YouTube Emerging Creator Finder & Growth Tracker

What does YouTube Emerging Creator Finder do?

Find smaller YouTube creators with evidence of strong recent video performance. Search a niche or supply channel URLs to build a shortlist with public subscriber counts, recent video statistics, breakout-video links and separate Shorts analysis. It is designed for AI and SaaS sponsorship research, and works with other search topics too.

The Actor reads public YouTube pages using HTTP requests. This version qualifies creators by their recent long-form sample; Shorts provide supplementary analysis. No YouTube API key or account cookies are required. Run it in Apify Console, call it through the API, export its dataset, or connect it to your research workflow. Apify tasks and schedules let you repeat a saved search; an optional monitor name enables comparisons between observations.

Why use this creator finder?

• Build a microcreator shortlist before reviewing channels for sponsorship fit.
• Compare recent views against a creator's subscriber count, with the supporting video URLs.
• Spot videos outperforming the creator's other sampled uploads.
• Keep Shorts performance separate from long-form performance.
• Recheck a saved shortlist and measure changes in public counters over the actual elapsed interval.

These are research signals. A view multiple is not an engagement rate, verified organic growth, audience demographics or a prediction of campaign results. The Actor does not estimate sponsorship prices or creator revenue.

How to find emerging YouTube creators

1. Enter a focused query, such as AI automation tutorials, or paste public channel URLs. Clear the search-query list for a URL-only run.
2. Set the subscriber range. The default is 1,000–100,000 subscribers.
3. Choose the maximum number of channels to return and the maximum number to investigate.
4. Run the Actor and open Creator shortlist in the Output tab. Inspect the source videos and warnings before selecting creators.
5. To measure observed growth, set a monitor name and reuse it on a later run with the same Apify account and Actor.

The first observation is a baseline. It does not claim that a creator grew before your first scan.

Input

The Input tab includes field descriptions and limits. This example searches an AI niche:

{
  "searchQueries": ["AI automation tutorials", "SaaS workflow tutorials"],
  "channelUrls": [],
  "maxChannels": 20,
  "maxCandidates": 60,
  "minSubscribers": 1000,
  "maxSubscribers": 100000,
  "recentVideos": 12,
  "includeShorts": true,
  "analysisWindowDays": 90,
  "minRecentVideos": 5,
  "outlierMultiplier": 3,
  "monitorName": "ai-saas-creators",
  "country": "US",
  "proxyConfiguration": { "useApifyProxy": true }
}

To analyze a known public channel, use a channel URL and an empty query list. This example uses a wider subscriber range; the channel still needs enough usable recent videos to qualify:

{
  "searchQueries": [],
  "channelUrls": ["https://www.youtube.com/channel/UCh6wT1M9abL0wiJz4eZTwdw"],
  "maxChannels": 1,
  "maxCandidates": 1,
  "minSubscribers": 0,
  "maxSubscribers": 1000000000,
  "recentVideos": 30,
  "analysisWindowDays": 365,
  "minRecentVideos": 3,
  "includeShorts": true
}

Field	Default	Meaning
searchQueries	Starter query when both source fields are omitted	Search topics. Explicitly empty queries and URLs are invalid.
channelUrls	[]	Public YouTube channel URLs to investigate.
maxChannels	20	Delivered unique channels, from 1 to 100.
maxCandidates	60	Candidates to enrich, from 1 to 200; must be at least maxChannels.
minSubscribers / maxSubscribers	1000 / 100000	Inclusive bounds on disclosed subscriber counts. Hidden counts cannot pass these filters.
recentVideos	12	Maximum fetched videos per format, from 5 to 30.
includeShorts	true	Fetch and analyze Shorts separately.
analysisWindowDays	90	Lookback window for recent-video statistics, from 7 to 365 days.
minRecentVideos	5	Long-form sample-size requirement, from 3 to 30 and at most recentVideos. Shorts do not qualify a channel. Breakout flags require at least 5.
outlierMultiplier	3	Breakout threshold against the other eligible videos' median, from 1.5 to 20.
monitorName	Empty	Optional persistent history; 1–40 lowercase letters, numbers or hyphens, starting with a letter or number.
country	US	YouTube search-region hint. It does not verify channel or audience location.
proxyConfiguration	Apify Proxy fallback enabled	Direct requests are tried first, then the configured proxy on failure. Custom proxy URLs are supported; residential proxy groups are not supported.

Output

Each dataset record contains one enriched channel observation. You can download the dataset in various formats such as JSON, HTML, CSV, or Excel. Use JSON to preserve the nested video, analytics and growth objects.

This abbreviated example illustrates the format; its numbers and creator are fictional:

{
  "schemaVersion": 1,
  "eventId": "example-observation-id",
  "rank": 1,
  "channelId": "example-channel-id",
  "channelUrl": "https://www.youtube.com/@example-creator",
  "name": "Example AI Creator",
  "subscriberCount": 12500,
  "subscriberCountText": "12.5K subscribers",
  "subscriberCountIsExact": false,
  "analytics": {
    "longform": {
      "medianViews": 6200,
      "averageViews": 10400,
      "medianViewsToSubscribersRatio": 0.496,
      "viewsAreExact": false,
      "ratiosAreExact": false,
      "breakoutVideos": []
    },
    "medianViewsToSubscribersRatio": 0.496
  },
  "growth": {
    "isBaseline": true,
    "previousObservedAt": null,
    "elapsedDays": null,
    "subscriberChange": null,
    "subscriberChangePercent": null,
    "matchedVideoViewsChange": null
  },
  "observedAt": "2026-09-24T12:00:00.000Z",
  "warnings": []
}

The default key-value store's SUMMARY record reports discovery, filtering, failed candidates, warnings and limits. Fewer delivered profiles than requested can be a valid result. A fatal run can also create an ERROR record.

Data fields

Field	Contents
channelId, channelUrl, name, description	Public channel identity and description.
subscriberCount, subscriberCountText, subscriberCountIsExact	Parsed count, original text and precision. 12.5K is approximate.
videoCount, totalViews, country, joinedAt	Public profile metadata when exposed; unavailable values are null. Total views, country and join date may be unavailable in the fetched header.
externalLinks	Public profile links visible in the fetched page, sometimes only one featured link.
discoverySources	Search or supplied-URL evidence explaining where the candidate came from.
videos, shorts	Separate bounded samples with URLs, titles, counts, date labels and available durations.
analytics.longform, analytics.shorts	Median and mean views, views/subscribers ratio, sample coverage and breakout evidence for each format.
growth	Actual interval and counter differences against the last saved observation of that channel.
rank, eventId, observedAt	Run-local shortlist position and observation identifiers.
warnings	Missing-data, small-sample and precision limitations.

How video signals are calculated

Only sampled videos with usable publication dates and known view counts inside analysisWindowDays enter view statistics. Date labels such as “2 weeks ago” produce approximate publication dates, marked with publishedAtIsApproximate. Undated videos remain evidence but cannot establish that they fall inside the window.

The views/subscribers ratio divides median eligible video views by the channel's disclosed subscriber count. A video is a breakout when its views meet outlierMultiplier times the median of the other eligible sampled videos in the same format. At least five eligible videos and a positive comparison median are required. Missing or zero baselines do not produce infinite multiples.

Statistics describe the available sample, not a channel's complete publishing history. Shorts never enter the long-form baseline. Inspect sampleCoverage alongside the numbers.

Channels qualify only when their usable recent long-form sample meets minRecentVideos. The shortlist ranks qualifying channels by descending long-form median views/subscribers ratio, then median long-form views, with channel ID as a stable tie-breaker. This ranking measures sampled reach relative to channel size; it is not a forecast of future growth.

How repeat-scan growth works

With a nonempty monitorName, each delivered channel updates its saved observation. Reuse that name with the same account and Actor to compare subsequent observations. A new name starts fresh history. An empty name produces independent snapshots.

The growth interval is the actual time between observations, expressed as elapsedDays; it is not automatically a 7-day or 30-day period. Matched-video view changes compare only videos present with known counts in both samples. Newly observed videos and videos that fell out of the bounded sample are counted separately. A newly observed video is not necessarily newly published, and an absent video is not necessarily deleted.

Public counters can be rounded, hidden, revised or decrease. Precision flags show when a comparison relies on approximate counts. Every run returns qualifying observations again; this is not a changes-only feed.

History retains the latest 1,000 delivered channels within a 16 MB state limit. A channel removed from retained history starts a new baseline if observed later. Use a focused monitor to preserve the comparisons you need.

Prepared output can survive an interrupted delivery. If a monitor has pending records, the next run drains those records only; a following run starts a fresh scan. Retried delivery after an ambiguous interruption can produce duplicate records and charges. Use eventId to deduplicate observations in your downstream data.

How much does it cost?

The launch price is $0.98 per 1,000 delivered creator profiles ($0.00098 per creator-profile event), plus a $0.00005 Actor-start event. The current Pricing tab is authoritative for the active rate and any additional platform charges.

At that profile rate, 20 delivered channels cost $0.0196 in profile events, and 100 cost $0.098, plus the start event. A channel's videos, Shorts and analytics are included in its profile event; they are not separately charged output records. Repeated observations of the same channel on later runs are charged again when delivered. Candidates excluded by filters are not profile events.

The minimum run budget is $0.01. This is the minimum spending cap you can set, not a minimum charge; actual event charges depend on the delivered profiles and the start event.

maxChannels and the Apify spending cap limit delivered output. maxCandidates limits investigation work; the spending cap can also reduce the candidate budget. Start with a narrow query and a small candidate cap to assess usefulness before increasing the sample.

Tips for useful sponsorship research

• Use queries specific to the product and buyer, such as “n8n CRM automation tutorial,” rather than just “technology.” Search relevance is evidence to review, not a verified niche label.
• Keep subscriber ranges broad enough for rounded counters near your cutoff.
• Increase recentVideos or the analysis window when channels publish infrequently. More returned videos still depend on what YouTube exposes in the first page.
• Use the same monitor for repeat observations. Do not run overlapping jobs for the same monitor.
• Review source videos, format, audience fit and recent activity manually before contacting a creator.

FAQ and support

Does it find every matching YouTube channel? No. Discovery and channel enrichment use bounded public-page samples, without exhaustive continuation pagination. Search results can vary by time and region. No minimum number of matching creators is guaranteed.

Does the country setting filter audience demographics? No. It is a search localization hint. The Actor uses English source labels for consistent parsing and does not infer a creator's spoken language or audience demographics.

Does it scrape email addresses or private analytics? No. It does not log in, solve email-reveal challenges, access YouTube Studio, or collect hidden contact details. Public profile links are included when visible.

Why are some dates, Shorts statistics or growth values missing? YouTube may omit dates or counters from a public listing. Unknown values remain unknown; first observations have no measured growth. Relative dates and abbreviated counts are approximate and explicitly marked.

Why can a run fail or return fewer channels? Source layouts, access restrictions, proxy availability, insufficient dated videos and the configured filters can reduce the result set. Check SUMMARY and each profile's warnings.

Use collected data in accordance with applicable law and the source platform's terms. For a reproducible problem or a custom research request, open an issue from this Actor's Issues tab and include the run ID and input with secrets removed.