Workflow Overview #
The Short Video Maker workflow runs as a three-stage content system inside n8n. Each stage depends on the output of the previous one. The workflow is designed to separate content planning, video production, and publishing into clear operational blocks so that the full process can run in an organized and repeatable way.
At a high level, the workflow does this:
- generates new video ideas
- turns one saved idea into a complete short video
- publishes the finished video to YouTube and updates the tracking sheet
Stage 1: Idea Generation #
Purpose #
The first stage exists to keep the content pipeline filled with new ideas that can later be turned into videos.
What happens #
When the idea generation trigger runs, the workflow sends a prompt to the AI idea generator. That AI returns a batch of 20 new video ideas. These ideas are then separated into individual items and added one by one into the Google Sheet.
Result #
Each generated idea becomes a new row in the tracking sheet and is stored as a future content item waiting for production. At this point, the row is saved with a production status showing that it has not yet been produced.
Why this stage matters #
This stage ensures the workflow always has a backlog of ideas available. Without it, the video creation section would have nothing to work with.
Stage 2: Video Creation #
Purpose #
The second stage takes one saved idea from the sheet and converts it into a complete short video.
Step 1: Select the next idea #
When the video creation trigger runs, the workflow checks the Google Sheet and selects the first row that is still marked Not Produced.
Step 2: Create the project folder #
After selecting the idea, the workflow creates a dedicated project folder in Google Drive for that row. Inside this folder, it creates the internal folder structure needed for production, including image and voice storage. This keeps each video project separated and organized using the sheet ID.
Step 3: Generate the creative assets #
The selected idea is then sent to the main AI agent. That AI generates the core creative package for the video:
- Title
- Six image prompts
- Sound prompt / narration script
- Caption
These outputs become the blueprint for the rest of the video creation process.
Step 4: Generate and store the narration audio #
The workflow takes the generated sound prompt and sends it to the audio generation step. The narration audio is created and stored in the voice folder. It is also uploaded publicly through Cloudinary so it can be used later in the video rendering step.
Step 5: Generate and store the images #
At the same time, the workflow takes the six image prompts and splits them into separate items. It loops through them one by one and performs the same process for each prompt:
- generate the image
- resize the image to the required vertical short-video format
- upload the image to the Google Drive image folder
After all images are created, the workflow collects the file links so they can be passed into the render step.
Step 6: Calculate image timing #
The workflow checks the total duration of the generated narration audio and divides that total equally across the six images. This gives each image the same amount of screen time.
Step 7: Render the final video #
Once the assets are ready, the workflow gathers:
- the six image links
- the narration audio link
- the per-image duration value
It then sends that data to JSON2Video, which renders the final short video. After a waiting period, the workflow fetches the rendered video, downloads it, uploads it into the project folder in Google Drive, and stores the final video link in the Google Sheet.
Stage 3: Save the Finished Video Metadata #
Purpose #
Before publishing can happen, the workflow must update the Google Sheet row with all finished video information.
What is saved #
Once the video creation stage is complete, the workflow updates the same row with:
- Title
- Caption
- Cover Image
- Video URL
- Production Status = Produced
- Posting Status = Not Published
Why this matters #
At this point, the row is no longer just an idea. It becomes a completed content item that is ready for the publishing stage. The sheet now contains everything the workflow needs for the final upload step.
Stage 4: Publishing #
Purpose #
The publishing stage takes the next completed video and uploads it to YouTube.
Step 1: Select the next unpublished video #
When the publishing trigger runs, the workflow searches the Google Sheet for the first row where Posting Status = Not Published.
Step 2: Download the final video #
The workflow uses the saved Video URL from that row to download the completed short video.
Step 3: Upload to YouTube #
The workflow then uploads the video using the YouTube node. During this step, it uses the saved:
- Title
- Caption
from the same row as the upload metadata.
Step 4: Update publishing status #
After the YouTube upload succeeds, the workflow gets the current date and time and updates the same row in Google Sheets by:
- changing Posting Status from Not Published to Published
- saving the Published Date
This completes the full lifecycle of that content item.
Workflow Logic Summary #
The easiest way to understand this workflow is to think of it as a content assembly line:
| Stage | Main Job | Output |
|---|---|---|
| Idea Generation | Create and save new ideas | 20 new idea rows in Google Sheets |
| Video Creation | Turn one idea into a video | finished short video + saved metadata |
| Publishing | Upload finished video to YouTube | published video + updated sheet status |
Because every step is tracked in the same Google Sheet, the workflow always knows:
- what ideas are still waiting
- what videos are already produced
- what finished videos are still waiting to be published
Key Dependency Rule #
The three stages must always happen in the correct order:
- ideas must exist before a video can be created
- a video must be created before it can be published
- the Google Sheet is the system that tells the workflow what comes next
If one stage is skipped or the tracking data is incorrect, the next stage may fail or find nothing to process.
In Simple Terms #
This workflow first fills the idea list, then converts one idea into a finished short video, and finally publishes the next finished video. It automates the full path from planning to publishing while using Google Sheets as the central tracking system for the entire pipeline.
