👥🔗 Instagram Followers ∩ Following Scraper — Mutuals, Fans & Reciprocity Analytics

Scrape a profile's followers, following, or both in a single run — and, in both mode, compute the set relationships between the two lists: mutuals, fans (they follow you, you don't follow back), and non-followers-back (you follow them, they don't follow back), plus a run-level reciprocity summary.

This actor is the relationship-analytics member of the Instagram audience family. If you only need a raw inbound or outbound list, use the dedicated followers-only / following-only actors; use this one when you want the derived set intelligence no single-list scraper produces.

What it does

• Three modes — followers (inbound audience), following (outbound interests), both (set analytics).
• Set-relationship labels on every user row (both mode): mutual, fan, non_follower_back, with booleans is_mutual, follows_target, followed_by_target.
• Run-level summary record — mutuals_count, fans_count, non_followers_back_count, reciprocity_rate, fan_rate, completeness, scrapedAt.
• Correctness-first — set labels are only asserted when the crawl is complete. A crawl truncated by maxData or interrupted by a block is flagged (completeness: partial) and the uncertain labels are downgraded to plain follower/following instead of guessing.
• Optional profile enrichment (enrichProfiles, OFF by default) — bio, follower/following/media counts, business flags, category, public email/phone, geo, HD avatar. Absent values are null (never guessed); a coverage summary reports real presence fractions.
• Strict superset of the base actor — every original identity field is preserved; relationship, derived, and enrichment fields are added on top.

How set relationships are defined

From the target profile's perspective, for each user row:

Label	Meaning	follows_target	followed_by_target	is_mutual
mutual	in both lists	true	true	true
fan	follows the target, target does not follow back	true	false	false
non_follower_back	target follows them, they do not follow back	false	true	false

fan and non_follower_back are only emitted when the relevant list is fully crawled. On a partial crawl the row is labeled follower / following with the uncertain booleans set to null.

• reciprocity_rate = mutuals / following_scraped — of the accounts the target follows, the fraction that follow back.
• fan_rate = fans / followers_scraped — of the target's followers, the fraction the target does not follow back.

Both rates are null unless both lists are complete.

Input

Field	Type	Notes
startUrls	array	Profile URLs or usernames.
scrapeMode	enum	followers | following | both. Default both.
maxData	integer	Cap per list. 0 = unlimited. For exact set math in both mode, keep 0.
enrichProfiles	boolean	OFF by default. One extra request/user for bio/counts/business/email/phone/geo/HD.
pageSize	integer	Users per API page (≤100). Default 50.
sessionId	secret	Instagram sessionid cookie. Login-walled endpoints make it effectively required; also enables sticky-IP anti-ban.
proxyConfiguration	proxy	Residential recommended. Honored from the first request.

Output

Per-user rows carry all base identity fields (pk, username, full_name, is_verified, is_private, profile_pic_url, …) plus record_type, relationship, is_mutual, follows_target, followed_by_target, source_profile, profileUrl, enrichment_status, the enrichment fields (null when not requested/absent), and scrapedAt. A single record_type: "summary" row carries the run-level set analytics. Two dataset views are provided: Users + Relationship and Run Summary.

Anti-ban / reliability

• Proxy configuration is honored on the first request — no direct-first override.
• With a sessionId, one sticky exit IP is used for the whole run (rotating IPs under one login cookie triggers checkpoints). Without a session, the exit IP rotates per request.
• On a soft-block the actor escalates to residential while preserving your configured country.
• A block mid-crawl marks the run aborted/partial and downgrades set labels — it never fabricates rows to look complete.

Notes on scale

both mode holds only the small following list plus a set of follower pk strings in memory (never full follower objects), streaming and emitting follower rows as it goes; the follower pk set is snapshotted to the key-value store for very large runs. For accounts whose follower pk set exceeds the run's memory budget, cap with maxData (the run is then flagged partial and set math is downgraded rather than being computed on an incomplete crawl).

Keywords

instagram followers scraper, instagram following scraper, mutuals finder, non-followers-back, unfollowers, fans, reciprocity analytics, instagram audience export, follower list export, social graph analysis, instagram lead generation, bio email extractor, instagram profile enrichment, instagram relationship analytics.

Compliance

Scrapes only data visible to the supplied logged-in session. Respect Instagram's Terms of Service and applicable privacy laws (GDPR/CCPA); use exported data lawfully.