Influencer Discovery

Find the creators worth reaching out to. Influencer Discovery searches Instagram, TikTok, or YouTube by niche, visits each candidate profile, and returns an enriched, contactable influencer lead — handle, display name, follower/engagement signals, bio, external links, and any public email the creator has published.

Point it at a topic (or a few seed profiles), set your follower band, and get a clean list of leads ready for outreach, sponsorships, and campaign shortlists.

---

What it does

1. Search — For each niche/keyword/hashtag, opens the platform's hashtag or search page and gathers candidate creator handles. You can also seed the run with specific profiles to expand from.
2. Enrich — Visits every candidate and pulls handle, display name, profile URL, followers, following, post count, bio, and external links.
3. Find contact — Extracts a public email from the bio and linked pages (extractEmails).
4. Score & filter — Estimates an engagement rate from publicly visible signals when available, then filters by follower band and minimum engagement.
5. Dedupe & cap — Removes duplicates by handle and stops at maxItems.

---

Input

Field	Type	Default	Description
platform	enum (instagram, tiktok, youtube)	instagram	Network to search.
niches	array of strings	–	Keywords/hashtags to discover creators (e.g. fitness, #vegancooking). The # is optional.
seedProfiles	array of strings	–	Optional profile URLs/handles to start from and expand (e.g. @nike).
minFollowers	integer	1000	Keep creators with at least this many followers.
maxFollowers	integer	1000000	Keep creators with at most this many followers.
minEngagementRate	number	0	Minimum estimated engagement rate (%). 0 disables the filter.
maxItems	integer	100	Maximum number of leads delivered to the dataset.
extractEmails	boolean	true	Parse public emails from bios/links.
proxyConfig	object	Residential	Proxy settings. Residential strongly recommended.

At least one of niches or seedProfiles is required.

Aliases: the original creator-finder field names are still accepted, so existing input configs keep working — proxyConfiguration maps to proxyConfig, keywords to niches, seeds to seedProfiles, and limit / maxResults to maxItems.

Example input

{
    "platform": "instagram",
    "niches": ["fitness", "#homeworkout"],
    "seedProfiles": ["@gymshark"],
    "minFollowers": 5000,
    "maxFollowers": 250000,
    "minEngagementRate": 1.5,
    "maxItems": 100,
    "extractEmails": true,
    "proxyConfig": {
        "useApifyProxy": true,
        "apifyProxyGroups": ["RESIDENTIAL"]
    }
}

---

Output

Each dataset item is one influencer lead:

{
    "platform": "instagram",
    "handle": "janedoefit",
    "displayName": "Jane Doe | Home Fitness",
    "profileUrl": "https://www.instagram.com/janedoefit/",
    "followers": 84500,
    "following": 612,
    "postCount": 1340,
    "bio": "Home workouts & meal plans. Coaching inquiries: jane@janedoefit.com",
    "externalLinks": ["https://linktr.ee/janedoefit"],
    "email": "jane@janedoefit.com",
    "hasEmail": true,
    "engagementRate": 3.21,
    "niche": "fitness",
    "source": "discovery",
    "scrapedAt": "2026-07-21T12:00:00.000Z"
}

Fields that cannot be extracted are returned as null (never omitted), so downstream schemas stay stable.

---

Pricing — pay per event

This actor uses Apify's pay-per-event model. It charges a single event, profile-scraped, once per influencer lead that is actually delivered to the dataset. Discovery/search pages are never charged, and charging is deliver-then-charge: the record is pushed first, then the event fires. When your run's event budget is reached, the crawl stops gracefully — you are only billed for leads you received.

Tip: set extractEmails: true and a realistic follower band to maximize the share of contactable leads per run.

---

Limitations & reality notes

• Anti-bot: Instagram and TikTok deploy aggressive bot detection, rate limiting, and login walls. Expect blocking without good residential proxies, and reduced yield on discovery pages that require authentication. YouTube is generally the most permissive.
• Selectors drift: these platforms change their DOM and embedded-JSON shapes often. All per-platform URL builders and selectors live in src/platforms.ts so they can be tuned in one place. Extraction is defensive — missing fields degrade to null instead of crashing.
• Engagement rate is an estimate: computed from publicly visible signals only when those signals are exposed; otherwise it is null and the engagement filter is skipped for that profile.
• Emails are only public ones: the actor extracts emails creators publish themselves in bios/links. It does not guess, verify deliverability, or bypass any access controls.

---

Legal & compliance

Only public data is collected. You are responsible for using the output in compliance with each platform's Terms of Service and applicable data-protection / marketing laws (e.g. GDPR / CAN-SPAM) when contacting leads.

---

Author: Assia Fadli · License: MIT