YouTube video description and SEO metadata from transcripts

Creating video content has become essential for most digital strategies, yet the administrative work that follows remains tedious. After uploading a YouTube video, you need to write a compelling description, craft SEO metadata, generate social media snippets, and sometimes create show notes. Most creators do this manually, copying transcript text between tools and rewriting content several times over.

This workflow solves that problem by automating the entire journey from transcript to optimised YouTube metadata. You'll feed a video transcript into a system that summarises it, generates multiple description variations, and produces SEO-optimised metadata, all without touching a single tool yourself.

The beauty of this particular combination is that it requires no coding experience and uses three focused tools that do one thing well each. Copy.ai writes persuasive descriptions, Quick Creator handles SEO optimisation, and Resoomer.ai extracts the essential points. By connecting them through an orchestration platform, you eliminate the manual copying and pasting entirely.

The Automated Workflow

We'll build this workflow using Make (formerly Integromat) as the orchestration layer, though Zapier and n8n work equally well. I'll show you Make first, then mention specific differences for the other platforms.

Prerequisites

Before starting, you'll need:

• A Copy.ai account with API access enabled
• A Quick Creator account (they offer API functionality through their enterprise plan or via Zapier integration)
• A Resoomer.ai account
• A Make.com account with at least a Standard plan (£10/month) to use multiple tools
• A YouTube channel where you control the video uploads

The Overall Flow

Here's how the data moves:

1. YouTube webhook or scheduled trigger sends a video URL or transcript text
2. Resoomer.ai summarises the transcript into key points
3. Copy.ai uses the summary to generate multiple description variations
4. Quick Creator optimises one variation and outputs SEO metadata (title suggestions, tags, keywords)
5. The final output lands in a Google Sheet or sends directly to your email

Setting Up the Make Workflow

Log into Make.com and create a new scenario. You'll see a blank canvas where scenarios are built from left to right.

Step 1: Trigger - Receiving the Transcript

The easiest start is a webhook that receives your video transcript. When you upload to YouTube, you can manually copy the transcript and send it via a simple form, or set up a YouTube automation trigger if you have Make's YouTube module.

For now, assume you're using a webhook. In Make, add a Webhook module as your trigger:

Copy
Webhook Module Settings:
- Choose "Watch Events"
- Copy the unique webhook URL
- Set it to accept JSON data

Example payload your application will send:
{
  "video_title": "How to Master Content Marketing",
  "video_url": "https://youtube.com/watch?v=...",
  "transcript_text": "Today we're going to discuss content strategy. First, let's cover the basics..."
}
``` For more on this, see [AI-Powered Content Marketing Pipeline](/blog/ai-powered-content-marketing-pipeline).

Make automatically parses this JSON, so you can reference `video_title`, `video_url`, and `transcript_text` in later steps.

**Step 2: Summarise with Resoomer.ai**

Add a new module and search for "HTTP" (since Resoomer.ai doesn't have a native Make connector, you'll use their REST API):

HTTP Module Configuration: URL: https://api.resoomer.com/summarize Method: POST Headers: Content-Type: application/json Authorization: Bearer YOUR_RESOOMER_API_KEY

Body (JSON): { "text": "{{1.transcript_text}}", "language": "en", "summary_level": 3 }

Copy
The `{{1.transcript_text}}` refers to the transcript from Step 1. The `summary_level` parameter controls brevity; 3 gives you a concise bullet-point summary.

Resoomer returns a response like this:

```json
{
  "summary": "Content strategy is essential. Key points include audience research, consistency, and multi-channel distribution. Most creators fail at one of these three.",
  "status": "success"
}

Map this response in Make so you can use the summary text in the next step.

Step 3: Generate Description Variations with Copy.ai

Copy.ai offers a powerful API that accepts brief briefs and generates multiple variations. Add another HTTP module:

Copy
HTTP Module Configuration:
URL: https://api.copy.ai/api/v1/projects/generate
Method: POST
Headers:
  Authorization: Bearer YOUR_COPY_AI_API_KEY
  Content-Type: application/json

Body (JSON):
{
  "template_id": "youtube_description",
  "inputs": {
    "video_title": "{{1.video_title}}",
    "summary": "{{2.summary}}",
    "target_audience": "marketing professionals",
    "tone": "informative"
  },
  "num_variations": 3
}

The Copy.ai API returns three different YouTube descriptions based on your summary. Store all three in Make's internal memory so you can pick the best one.

If Copy.ai's API template_id doesn't match exactly, check their API documentation for the correct endpoint. Some accounts may use a more generic /generate endpoint:

Copy
URL: https://api.copy.ai/api/v1/generate
Method: POST

Body:
{
  "prompt": "Write a YouTube video description for a video titled '{{1.video_title}}'. The video summary is: {{2.summary}}. The audience is marketing professionals. Write it in an informative tone. Generate 3 variations separated by '---'.",
  "max_tokens": 500
}

Step 4: Optimise for SEO with Quick Creator

Quick Creator specialises in SEO metadata. Since their API support is limited, you have two options: use their Zapier integration (which Make can trigger) or manually use their web interface. For full automation, we'll show the Zapier bridge approach.

Actually, if you're already in Make, a better approach is to use Quick Creator's direct API if available, or use a HTTP module to trigger their system:

Copy
HTTP Module Configuration:
URL: https://api.quickcreator.com/api/v1/seo-optimize
Method: POST
Headers:
  Authorization: Bearer YOUR_QUICK_CREATOR_API_KEY

Body:
{
  "content": "{{3.variations[0]}}",
  "title": "{{1.video_title}}",
  "target_keywords": ["content marketing", "strategy", "audience"],
  "content_type": "youtube_description"
}

This returns optimised metadata including:

Copy{
  "seo_score": 87,
  "suggested_tags": ["content marketing", "marketing strategy", "creator tips"],
  "suggested_title": "How to Master Content Marketing: 5 Proven Strategies",
  "keyword_density": {
    "content marketing": 0.08,
    "strategy": 0.12
  },
  "word_count": 156
}

Step 5: Output to Google Sheets or Email

Add a final module to capture everything. Most teams want results in Google Sheets for review before publishing:

Copy
Google Sheets Module:
Action: Add a row
Spreadsheet: [Your spreadsheet ID]
Sheet: YouTube Metadata

Columns to populate:
- Video Title: {{1.video_title}}
- Original Transcript Summary: {{2.summary}}
- Description Option 1: {{3.variations[0]}}
- Description Option 2: {{3.variations[1]}}
- Description Option 3: {{3.variations[2]}}
- SEO Score: {{4.seo_score}}
- Suggested Tags: {{4.suggested_tags}}
- Suggested Title: {{4.suggested_title}}
- Processing Date: NOW()

Alternatively, send via email:

Copy
Email Module:
To: your@email.com
Subject: YouTube Metadata Ready: {{1.video_title}}
Body:
Video: {{1.video_title}}

Best Description ({{4.seo_score}} SEO score):
{{3.variations[0]}}

Tags: {{4.suggested_tags}}
Keyword Density: {{4.keyword_density}}

Review and publish when ready.

Testing Your Workflow

Before going live, test with a real YouTube transcript. In Make's testing interface, paste sample data into your webhook step. Make will simulate the entire workflow and show you the output at each stage. Check that:

• The summary captures key points without being truncated
• All three descriptions read naturally
• SEO suggestions match your content
• The final output format is what you expected

Adapting for Zapier or n8n

If you're using Zapier instead:

• Zapier has native integrations for Copy.ai and Quick Creator, making this simpler
• Search "YouTube" for transcript triggers; some exist but may require manual transcript input
• Zapier charges per task, so estimate volume before choosing your plan

For n8n (self-hosted):

• Download and run n8n locally or on a server
• Use HTTP Request nodes exactly as shown above
• n8n is free but requires more technical setup
• You get unlimited tasks, making it best for high-volume channels

The Manual Alternative

Some creators prefer human control over each step. You can run this process manually in about 15 minutes per video:

1. Upload your YouTube video and copy the auto-generated transcript
2. Paste it into Resoomer.ai and click Summarise
3. Copy the summary and paste it into Copy.ai with your video title
4. Generate descriptions and pick your favourite
5. Paste that description into Quick Creator's web interface
6. Copy the suggested tags and metadata into YouTube directly

This works fine if you upload once a week. For channels posting daily, the automation pays for itself immediately in time saved.

Pro Tips

Tip 1: Batch Processing for Efficiency

If you upload multiple videos, batch them into a single workflow run. Modify your webhook to accept an array of transcripts, then loop through each one. This reduces API calls and costs significantly.

Copy
Modified webhook body for batch processing:
{
  "videos": [
    {
      "video_title": "Video 1",
      "transcript_text": "..."
    },
    {
      "video_title": "Video 2",
      "transcript_text": "..."
    }
  ]
}

Then use Make's Iterator module to process each video in sequence.

Tip 2: API Rate Limits

Copy.ai allows 100 requests per day on their free plan. If you're posting more than one video daily, upgrade to a paid plan or add delays between requests using Make's "Sleep" module. Resoomer.ai has a 10 requests per minute limit; again, the Sleep module prevents hitting this.

Copy
Sleep Module Settings:
Duration: 15 seconds
(Add this between Resoomer and Copy.ai steps for safety)

Tip 3: Handle Missing Data Gracefully

Sometimes transcripts are incomplete or API calls fail. Add error handlers in Make:

Copy
For each HTTP module, add a conditional route:
- If Status = 200, continue normally
- If Status ≠ 200, send alert email and stop

This prevents garbage data reaching your Google Sheet.

Tip 4: Monitor Costs Before Full Rollout

Test your workflow 5 times and note the costs. Copy.ai charges per request, Quick Creator per analysis, and Resoomer.ai per summarisation. Make charges per workflow execution. Run the numbers for your posting frequency before automating everything.

Tip 5: Customise Prompts for Your Brand

The descriptions generated by Copy.ai are generic unless you customise the prompt. In Step 3, add your brand voice to the inputs:

Copy
"brand_guidelines": "We use conversational tone with questions. We always link to related resources. We avoid jargon."

This makes generated descriptions feel authentically yours rather than AI-generic.

Cost Breakdown

If you're posting once daily, consider the annual cost of £827.88 against the time savings. Most creators processing one video per day recover costs within a fortnight.

This workflow transforms a 20-minute manual task into a 2-minute one. You focus on creating great content; the system handles the administrative metadata work. Start with one video, refine the prompts based on results, then scale to your full publishing calendar.