Urgent.News

What's breaking now, across thousands of outlets.

Tech

I mapped a YouTube channel's entire catalog in two runs — for under a cent

Competitor research on YouTube usually dies in one of two places: you click through forty videos by hand and give up, or you start writing a scraper and give up slightly later, somewhere around the third rate-limit. I needed a channel's full catalog — titles, URLs, and the view numbers that show which topics actually land — for a content teardown. Instead of building anything, I ran two existing…

Competitor research on YouTube typically ends after either manually viewing forty videos or attempting to write a scraper, often failing after the third rate limit. To obtain a channel's full catalog, I utilized two existing tools sequentially, which took less than a minute and cost under a cent. Here's the process:

Step 1: Enumerate the catalog quickly and inexpensively. Running the first tool with a 30-video cap for @OpenAI took about 7 seconds and returned 30 rows. The output included videoId, title, URL, channel name, and channel ID, but excluded view counts, durations, and publish dates. This separation of tasks allows for a nearly free enumeration pass, enabling affordable enumeration on multiple competitors.

Step 2: Enrich only the necessary information. From the 30 rows, I input three video URLs into the companion details tool. This step revealed view counts, durations, and likes. For instance, "Cyber at the Frontier: Introducing Daybreak Red Team & Daybreak Defense Factory" had 51,193 views, a 34-minute duration, and 70 likes. The enrichment process encountered gaps, such as a row lacking like counts, as public pages might not consistently provide all metrics.

The two-step approach is beneficial because enriching during enumeration would cost more for a large channel. Instead, enumerating all videos is inexpensive, giving a full topic map of published content, frequency, and titles. This allows for local sorting and filtering based on keywords, series, or recency, followed by enriching the shortlisted rows with per-video costs.

This pattern can be applied to other platforms like TikTok or search results. The cost-effective strategy also applies to evaluating any data tool by assessing enumeration pass costs before considering detail pass returns.

Both tools are accessible as Apify actors, requiring no coding. By inputting a channel handle and setting a cap, the channel catalog can be downloaded as JSON or CSV using https://apify.com/bin_ai_tools/youtube-channel-videos. Video details can be accessed via https://apify.com/bin_ai_tools/youtube-video-details. Automating these steps through the Apify API provides a straightforward integration into scripts or scheduled jobs, such as a weekly competitor-catalog comparison, which can be achieved in about twenty lines of code.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

RunsOn vs Blacksmith for GitHub Actions runners

RunsOn vs Blacksmith is a choice between two cost models rather than two rate cards: RunsOn is a flat annual license from 300 euros that installs into your own AWS account and leaves the compute on…

  • RunsOn pricing model costs $29/month for 10,000 Linux minutes
  • Blacksmith charges $0.004 per Linux minute with 3,000 free minutes/month
  • Latchkey offers Linux x64 pricing from $0.0025 to $0.0025 per minute

PNG vs JPEG for PDF pages: the chart page agreed, the text page didn't

"Save text as PNG, it's smaller than JPEG" is a rule of thumb for storing PDF pages as images, whether that's lecture slides for revision notes or a paper's figures.

  • PNG generally smaller than JPEG for chart page (127,051 bytes vs 178,232 bytes)
  • Dense text page 5% larger in PNG (838,404 bytes vs 796,608 bytes)
  • Lossless WebP smallest file for both text and chart pages

I Generated 76 OBD-II Pages. Two of Them Are the Same Page.

Small car-specs site, generated pages, one per diagnostic trouble code. The build script decides what goes in. Nobody reads the output. So I read it.

  • 76 pages generated for diagnostic trouble codes
  • P0303 and P0304 pages are duplicates
  • Pages sourced from SAE J2012/ISO 15031-6 standard

What we write down after every database restore drill

A restore drill that passes tells you that one backup restored on one day. That's worth knowing, but most of what drills can teach you only shows up when you compare them: restore times that creep up…

  • Restore drill logs record date, person, backup, target environment, and restore time.
  • Logs check success, table counts, row counts, app connectivity, and file integrity.
  • Key insight: dump size doesn't predict restore time; custom-format includes index rebuild.

More from Sunday 15 November →