Company Expansion Keyword Monitor

Monitor public company websites, newsrooms, blogs, careers pages, and announcements for expansion keywords — hiring, funding, product launch, partnership, market entry, new offices, and acquisitions. You give it company domains and/or specific page URLs; it returns one flat, CSV-ready row per matched signal with the matched keyword, a snippet, a signal category, and a transparent signal score (0–100).

Built for B2B sales, lead generation, agencies, market researchers, and competitor-intelligence teams who want pre-classified buying/expansion signals without configuring a generic crawler or writing keyword rules from scratch.

• ✅ No login, no cookies, no API keys. Public HTML only.
• ✅ One row per matched signal, classified into a category.
• ✅ Transparent, non-AI scoring you can audit field-by-field.
• ✅ Shallow, targeted discovery — homepage plus likely signal pages, bounded per company.

---

What it does

For every company domain or URL the actor:

1. Normalizes the input (domain → homepage; URLs are fetched directly). Plain company names without a domain are skipped in V1.
2. Fetches the start page over plain HTTP and, if enabled, shallow-discovers likely expansion pages on the same domain (news, press, blog, careers, about, locations, investors, announcements), bounded by maxPagesPerCompany.
3. Matches your selected keyword groups + custom keywords against the visible page text.
4. Scores each matched category with a transparent signal_score and emits one flat row per signal, including a snippet and reason tags.

It does not do deep crawling, login/session scraping, browser rendering, contact/email enrichment, AI summaries, or historical diffing.

---

Input

Field	Type	Default	Description
startUrls	array of strings	[]	Exact company/announcement page URLs to monitor (max 500).
companies	array of strings	[]	Company domains or website URLs. Plain names without a domain are warned + skipped in V1.
enableShallowDiscovery	boolean	true	Discover likely expansion pages from each company homepage.
maxPagesPerCompany	integer	10	Pages fetched per company/domain, incl. the start page (1–50).
maxResults	integer	100	Global cap on saved signal rows (1–10000).
keywordGroups	array of strings	all 8 groups	expansion, hiring, funding, partnership, product_launch, market_entry, office, acquisition.
customKeywords	array of strings	[]	Your own expansion keywords/phrases (max 100, ≤80 chars each).
matchMode	string	phrase	phrase (case-insensitive substring) or word_boundary (whole word).
minSignalScore	integer	20	Drop matches below this score (0–100).
includeSnippet	boolean	true	Include a short snippet around the matched keyword.
deduplicate	boolean	true	Merge/skip duplicate signals.
proxyConfiguration	object	{ "useApifyProxy": true }	Datacenter, no proxy, or custom proxy URLs. Apify Residential rejected at startup.

You must provide at least one valid URL or domain across startUrls / companies.

Example inputs

1. Monitor company homepages + a newsroom for the default expansion signals

{
    "companies": ["canva.com", "atlassian.com"],
    "startUrls": ["https://www.canva.com/newsroom/"],
    "enableShallowDiscovery": true,
    "maxPagesPerCompany": 10,
    "maxResults": 100,
    "minSignalScore": 30,
    "proxyConfiguration": { "useApifyProxy": true }
}

2. Build a funding/office watchlist with your own keywords

{
    "companies": ["stripe.com", "notion.so"],
    "keywordGroups": ["funding", "office", "market_entry"],
    "customKeywords": ["new regional headquarters", "expanding into Australia"],
    "matchMode": "word_boundary",
    "minSignalScore": 40,
    "proxyConfiguration": { "useApifyProxy": true }
}

3. Minimal run — just give it companies

{ "companies": ["canva.com", "https://www.atlassian.com/company/news"] }

---

Output

One flat row per matched company/page/keyword signal:

• Identity/source: input_value, company_domain, company_name_guess, source_url, canonical_url, page_title, page_type (home/news/press/blog/careers/locations/investors/about/other), http_status
• Match: signal_category, matched_keyword, match_snippet, matched_count, first_match_text_hash
• Scoring: signal_score (0–100), signal_label (weak/moderate/strong/very_strong), reason_tags
• Runtime: scraped_at

Expansion signals — all fields (table view)

Sample record

Real output — an office-expansion signal detected on Stripe's newsroom (one of several categories the same page produced — see the run note below):

{
    "input_value": "https://stripe.com/newsroom/news",
    "company_domain": "stripe.com",
    "company_name_guess": "Stripe",
    "source_url": "https://stripe.com/newsroom/news",
    "canonical_url": "https://stripe.com/newsroom/news",
    "page_title": "Stripe Newsroom: The Latest News & Announcements",
    "page_type": "press",
    "signal_category": "office",
    "matched_keyword": "new headquarters",
    "match_snippet": "…Corporate Stripe opens new headquarters in Dublin as Ireland's internet economy surges October 9, 2025 Product Stripe launches new products…",
    "matched_count": 1,
    "signal_score": 65,
    "signal_label": "strong",
    "reason_tags": "press_page|multiple_keywords|new_office_signal|multiple_categories",
    "first_match_text_hash": "3c58acf7",
    "http_status": 200,
    "scraped_at": "2026-06-17T05:38:16.273Z"
}

From that single Stripe newsroom page the run produced distinct rows for office, partnership, product_launch, market_entry, acquisition, funding, and expansion — each its own commercially-distinct signal with its own keyword, snippet, and score.

A run summary is stored in the default key-value store under RUN_SUMMARY with counters such as inputs_total, pages_fetched, raw_matches_found, results_saved, duplicates_removed, filtered_out, and charged_events.

---

Signal score

A transparent 0–100 weighted sum (see PRD §7), based only on visible public fields:

• +20 for any selected keyword match
• +10 if the keyword appears in the page title or H1
• +10 if the page type is press, news, careers, locations, or investors
• +10 if at least 3 related keywords appear on the page
• +15 if the category is high-intent (funding, market_entry, office, acquisition, partnership, product_launch)
• +10 if the page looks recent (visible date or URL year/month within ~18 months)
• +10 if multiple expansion categories appear on the same page
• +15 if a custom keyword matched

Labels: 0–24 weak, 25–49 moderate, 50–74 strong, 75–100 very_strong.

---

Pricing

Pay Per Event. One event, expansion-signal-result, is charged only after a valid, unique signal row is successfully pushed to the dataset. Duplicate signals, filtered-out matches, failed pages, blocked requests, and skipped invalid inputs are never charged. The actor honours your per-run spending limit and stops cleanly when it is reached.

---

🚦 Proxy policy

Use Apify Datacenter proxy or no proxy for normal runs — both work reliably for public company pages at this actor's conservative concurrency.

Apify Residential proxy is not supported. The actor fails at startup if apifyProxyGroups includes RESIDENTIAL. Reason: in pay-per-event actors, residential bandwidth (~$8/GB) is billed to the developer, not the run user, so a single bandwidth-heavy run could exceed the per-result event revenue.

If you genuinely need residential routing, supply your own residential provider via the proxy editor's Custom proxy URLs field — that traffic goes through your provider, not Apify, and is unaffected:

http://user:pass@proxy.iproyal.com:12321
http://user:pass@proxy.brightdata.com:22225
http://user:pass@proxy.oxylabs.io:7777

---

Notes & limitations

• Public HTML only, HTTP-first. No headless browser in V1, so pages that render their entire body in client-side JavaScript may yield little visible text.
• Shallow by design. Discovery follows only likely-signal links from the homepage, bounded by maxPagesPerCompany — it is not a whole-site crawler.
• Plain company names (no domain) are skipped in V1; supply a domain or URL.
• phrase mode is a case-insensitive substring match (broad recall); use word_boundary for stricter whole-word matching.
• This actor is built to clone into specialised variants (funding monitor, office-expansion monitor, product-launch monitor, partnership monitor).