer 0–168	24	Re-use cached results within this window. Set to 0 for always-fresh data.

Tip: We recommend keeping runs to 100 companies or fewer. For larger prospect lists, split into batches of 100 and use cacheTtlHours to avoid re-processing.

---

Using via MCP (AI Agent Pipelines)

This Actor integrates directly with Claude and other MCP-compatible AI agents via Apify's off-the-shelf MCP server — no custom wrapper needed.

$npm install -g @apify/actors-mcp-server

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": ["-y", "@apify/actors-mcp-server"],
      "env": {
        "APIFY_TOKEN": "your-apify-api-token",
        "ACTORS": "<your-apify-username>/b2b-sales-trigger-intelligence"
      }
    }
  }
}

Use outputFormat: "mcp" when calling from an agent — it returns compact fields plus a pre-rendered plain-text summary sorted grade A → D, optimised for LLM consumption.

---

Homonym Filtering — The Hard Part

The biggest technical challenge in a news-based signal scraper is name collision. Startup names like "Arc," "Forge," "Mesh," "Loop," "Bolt," and "Wave" appear constantly in unrelated contexts — Intel Arc GPUs, SpaceX Falcon rockets, Wi-Fi mesh routers, Boring Company tunnel loops. Without aggressive filtering, these articles inflate scores and produce false positives.

This Actor maintains a continuously-updated collision blocklist (currently 17+ refinement cycles) that detects when a company name appears in the context of a known unrelated entity and rejects those articles before scoring.

Despite this, edge cases exist. The filter is a living system, not a finished product. If you see a company scoring unexpectedly high or low:

1. Check the rationale field — it explains exactly which signal drove the grade
2. Check topHeadlines in full output to spot-check article relevance
3. Report the issue via the contact form on this Actor's page — patches for specific company names are typically deployed within 24 hours

This transparency and responsiveness is part of what the $0.10/company price reflects.

---

Architecture

src/
├── main.ts               # Entry point — reads input, drives concurrency
├── orchestrator.ts       # Per-company sweep — runs scrapers in parallel
├── types.ts              # TypeScript interfaces
├── config.ts             # Thresholds, URLs, keyword dictionaries, funding overrides
├── scrapers/
│   ├── newsScraper.ts    # Google News RSS + homonym collision filtering
│   ├── fundingScraper.ts # Crunchbase + Google News fallback + public company guard
│   └── jobsScraper.ts    # ATS boards (Greenhouse/Lever/Ashby) → LinkedIn → Indeed
├── scoring/
│   └── scorer.ts         # A–D composite scoring model
└── output/
    └── formatter.ts      # Profile assembly + rationale text generation

Each scraper runs independently via Promise.allSettled. If LinkedIn is rate-limited, the profile still gets funding and news signals and a valid (reduced) score. Errors are captured in the errors[] array — they never fail the run.

---

Development

npm install
npm run build
apify run

---

Keywords

B2B sales intelligence, sales trigger data, buying intent signals, lead prioritization, account scoring, prospect enrichment, outbound sales automation, trigger-based selling, company news monitoring, funding alerts, hiring signals, executive change detection, account-based marketing, ABM tools, SDR tools, RevOps automation, CRM enrichment, Salesforce enrichment, HubSpot enrichment, sales prospecting tool, revenue intelligence, pipeline prioritization, intent data alternative, ZoomInfo alternative, Apollo alternative, Crunchbase alternative, VC deal sourcing, competitive intelligence, MCP tool, AI agent sales tool, Apify sales scraper

---

April 2026 — actively maintained.