YouTube Channel Scraper in Python: Filter Active Creators

A Python script for scraping YouTube channels can extract channel names, subscriber counts, and URLs. However, it doesn't determine if a channel is a suitable candidate for outreach. The costly error is overlooking a channel that ceased publishing, no longer generates views, utilizes an inappropriate content format, or lacks a public business contact method. This guide creates a Python workflow that outputs active creator prospects rather than an unfiltered channel list.

First, define what constitutes an "active" channel. The minimum criteria might include at least one video per month, a public upload within the last 60 days, a median view count of 1,000 across a recent sample, between 5,000 and 250,000 subscribers, and content that matches the intended offer. These thresholds may vary depending on the specific requirements of the stakeholder. Before examining the results, establish the rule beforehand.

To begin, install the Apify client with the command: pip install apify-client. Load the API token from an environment variable to prevent committing it to the script or Git repository: export APIFY_API_TOKEN = ...

Next, initiate a targeted channel search using the following parameters:

- Niches: AI automation or B2B SaaS

- Languages: English (en)

- Countries: United States (US), Canada (CA), United Kingdom (GB)

- Minimum subscribers: 5,000

- Maximum subscribers: 250,000

- Minimum videos per month: 1

- Last upload within: 60 days

- Minimum recent median views: 1,000

- Content formats: long-form or Shorts

- Require public business contact: No

- Maximum channels: 10

- Videos analyzed per channel: 10

- Maximum search calls: 12

Run the actor 'kazkn/youtube-creator-lead-finder' with these run_input parameters. The actor identifies prospects from the specified niches and queries. Channel URLs are optional. It eliminates duplicates based on channelId and generates a dataset with one entry per qualified channel.

Inspect recent performance rather than lifetime totals for each returned creator. Print the following useful review fields:

- Channel name

- Channel URL

- Subscriber count

- Last upload date

- Estimated videos per month

- Recent median views

- Opportunity score

- Opportunity reasons (if any)

- Warnings (if any)

Median recent views are particularly valuable as a single viral upload can skew an average. Remember that the sample remains bounded, so keep the warnings and avoid presenting the metric as a complete channel history.

Match the content format to the offer by considering the specific ratio metrics relevant to the service. For example, a long-form editing service may prioritize longFormRatio, while a vertical-video agency may focus on shortsRatio, and a webinar or streaming product may emphasize liveRatio. The output also includes a format confidence value, and Shorts classification relies on a duration heuristic from the public API, so uncertain cases should be clearly identified.

Lastly, separate geography and language evidence in the search results. Fields such as declaredCountry, inferredCountries, countryConfidence, primaryLanguage, detectedLanguages, and languageConfidence provide distinct information. If the brief requires French-speaking creators worldwide, leave the countries field empty and set worldwideLanguageMode to True. Do not assume the search region matches the creator's declared location without explicit confirmation.