Instagram Contact & Lead Scraper

Enrich Instagram profiles into structured contact leads. For every target this actor fetches Instagram's authoritative web_profile_info payload and emits the profile's real contact data — business phone, business email, external URL, and bio-extracted emails / phones / WhatsApp / Linktree — together with profile attributes (category, verified & business flags, follower count).

It replaces snippet-regex guessing with authoritative profile data: phone numbers are validated and normalized to E.164, so prices, follower counts, years and post IDs are rejected instead of being reported as phones.

Input

Provide at least one of:

• usernames (array) — direct targets. Accepts:
  • plain usernames: natgeo
  • @handles: @natgeo
  • profile URLs: https://instagram.com/natgeo
  • post/reel URLs: https://instagram.com/p/CXXXXXX/ (the post owner is resolved and enriched; the row keeps postUrl + postShortcode)
  • numeric user IDs: 173560420
• searchKeywords (array) — optional Google-SERP discovery. Each keyword runs site:instagram.com "<dial code>" "<keyword>", then every candidate profile is enriched exactly like a direct target.

Other parameters:

Field	Meaning
country	Dial code used to localize the discovery query and to parse local (non-+) phone numbers into E.164.
maxProfiles	Max candidate profiles discovered per keyword (does not cap direct targets).
hasPhoneOnly / hasEmailOnly / businessOnly / verifiedOnly / minFollowers	Lead filters. Error rows are always kept.
concurrency / maxRetries / delayMs	Enrichment tuning.
proxyConfiguration	Proxy for enrichment requests (RESIDENTIAL recommended; a fresh IP per profile). SERP discovery always uses the Apify GOOGLE_SERP proxy.

Example input

{
  "usernames": ["nike", "@natgeo"],
  "searchKeywords": ["real estate agent"],
  "country": "United States (+1)",
  "hasPhoneOnly": false,
  "businessOnly": false,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }
}

Output

One row per target (success or error). Unavailable values are null (never a fabricated 0 or a false-positive number).

Field	Description
inputType	username / profileUrl / postUrl / userId / invalid.
mode	direct (from your usernames list) or discovery (found via keyword SERP).
username, userId, fullName, profileUrl	Profile identity.
postUrl, postShortcode	Set when the input was a post/reel URL.
keyword, serpTitle, serpSnippet, serpUrl	Discovery context.
businessPhone	Authoritative phone from web_profile_info (business/contact/public phone).
phoneE164, phoneCountry	Normalized + validated phone and its country.
businessEmail, emailDomain	Authoritative email and its domain.
externalUrl, whatsapp, linktree, bioLinks	Links/contacts from the profile.
emailsFromText, phonesFromText	Deterministic extraction from bio + SERP snippet (phones E.164-validated).
biography, categoryName, isBusinessAccount, isVerified, isPrivate, followerCount	Profile attributes.
scrapedAt	ISO-8601 UTC timestamp.
error	Reason string when a target could not be enriched (not-found, private, auth-gated, etc.).

Notes on coverage

• Reachability (anti-bot reality). Instagram's logged-out anti-bot returns an intermittent 401/403/429 that is variance per IP — the same account can fail on one residential IP and enrich cleanly on another. To counter this, each retry runs on a fresh residential proxy IP with a fresh Chrome-TLS client and a fresh app_id/csrf bootstrap handshake, with exponential backoff + jitter. This materially raises reachability versus retrying on one IP: on a fixed 18-account benchmark, reachability rose from 12/18 (66.7%) with same-IP retries to 17/18 (94.4%) with fresh-IP retries at maxRetries: 3. It does not reach 100% (Instagram can block every IP tried within the budget). Raising maxRetries trades run time for a higher hit rate. Unreachable targets return an honest error row (never fabricated data).
• businessPhone / businessEmail are only present when the account has published them and Instagram returns them unauthenticated. Instagram gates these behind login for logged-out requests on most accounts, so they are frequently null. Lead value in that case comes from the E.164-validated phones/emails/WhatsApp/Linktree extracted from the bio, bio_links and (in discovery) the SERP snippet. null here is honest absence, not a bug.
• Private and non-existent profiles produce a clean row (private accounts enrich with isPrivate: true and limited fields; non-existent ones return an error row) rather than fabricated data.
• The userId input mode depends on Instagram's i.instagram.com/users/{id}/info endpoint, which is auth-gated and blocked from Apify's datacenter network — it returns a clean error row.
• Every input produces exactly one row (success or an error row); one bad profile never aborts the batch. Every network call is bounded by a timeout and each profile by a total time budget.
• Dedup is global (by resolved username / user ID / URL) across direct and discovery targets.

Compliance

Use only for publicly available data and in accordance with Instagram's terms and applicable data-protection law. Private profiles are not accessed.