🐦 X/Twitter Lead Email Finder

Find public email addresses on X (Twitter) — search by keyword, filter by email domain and location, and get every lead scored, deduplicated, and MX-verified automatically. Built on a Google-search extraction pipeline (no X API key needed).

---

🚀 Why Choose This Actor?

• 🧮 Confidence scoring — every lead gets a 0-100 score from result type (profile vs. tweet), contact-intent phrasing, local-part quality, domain-filter match, MX validity, and cross-keyword corroboration
• 🔗 Real cross-keyword deduplication — the same email found under two different keyword searches is merged into one row instead of duplicated, and the corroboration itself boosts confidence
• 🔬 MX domain verification — every discovered email's domain gets a cached, per-run MX-record check
• 🧩 Result-type + handle extraction — every lead is tagged profile or tweet and gets a clean handle field, so you don't have to parse URLs yourself
• 📬 Contact-intent detection — flags leads whose bio/tweet text explicitly invites contact ("DM for collabs", "business inquiries:", "email me")
• 🌍 Keyword + location + domain targeting — narrow results to a niche, a city/region, and specific email providers or business domains

---

✨ Key Features

• 🔍 Multi-keyword search — process as many search terms as you like in one run
• 📧 Domain filtering — keep only Gmail/Outlook/Yahoo/etc., or your own business domain
• 🧩 resultTypeFilter — keep only profile-bio leads, only tweet mentions, or both
• ✅ minConfidenceScore — automatically drop low-quality matches
• 🔬 verifyEmailDomains — MX-record validation, cached per domain per run
• 📍 Location targeting — narrow to a city/region
• 📊 Run analytics — a RUN_SUMMARY written to the key-value store every run

---

📥 Input Configuration

{
  "searchKeywords": ["marketing", "photographer"],
  "targetPlatform": "Twitter",
  "locationFilter": "London",
  "emailDomainFilter": ["@gmail.com"],
  "maxEmailsPerKeyword": 15,
  "resultTypeFilter": [],
  "minConfidenceScore": 40,
  "verifyEmailDomains": true,
  "scraperEngine": "legacy",
  "proxyConfig": { "useApifyProxy": true }
}

Field	Type	Description
searchKeywords	array	🔍 Keywords to search for (required)
targetPlatform	string	🌐 Target site (currently Twitter)
locationFilter	string	📍 Optional city/region filter
emailDomainFilter	array	📧 Keep only emails from these domains (e.g. @gmail.com); empty = all domains
maxEmailsPerKeyword	integer	📊 Raw hits to collect per keyword before cross-keyword dedup (default 10)
resultTypeFilter	array	🧩 Keep only profile and/or tweet results; empty = both
minConfidenceScore	integer	✅ Drop leads below this 0-100 confidence score
verifyEmailDomains	boolean	🔬 Run an MX lookup per discovered domain (default true)
scraperEngine	string	⚙️ Extraction engine (currently legacy)
proxyConfig	object	🛡️ Proxy config; leave empty for the auto-selected default

---

📤 Output Format

{
  "emailAddress": "jane.doe@gmail.com",
  "domain": "gmail.com",
  "hasValidMxRecords": true,
  "confidenceScore": 83,
  "handle": "janedoephoto",
  "resultType": "profile",
  "contactIntentDetected": true,
  "crossKeywordMatchCount": 1,
  "matchedKeywords": ["photographer"],
  "sourcePlatform": "x.com",
  "resultTitle": "Jane Doe Photography (@janedoephoto) / X",
  "snippetText": "📸 Wedding & portrait photographer. Bookings: jane.doe@gmail.com",
  "sourceUrl": "https://x.com/janedoephoto",
  "matchedLocation": ""
}

Field	Description
emailAddress	✅ The discovered email address
domain	🌐 Email domain, used for MX verification and scoring
hasValidMxRecords	🔬 Whether domain has valid MX records (null if verifyEmailDomains is off)
confidenceScore	🧮 0-100 score — see scoring breakdown below
handle	👤 New. The X/Twitter @handle, parsed from the URL or (when Google's link is undecodable — see note below) recovered from the visible card text
resultType	🧩 New. profile (bio page) or tweet (a specific post)
contactIntentDetected	📬 New. True if the title/snippet uses contact-inviting phrasing
crossKeywordMatchCount / matchedKeywords	🔗 New. How many, and which, of your search keywords surfaced this same email
sourcePlatform	🌐 Resolved platform domain (x.com)
resultTitle / snippetText	📝 Google's result title and snippet
sourceUrl	🔗 Link to the profile or tweet (see note below on when this is a Google redirect link)
matchedLocation	📍 The location filter used, if any

A RUN_SUMMARY object is written to the key-value store at the end of every run: totalEmailsCollected, skippedLowConfidence, skippedByResultTypeFilter, uniqueDomains, domainsWithValidMx/domainsWithInvalidMx, resultTypeBreakdown, avgConfidenceScore.

---

🧮 How confidence scoring works

base score        = 60 (profile) or 40 (tweet)
+ local-part bonus = +10 role-style (info@, press@) · +6 structured (jane.doe@) · +0 plain
+ contact intent    = +15 if the text says "DM for collabs", "business inquiries:", "email me", etc.
+ domain filter      = +10 if you set emailDomainFilter and it matched
+ MX record          = +8 valid · −20 invalid · +0 not checked
+ cross-keyword       = +5 per extra keyword that also found this email, capped at +15
= clamped to 0-100

---

🔬 Honest notes on real search behavior (verified live)

• A broad keyword with no domain filter can legitimately return 0 emails. Emails are relatively rare inside Google's truncated snippets — a live test on "marketing agency" with no domain filter returned 18 real matching profiles/tweets but 0 with a visible email in the snippet. Add an emailDomainFilter entry (e.g. @gmail.com, or your own business domain) to sharply narrow the search toward pages that actually contain an email — a live test on "photographer" + @gmail.com found 3 verified leads immediately.
• sourceUrl is sometimes a Google redirect link, not the raw profile/tweet URL. Google increasingly serves an opaque /goto?url=... redirect token for these results that cannot be decoded back to the real destination (verified live — this affected 100% of results in a recent sample and, unfixed, would silently drop every result; this actor's extraction now handles it). The link still resolves to the right page when opened, but when it happens, resultType defaults to profile (it can't be determined from an undecodable URL) and handle falls back to any @handle mention visible in the snippet text — which is not always present.

---

⚠️ Legal & Ethical Notice

• This actor only collects data from publicly available Google search results
• Only emails that appear in public X/Twitter posts or bios are collected
• Users are responsible for compliance with GDPR, CAN-SPAM, CCPA, and applicable laws
• Do not use collected emails for spam or unsolicited bulk messaging

---

❓ Frequently Asked Questions

Q: Does this need an X/Twitter account or API key? A: No — it searches Google for public X/Twitter content.

Q: Why did my run return 0 emails? A: Broad keywords without a domain filter often have a low email-in-snippet rate. Add emailDomainFilter entries and/or more specific keywords.

Q: What's the difference between profile and tweet results? A: profile results are bio pages — durable, intentional contact info. tweet results are a specific post that happened to mention an email — a one-off, weaker signal, reflected in the lower base confidence score.

Q: Do I need Apify proxies? A: Leave proxyConfig empty and the actor auto-selects the Google-search proxy tier; recommended for reliability.

---

📞 Support & Feedback

• 🐛 Report issues in the Issues tab
• 💡 Custom features: dev.scraperengine@gmail.com
• ⭐ If this actor saves you time, please leave a review!